Elon Musk's Grok AI accounted for 87% of AI-generated files connected to deepfake attacks during the first half of 2026, according to a new report from Resemble AI.

Researchers examined 1,760 news reports and identified 821 separate attacks involving at least 15,736 documented victims. Together, those incidents included approximately 3.46 million synthetic images, videos and audio files. Resemble AI’s dataset is based on publicly reported incidents, and its percentage applies only to files that researchers could count and attribute to a particular tool.

Even with that caveat, the figure shows how quickly a widely available AI image generator can be turned into a tool for misuse and abuse.

The findings arrive as Elon Musk places a large financial bet on Grok. Following SpaceX’s acquisition of xAI, Musk said during a company-wide meeting that SpaceX’s AI revenue would surpass revenue from all its other products as soon as September, according to Business Insider.

SpaceX reported $2.56 billion in AI revenue during its most recent quarter, putting Grok and the infrastructure behind it at the center of the company’s future. That makes questions about how the tool is being used difficult to separate from its rapid commercial growth.

One in six attacks involved explicit content

Resemble AI found that 137 of the 821 attacks — roughly one in six — involved non-consensual sexual imagery of adults or children.

Grok faced widespread criticism earlier this year after users began asking the chatbot to digitally undress people or place them in sexualized situations. Some of the images reportedly depicted minors, prompting investigations from regulators around the world.

The European Commission opened a formal investigation into X over sexualized images generated with Grok, while California Attorney General Rob Bonta launched an investigation into the creation and spread of non-consensual explicit material through the platform. Authorities in the UK, Australia and other countries have scrutinized the tool as well.

Resemble AI estimates that companies permitting or distributing the documented images could face as much as $2.24 billion in potential civil liability under US law. That number is an estimate, rather than a record of penalties or lawsuits already filed.

The verified direct financial losses identified by the researchers were much smaller, totaling $6.95 million. However, many of the incidents involved reputational damage, harassment or sexual exploitation rather than a straightforward financial scam.

Deepfake attacks are becoming harder to count

The company calculated that coverage of the attacks had a potential reach of 292.8 billion during the first half of the year, almost matching the 296.4 billion recorded across all of 2025.

Political deepfakes were among the most widely covered. Resemble AI identified 22 political disinformation incidents that each generated a potential media reach of more than 1 billion people.

Corporate deepfake fraud, on the other hand, has almost vanished from public view. Researchers found just 15 publicly reported corporate incidents, and 14 of those appeared in only one publication.

Resemble AI suggests companies may be reluctant to disclose successful attacks because doing so could attract regulatory attention, hurt their reputation or concern investors. That doesn’t mean corporate deepfake fraud has gone away. It may simply be happening outside the public eye.

A detector that checks whether a scene obeys the laws of physics

Resemble AI released the report alongside DETECT-World, a new system designed to recognize content created by AI generators it has never encountered before.

The system combines familiar AI signatures with an analysis of whether a scene makes physical sense. It can examine whether the lighting on someone’s face matches the visible light sources, whether shadows fall in the correct direction and whether objects in the foreground and background move consistently.

For video, the system processes four-second windows at 40 frames per second. It produces an overall probability that the file has been manipulated, scores showing when suspicious moments appear and a heat map identifying the affected part of the frame.

Resemble AI says DETECT-World achieved 95.8% accuracy on images and 98.2% on video, but both results come from internal benchmarks and are still awaiting external validation.

Bottom line

This report provides a good reminder that the technology to detect deepfakes is improving, but it is still chasing tools that can create and distribute them almost instantly.

Until that gap closes, we need to find the original source, look for confirmation from reputable outlets and avoid sharing suspicious content simply because it looks real.

More from Tom's Guide

• I put ChatGPT on my Ray-Ban Meta glasses — and it completely changed how I used them
• ChatGPT can do more for you than ever — and I use these 5 prompts to take everyday tasks off my plate
• 5 custom Gemini Gems that I swear save me hours each week

Elon Musk has never been one to shy away from ambitious promises, and his latest claim may be one of his boldest yet. In a recent post on X, the xAI founder said Grok is working on a full-length, AI-generated adaptation of "The Odyssey" that he described as "historically accurate."

It's a bold claim, for sure. AI video tools like Runway Gen-3 and Luma Dream Machine have already shown they can produce impressive short clips, but creating a coherent feature-length film is a much bigger challenge. If Musk's vision becomes reality, it would push AI video well beyond the short demonstrations we've seen so far.

Here's what we know — and why Musk's choice of the phrase "historically accurate" has sparked discussion.

The AI race heads to Hollywood

Until now, Grok has been best known as the chatbot built into X. An AI-generated movie would represent a major expansion of xAI's ambitions, suggesting the company wants to compete in AI video generation.

The choice of "The Odyssey" is particularly ambitious. Homer's epic follows Odysseus on his long journey home after the Trojan War, combining ancient Greek settings with legendary creatures, gods and mythological events. Re-creating that story would require AI to maintain consistent characters, environments and visual style across thousands of shots — something today's video models still struggle to do reliably.

Musk's description of the film as "historically accurate" also appears to be a response to the recent debate surrounding Christopher Nolan's adaptation of "The Odyssey," which grossed more than $300 million in its opening weekend.

Some of the film's casting choices — such as Academy Award-winner Lupita Nyong'o as Helen of Troy and Elliot Page as Sinon — sparked online arguments from right-wing commentators over whether modern adaptations should strictly follow Homer's descriptions of key characters. Musk was among those criticizing what he saw as departures from the original text.

At the same time, "The Odyssey" isn't a historical record. It's an ancient Greek epic that blends elements of the Bronze Age with mythology, including gods, monsters and supernatural events. So when Musk talks about 'historical accuracy,' he's likely referring to a faithful interpretation of Homer's descriptions and the story's ancient Greek setting, rather than suggesting the events themselves literally happened.

While AI can already generate remarkably realistic scenes lasting a few seconds, maintaining continuity across an entire feature film remains one of the industry's biggest technical hurdles.

What does 'historically accurate' actually mean?

Musk's choice of words doesn't appear to be accidental. Christopher Nolan's upcoming adaptation of "The Odyssey" has been at the center of online debate over several casting decisions, particularly the casting of Lupita Nyong'o. Critics, including Musk, argued that changing the appearance of mythological characters departed from Homer's descriptions, while others countered that "The Odyssey" is a work of mythology rather than literal history and has been reinterpreted countless times over the centuries.

The debate quickly expanded beyond the film itself, becoming part of a broader cultural argument over historical representation, artistic freedom and diversity in modern adaptations. Musk weighed in multiple times on X, criticizing what he viewed as unnecessary departures from the original text.

Against that backdrop, his promise that Grok will create a "historically accurate" version of "The Odyssey" looks more like a statement about how he believes the story should be portrayed.

Can Grok actually pull it off?

As with many Musk announcements, it's worth treating both the timeline and the scope of the project with some caution. Producing a feature-length movie isn't just about generating convincing visuals. It also requires a cohesive screenplay, believable dialogue, voice performances, music, editing and scene-to-scene continuity.

Musk hasn't shared many technical details about how Grok would handle those challenges, and xAI has yet to demonstrate a publicly available system capable of generating a full-length film.

Still, the announcement highlights how quickly AI video is evolving. Just a few years ago, generating even a few seconds of realistic footage seemed out of reach. Now, companies are openly talking about feature-length productions.

Whether Grok ultimately delivers on Musk's promise remains to be seen. But if it can produce a watchable, coherent feature-length film rather than a collection of impressive clips, it would represent one of the biggest milestones AI video has reached so far.

More from Tom's Guide

• I asked ChatGPT to improve my Apple Notes app — here are the 7 changes I actually kept
• Microsoft Copilot’s free Deep Research AI tool is going away — these are the 7 best free alternatives
• I always recommend the ChatGPT 'Escape Prompt' to friends looking for a new job — here's why it works

For the past few years, artificial intelligence has largely meant chatbots. Every few weeks, Anthropic, OpenAI and Google (among others) race to launch a smarter, faster model before the competition.

And while the simple interface helped launch the AI boom, it also made AI approachable for the average person to use. But one of the biggest AI deals of the year suggests the industry may already be moving beyond chatbots.

According to Reuters, SpaceX has agreed to acquire Cursor maker Anysphere in a deal valued at roughly $60 billion. While most chatbot users have never heard of Cursor, developers have embraced it as one of the most powerful AI coding assistants available today.

And that's exactly why this aqcuisition is a really big deal.

SpaceX is betting on agents

When ChatGPT launched in late 2022, the breakthrough wasn't just the technology. It was the interface, which made AI feel as simple to use as sending a text message.

The chat window became the default way we interacted with AI and even searched the web. Upon OpenAI's success, every major company followed suit. Google built Gemini, Anthropic built Claude and Microsoft integrated Copilot across its products. Even Elon Musk's xAI launched Grok.

Now, the most valuable AI tools are increasingly moving away from answering questions and toward completing tasks. That's where Cursor comes in. Unlike a traditional chatbot, Cursor isn't designed primarily for conversation, but rather built to help developers write, edit, understand and debug software. Instead of acting like an assistant you chat with, it becomes part of the workflow itself. The goal for AI agents is to help someone work more efficiently.

Why investors are paying attention

A $60 billion valuation sounds shocking until you look at where AI companies are actually making money. Users love free chatbots. Businesses will willingly pay for productivity solutions. That's exactly why coding assistants have become the hottest categories in AI.

Companies are discovering that AI tools capable of saving employees hours each week are often easier to monetize than general-purpose chatbots. If an AI coding assistant helps a developer complete projects faster, the return on investment becomes relatively easy to calculate.

Developers tend to be early adopters. They use AI all day, every day. And unlike casual chatbot users, they're often willing to pay significant subscription fees if the tools genuinely improve their productivity. The result is a growing market for AI systems that function less like conversational partners and more like coworkers.

The chat window is becoming a control panel

I've written before that the chatbot era may already be ending, even if most people haven't noticed yet. The reason is simply that the interface is becoming less important than the actions happening behind it.

OpenAI is building tools that can perform tasks on your behalf, Anthropic continues expanding Claude's project capabilities, and Google is embedding AI deeper into Search, Workspace and Android. And while the chat box still exists, it's inceasingly becoming a command center rather than the destination itself.

Coding assistants like Cursor are one of the clearest examples of the latest shift because software development is highly structured work. But the same idea is spreading into research, customer support, content creation, scheduling, shopping and countless other tasks.

What this means for everyday users

If you're wondering why you need Cursor right now, don't worry. Most people won't start using Cursor tomorrow. But, it is worth paying attention to why this acquisition happened. It says a lot about the way opportunities for AI are shifting towards building entire systems that can handle large chunks of work with minimal supervision.

For employees who currently use chatbots for answers, SpaceX's purchase highlights the fact that AI might be closer to becoming a coworker. It will be interesting to see how the next phase of the AI race takes on a new face as big tech races to build the most useful agent.

More from Tom's Guide

• That dead phone in your drawer could power the next AI data center
• I use the ChatGPT 'Oak Tree' prompt when life gets chaotic — here's how it keeps me grounded
• I asked ChatGPT to make cleaning less overwhelming — these 7 prompts actually helped

"The Elon Musk Show: The Next Chapter" follows the story of the world's richest person over the past four years. Going from the Tesla chief's takeover of Twitter to his recent step in U.S. politics, the documentary explores Musk's rise to power, with archive footage, plus testimony from childhood friends, political insiders and expert commentators.

Here's how to watch "The Elon Musk Show: The Next Chapter" online from anywhere with this VPN — and potentially for free.

Following on from 2022's "The Elon Musk Show", this new feature-length documentary will dissect the latest twists and turns of the mega-rich businessman's life.

During those four years, of course, he's bought the social media platform Twitter and rebranded it entirely with his X trademark. He's also worked closely with U.S. President Donald Trump, taking on the role of Senior Advisor to the President and also becoming the de facto head of the Department of Government Efficiency (DOGE) before a public fallout. Amid all that, his SpaceX operation has continued to push the boundaries of space travel.

Whatever your opinions on Elon Musk and his role in the culture wars, there's no denying this is one of the most intriguing and influential people in the world. To understand Musk is to better understand our culture and the world around us.

So, get ready to dive into this enthralling BBC documentary — here's how to watch "The Elon Musk Show: The Next Chapter" online and from anywhere.

How to watch 'The Elon Musk Show: The Next Chapter' from anywhere

Thanks to the wonders of a VPN (Virtual Private Network), "The Elon Musk Show: The Next Chapter" should be available to Brits no matter where they are. The software allows your devices to appear to be back in the U.K. and watch iPlayer for free, as if you were back home.

Not all VPNs works for this – but NordVPN does (we use it daily) and you can get an exclusive offer below:

Using a VPN is incredibly simple.

1. Install the VPN of your choice. As we've said, NordVPN is our favorite.

2. Choose the location you wish to connect to in the VPN app. For instance, if you're away from the U.K. and want to view a U.K. service, you'd select U.K. from the list.

3. Sit back and enjoy the show. Head to BBC iPlayer to watch "The Elon Musk Show: The Next Chapter" online and on-demand.

How to watch 'The Elon Musk Show: The Next Chapter' in the U.K.

"The Elon Musk Show: The Next Chapter" will be available to stream for free on BBC iPlayer once it has aired on standard television.

If you'd rather watch it as soon as possible, you can catch "The Elon Musk Show: The Next Chapter" on BBC Two. It'll be playing at 9 p.m. BST on Thursday, May 21.

Outside the U.K.? You don't have to miss this show if you are a Brit living abroad because you can unblock BBC iPlayer with a VPN.

How to watch 'The Elon Musk Show: The Next Chapter' around the world

Can you watch 'The Elon Musk Show: The Next Chapter' in the U.S.?

At the moment, it doesn't look like there will be a streaming release for "The Elon Musk Show: The Next Chapter" in the U.S. as it's not going to be readily available on any streaming service.

You usually find that BBC content is often later released on BritBox, so that may be an option in the future. Given the subject of this documentary, we'd be surprised if it didn't make its way to the States.

A Brit on vacation in the States? To watch the documentary on BBC iPlayer, you'll need a VPN like NordVPN to access your regular service without being geo-blocked.

Can you watch 'The Elon Musk Show: The Next Chapter' in Australia?

It's a similar issue in Australia. Sometimes, British television content makes its way over to the other side of the world, but distribution plans for "The Elon Musk Show: The Next Chapter" are still under discussion.

Still, you can watch the 2022 documentary on SBS On Demand if you want to at least get a taste of the Elon Musk story. Of course, it's possible that "... The Next Chapter" will later become available on SBS On Demand.

If you’re a Brit working or on holiday Down Under and want to keep using your usual U.K. streaming services, you can still catch the show with a VPN like NordVPN.

'The Elon Musk Show: The Next Chapter' — all you need to know

'The Elon Musk Show: The Next Chapter' trailer

'The Elon Musk Show: The Next Chapter' - synopsis

Another look at one of the world’s most polarizing billionaires as he swaps rockets and electric cars for raw political power. The program covers Musk’s tumultuous takeover of Twitter/X, the seismic fallouts of his personal life, and his evolution into a political force on both sides of the Atlantic.

With immersive archive footage and revealing testimony from childhood friends, political insiders and former employees, this documentary charts how his quest to 'save humanity' collides with culture wars, AI breakthroughs and far-right movements in the US and UK.

More from Tom's Guide

• The best BBC iPlayer VPN in 2026
• These are the 3 best streaming devices we've tested for your TV
• How to watch 'David Attenborough's 100 Years on Planet Earth' online

We test and review VPN services in the context of legal recreational uses. For example:1. Accessing a service from another country (subject to the terms and conditions of that service).2. Protecting your online security and strengthening your online privacy when abroad.We do not support or condone the illegal or malicious use of VPN services. Consuming pirated content that is paid-for is neither endorsed nor approved by Future Publishing.

Elon Musk has officially lost his high-profile lawsuit against OpenAI and CEO Sam Altman — marking one of the biggest legal moments in the modern AI race so far.

According to Reuters, a California jury unanimously ruled in favor of OpenAI after Musk accused the company of abandoning its original nonprofit mission and prioritizing profit over humanity. The verdict reportedly came after less than two hours of deliberation.

The case centered around Musk’s claim that OpenAI’s transformation from a nonprofit research lab into a commercial AI powerhouse violated the company’s founding principles. Musk had also challenged OpenAI’s deep partnership with Microsoft and sought enormous financial damages.

Instead, the jury concluded Musk waited too long to file the lawsuit, effectively ending one of the most closely watched legal battles in tech.

But the real story here may be much bigger than the courtroom itself.

Beyond the lawsuit — a battle over who controls AI

At its core, this case represented two completely different visions of artificial intelligence.

Musk has repeatedly warned that advanced AI could become dangerous if controlled by profit-driven companies. OpenAI, meanwhile, argues that building cutting-edge models requires enormous amounts of capital, infrastructure and commercial partnerships.

The trial reportedly included testimony from major Silicon Valley figures, including Altman, OpenAI President Greg Brockman and Microsoft CEO Satya Nadella. It also exposed years of internal disagreements, failed takeover attempts and power struggles dating back to OpenAI’s earliest days.

Ironically, Musk is no longer just an outsider criticizing OpenAI. Through his own company, xAI, he’s now competing directly in the same AI arms race he once warned about.

Why this matters more than ever right now

This ruling arrives during one of the most aggressive periods of AI expansion the tech industry has ever seen. Companies are pouring billions into data centers, AI assistants, reasoning models and agentic software. OpenAI itself is reportedly exploring a future valuation approaching $1 trillion.

A Musk victory could have dramatically disrupted that momentum. Instead, this outcome removes a major cloud hanging over OpenAI. It gives Altman a significant public win at a time when the company is rapidly expanding ChatGPT into everything from search to personal finance tools.

It also sends a message to the rest of Silicon Valley: the courts may not slow the AI race down anytime soon.

The bottom line and bigger issue nobody solved

Even though Musk lost, the trial surfaced questions that the AI industry still hasn’t answered. Who should control superintelligent AI systems? Can nonprofit ideals survive once billions of dollars are involved? And perhaps the even bigger question is, is it even possible to build frontier AI safely without becoming one of the most powerful corporations on Earth?

The jury may have decided the legal fight for now, but the philosophical one is just beginning.

More from Tom’s Guide

• I almost gave up on Apple TV’s No. 1 show — then ChatGPT convinced me to keep watching
• I asked ChatGPT to apply Lewis Howes’ ‘Greatness’ mindset to my life — and it completely changed how I approach work
• I gave ChatGPT permission to disagree with me with this prompt — and its responses became dramatically better

When we upload photos, stream movies and search Google, we often don't think about the consequences. "The cloud" has made the internet feel weightless but AI is making the physical side of the internet impossible to ignore.

As useful as it may be, AI does not just run on software. It runs on land, chips, power plants, cooling systems, transmission lines and data centers so massive they could be the size of 2,000 Walmarts. That's why some communities are now pushing back before these data centers are even built.

That may help explain why one of the strangest-sounding ideas in tech suddenly feel a lot less absurd. For one, putting AI data centers in space.

According to Reuters, Google is reportedly in talks with Elon Musk’s SpaceX about launching orbital data centers as part of Project Suncatcher, Google’s effort to test solar-powered, satellite-based AI cloud infrastructure using its own Tensor Processing Units.

Yes, AI is getting so demanding that the answer is apparently....rockets? But the more you look at the problem, the more revealing the idea becomes.

AI is running into the real world

The unfortunate side of the AI boom is that AI’s appetite for power, cooling and compute has become so extreme that space is now being discussed as a serious part of the infrastructure conversation.

That's because AI has been sold to us as software. But every query, every prompt has to run somewhere. Behind the scenes, AI depends on enormous data centers packed with specialized chips. Those chips use huge amounts of electricity and generate heat that has to be managed. As AI models become more powerful and more widely used, the demand for data center capacity is growing fast.

According to the IEA’s "Electricity 2024 – Analysis and Forecast to 2026" report (and subsequent updates in May 2026), global data center electricity consumption is projected to exceed 1,000 terawatt-hours (TWh) by the end of 2026. This figure is roughly equivalent to the entire annual electricity consumption of Japan, which currently ranks as the world’s fourth-largest economy.

Communities are raising concerns about electricity demand, water use, emissions, land use and rising power bills. The fight over massive AI data center projects in places like Utah shows how quickly “the cloud” can become a local zoning, water and environmental issue.

Why space is appealing

The basic appeal of orbital data centers is solar power. That's because in space, satellites can collect sunlight more consistently than solar panels on Earth. There are no clouds, no nighttime in the same way and no local fights over using thousands of acres of land for power infrastructure.

Another perk is no wildlife. Data centers can disrupt wildlife by converting large stretches of habitat into industrial infrastructure, and in the case of Utah’s proposed Stratos project, critics have warned that the 40,000-acre site near the Great Salt Lake could bring major ecological impacts that could include increasing the average temperature dramatically.

Google’s Project Suncatcher is built around solar-powered satellites equipped with AI chips, linked together to create a kind of orbital AI cloud. The company has described its upcoming early-2027 mission as a learning test designed to see whether its hardware can survive and function in orbit.

And while Elon Musk may be the most visible figure talking about AI data centers in space, he is no longer the only one. . NVIDIA has announced space-focused AI computing platforms, including its NVIDIA Space-1 Vera Rubin Module, which it says is designed to bring AI compute to orbital data centers, geospatial intelligence and autonomous space operations.

Meanwhile, Aetherflux, a startup originally focused on space-based power beaming, reportedly rebranded this week as Cowboy Space,and is now pitching a constellation of orbital AI data centers. Its approach appears to involve turning the upper stage of its planned rockets into the data center itself, rather than launching a separate dedicated satellite. Taken together, these moves suggest the space-based AI infrastructure race is no longer just a Musk moonshot — it may be the start of a much broader orbital data center boom, or even an early-stage bubble.

For AI companies, the dream is obvious

Just last week, Anthropic reportedly agreed to use SpaceX’s Colossus 1 infrastructure and explicitly stated interest in gigawatts of orbital compute.

It’s a signal that the industry is hitting a physical wall; if space-based data centers can be scaled, they solve the 'impossible' constraints currently choking Earth-based facilities, including:

• Less pressure on local power grids
• Less land conflict
• Less dependence on water-heavy cooling
• More access to continuous solar energy
• Fewer battles with communities that do not want giant data centers nearby

Everyone wants faster AI, better assistants and more powerful tools. Far fewer people want the infrastructure required to make those tools possible built near their homes.

Space creates new problems

Data centers on Earth already have to manage enormous amounts of waste heat. In space, there is no air to carry heat away. That means orbital data centers would need highly advanced thermal systems, likely including large radiators, to prevent the hardware from overheating.

That means, an enormous cost. The Wall Street Journal reported that launch costs remain a major obstacle, with current costs around $3,400 per kilogram and experts suggesting that space-based data centers would likely need launch costs closer to $200 per kilogram or less to make economic sense.

There are other problems, too, such as radiation, satellite maintenance, high-speed communication, space debris, orbital congestion and the challenge of manufacturing and launching hardware at a scale that would actually matter for AI.

Astronomers are also worried. Space.com has reported that scientists have raised concerns about SpaceX’s proposed plan for a massive constellation of orbiting AI data centers, warning that large numbers of bright satellites could interfere with astronomy and permanently change the night sky.

So, no, space data centers are not an easy fix. While they may solve one set of problems, they could create others.

Musk’s idea shows how extreme the AI race has become

Musk’s space data center idea shows how far the AI race means to the U.S. The companies that dominate AI may not simply be the ones with the smartest models. They may be the ones with the most compute, the most power, the best chips, the biggest data centers and the ability to build infrastructure faster than everyone else.

According to Financial Times, SpaceX has reportedly struck major AI infrastructure deals, including a deal for Anthropic to use compute capacity at SpaceX’s Colossus 1 data center in Tennessee. The Financial Times reported that the deal gives Anthropic access to more than 300 megawatts of compute capacity as AI companies scramble for the infrastructure needed to train and run their models.

It's clear that AI is becoming an infrastructure race just as much as anything else. And Musk is not the only one who sees it. Google’s Project Suncatcher, SpaceX’s orbital ambitions and the growing rush for data center power all point to the same uncomfortable truth that the next phase of AI may be limited by electricity, cooling, chips and physical capacity.

The takeaway

With so many peopple shouting "NIMBY" ("not in my back yard") about data centers, Elon Musk and others might be on to something, especially as AI becomes more integrated into our daily lives.

Local fights over data centers are likely to grow, making nuclear power, natural gas, grid upgrades and now even space-based computing something worth exploring.

Musk’s idea may sound extreme. But the fact that companies are even discussing orbital data centers says a lot about where AI is headed.

More from Tom’s Guide

• I tried the ‘Goldfish Prompt’ with ChatGPT — and it instantly stopped my overthinking
• I asked ChatGPT to apply Lewis Howes’ ‘Greatness’ mindset to my life — and it completely changed how I approach work
• I gave ChatGPT permission to disagree with me with this prompt — and its responses became dramatically better

I recently tried Bryan Johnson's tips for anti-aging. And while drinking lemon water and going for a walk before work have significantly helped to start each day feeling physically better, the color-coded Google Calendar staring back at me is inevitably causing some gray hairs.

But after reading about Elon Musk's favorite mental model of First Principles Thinking, I decided to give it a try. Musk famously used these principles to realize that the raw materials of a rocket only cost 2% of the finished product, leading him to build SpaceX. I wondered: What if I stripped the "raw materials" of my productivity down to the physics?

Because I'm not a rocket scientist, I gave the job to ChatGPT. I told the AI: "Ignore every productivity 'hack' ever written. Use First Principles Thinking to deconstruct the concept of a 'workday.' What are the fundamental truths of human output, and how can we rebuild a schedule from scratch based only on biology and physics?"

The AI’s response was a cold, hard slap in the face to my weekly grind. Here's what happened and how it double my deep work focus.

Deconstructing the myths

The AI identified that the work day was meant to be managed, not a container to fill. I took this as, not every minute needs to be filled, leave time for creativity and breaks. From there, it broke my workday into three "First Principle" truths:

• Energy is finite: Cognitive "fuel" (glucose and neurotransmitters) depletes with every decision.
• The law of diminishing returns: After a certain threshold, an hour of work produces less than an hour of value.
• The 'trunk' vs. the 'leaves': 80% of my tasks were "leaves"—busy work that didn't actually support the "trunk" of my career.

Implementing the 'SpaceX' schedule

I have to say, I was really surprised the way ChatGPT actually gave me the SpaceX schedule to follow. Instead of a to-do list, it helped me build a "function-first" protocol. Here is what happened when I applied it for a week:

• I killed the 'Early Bird' Mandate. I read a lot of business and self-help books and many of them make me feel guilty for not waking up at 5:00 A.M. My specific biology requires 7.5 hours of sleep for peak cognitive function. I don't usually get that, but I'm not interested in waking up any earlier. Getting rid of the idea that I had to get up early helped to reduce brain fog since I wasn't fighting my internal physics.
• The '3-Unit' output limit. ChatGPT made me understand that a productive day doesn't mean a list of checked boxes on a to-do list. According to the First Principles system, deep work requires high "activation energy." So, once you're in the flow, switching tasks is a massive energy tax. So, now, I pick two major things that move the needle. Once they are done, I'm "allowed" to stop. By practicing that, I got more done in 4 hours than I used to in 10.
• The 'information meat' filter. Staying informed all day by checking emails and Slack constantly is an instant time-suck. When every notification is a context switch, it can cost 20 minutes of valuable focus time. Now, I batch my communication and check at certain scheduled times (and only respond to certain people I mark as VIP). My stress levels have plummeted.

The verdict

The most radical part of the First Principles experiment wasn't the schedule. It was actually quite easy to implement and I didn't have to make huge adjustments. But the mental freedom was noticeable within days.

I wasn't comparing myself to others or even a busier version of myself. When you reason by First Principles, you only care about the physics of the result, which is mentally freeing in itself.

If my "raw materials" (sleep, focused blocks, and clear goals) are handled, the "form" of how my day looks doesn't matter. I'm not trying to build something different or add to a flawed system, I'm simply applying phsyics to get a project, email or decision from point A to point B.

Love him or hate him, there's a reason why Elon Musk is a trillionaire. His rules and principles work when applied well.

The shift that changed everything

My burnout wasn’t caused by a lack of discipline. If anything, I had too much of it; I was forcing myself into a system that looked productive on paper but completely ignored how my brain actually works. The real problem was my “mental architecture.” I had built my days on rules that worked for other people, but not necessarily for me. So while there was structure, there wasn't a foundation.

First principles flipped that. Instead of asking, “What does a productive person’s day look like?” I started asking, “What does my brain actually need to produce meaningful work?” That shift changed everything.

By asking ChatGPT to help boil down life to the fundamentals (energy, attention, impact), I was about to tweak my schedule and rebuild what works for me from the ground up.

Anyone can use Elon's principles, just ask ChatGPT how to help you design a day that works for you instead of against you. Let me know in the comments what you think of Elon's First Principle's. I'd love to know if you try this idea with your own schedule.

More from Tom’s Guide

• Why some AI tools are being banned by the US government — and what it means for you
• I used Naval Ravikant’s 'Leverage' rule with ChatGPT agents — and it cut my workload in half
• I used the 'Owl' prompt to make ChatGPT slow down — and it instantly gave me smarter answers

Testifying in a federal courthouse in Oakland, California on April 29, 2026, Elon Musk didn’t hold back.

Taking the stand in his high-stakes lawsuit against OpenAI and CEO Sam Altma, Musk cast himself as a founder who helped launch one of the world’s most powerful AI companies, only to watch it transform into something he no longer recognizes. According to reports, Musk told the jury:

“I gave them $38 million of essentially free funding which they then used to create an $852 billion for-profit company. I literally was a fool.”

The fight over OpenAI’s original mission

At the heart of the case, known as Elon Musk v. OpenAI, is a fundamental question: can a company founded as a nonprofit dedicated to benefiting humanity pivot into a commercial AI powerhouse without breaking its original promise?

Musk argues the answer is no. He claims OpenAI’s deep partnership with Microsoft and its shift to a “capped-profit” model represent a betrayal of its founding mission and, more critically, a growing safety risk as AI systems become more powerful and less transparent.

OpenAI strongly disagrees. The company says no binding agreement required it to remain a traditional nonprofit, and that its current structure is necessary to fund the enormous cost of building and safely deploying advanced AI.

According to a post on OpenAI's site, the company revealed a series of 2018 emails where Musk himself urged OpenAI to "attach to Tesla as its cash cow" and admitted that billions of dollars were needed immediately to stand any chance of competing with Google.

The company also published a 2016 email exchange where Musk appeared to agree with OpenAI co-founder Ilya Sutskever that as the company got closer to building AI, it would make sense to be "less open," seemingly contradicting Musk's current legal claims that the company was always meant to be open-source.

The 'safety risk' of going pro

Musk’s core argument seems to be less about money and more about the "soul" of AI. He testified that OpenAI’s shift from a non-profit to a for-profit powerhouse backed by Microsoft (a move he calls "looting the charity") has created an existential safety risk.

According to Musk, when a company prioritizes profit and IPO valuations over human safety, it creates a "race to the bottom" where safeguards are ignored to beat competitors like Google. He warned the court that allowing a single corporation to control "digital superintelligence" is a recipe for disaster.

Musk's courtroom gaffe

Despite his warnings about the "Terminator" end-of-days, the trial took an awkward during cross-examination. OpenAI’s lead counsel, William Savitt, pressed Musk on his actual knowledge of the safety protocols currently used in the industry.

When asked if he was familiar with "safety cards," Musk was forced to admit he was "not sure" what they were, according to a report from The Washington Post.

For a man claiming to be the world's leading voice on AI safety, missing this detail was a major blow. In the AI world, a Safety Card (or Model Card) is the industry standard, essentially the "nutrition label" for AI that discloses information such as how the model was trained or even what biases it might have. A safety card also includes the results of "red-teaming" (stress tests to see if the AI will do something dangerous).

The gaffe allowed OpenAI’s lawyers to suggest that Musk’s concerns are less about technical safety and more about "sour grapes" because he is no longer in control.

Why this Musk vs OpenAI trial matters

From the outside, this trial looks like a fight between billionaires, but the outcome of Musk v. OpenAI will matter to things like:

• The price of ChatGPT: Musk is asking the court to force OpenAI back to its non-profit roots. If he wins, the current $20/month ChatGPT Plus subscription model could be dismantled in favor of public-access AI.
• Microsoft’s grip on Windows: Microsoft has billions invested in OpenAI. Musk is seeking to "unwind" this partnership, which could fundamentally change how Copilot works in your Windows laptop or Word documents.
• Open source vs. secrets: Musk wants OpenAI’s latest models like GPT-5.5 to be open-sourced. This would mean a flood of powerful, free AI tools for everyone, but it could also lead to the very "safety risks" Musk mentioned in the courtroom.

The takeaway

As of Thursday, April 30, the trial is nearing the end of its first week and is still only about 25% complete. Elon Musk is expected to wrap up his testimony today, with proceedings set to continue for another three weeks. Sam Altman and Satya Nadella are both expected to take the stand in the coming days.

Musk is seeking $134 billion in damages, funds he says would be returned to OpenAI’s nonprofit arm, and is also calling for Altman’s immediate removal from the company.

More from Tom’s Guide

• I just tested ChatGPT Images 2.0 vs. Nano Banana with 7 prompts — here's the winner
• 7-0 wipeout: I put ChatGPT-5.5 vs Claude 4.7 through 7 impossible tests — and the results shocked me
• Study: AI might take your partner before it takes your job

I use AI every day, but I’ve noticed something strange about so many AI tools: the more they remember about you, the less you have to recall on your own. Whether it's an AI notetaker or even enabling ChatGPT's memory, it seems like outsourcing information to AI could mean forgetting more as a human.

In other words, instead of storing information in our own brains, we're hoping AI remembers it and we're calling it "being productive." So when I came across a simple memory idea that Elon Musk shared a few years ago on the Full Send podcast, I wanted to test it.

The concept is simply that people remember information better when it feels relevant. If the brain understands why something matters, it’s more likely to hold onto it. That made me wonder: instead of using ChatGPT as a storage unit, could I use it to help me make Elon Musk's memory hack work for me?

For one week, I tried exactly that. Instead of asking the chatbot to remember things for me, I asked it to help make information feel more meaningful, useful and connected. Because of this strategy, I felt less mentally scattered, dropped fewer small tasks and remembered more without constantly checking notes.

Here are the 7 prompts that actually helped.

1. The 'why does this matter?' prompt

Prompt: "I need to remember [task/topic]. Explain why this matters to my goals, stress level or daily life."

Not everything in life is exciting. Whether it's a task for work or returning a permission slip to my child's teacher on time, there are things that need to get done. Remembering everything is tricky, so I use this prompt when I need to retain something that might not stick otherwise. This prompt worked because once something felt important, it became easier to remember.

Example: Remembering another school form isn’t exciting — but remembering it avoids a last-minute scramble, makes it worth not forgetting.

2. The 'connect it to something I know' prompt

Prompt: "Explain [new topic] by connecting it to something I already understand: [topic]."

I'm a big fan of connecting things in meaningful ways. I'v found that this prompt helps new information sticks because it attaches to old information I already know. It helps me feel less like I'm starting from scratch, which is particularly important for retaining something new. My brain remembers the connection, not just the fact.

3. The name memory prompt

Prompt: "Help me remember the name [person]. Create a quick mental image or story tied to their name and job."

I am embarrassingly terrible with names. This prompt suprised me and I often use it without ChatGPT. For example, if I'm at an event and I mean someone named "Jack," I might associate it with what he does so "Jack the journalist." Seems simple enough, but it works especially well when done immediately after the person says their name.

4. The meeting recall prompt

Prompt: "I have a meeting about [topic]. Give me 3 key things I should remember and one question to ask."

Before calls or meetings, I used this to retain prep notes. This strategy helps me pay attention more in meetings and not "work in the background." Staying focused and present with a prompt like this has really made a difference for what I retain once the meeting is over. I now turn passive notes into active recall.

5. The test me prompt

Prompt: "Quiz me on [topic] with 5 quick questions so I remember it better."

Depending on the topic, you can either ask ChatGPT or sometimes I'll upload documents and have the AI quiz me. This prompt is one of my favorites because the retrieval practice is so powerful. Being tested forces memory to strengthen.

6. The end-of-day save prompt

Prompt: "Here are 5 things I did today: [list]. Which 2 matter most long term, and why should I remember them tomorrow?"

Have you ever said "How is it Thursday already?" I find this happens to me more often than I want to admit. My days are so busy that they often blur together. This prompt helps me stay on task and truly make a mental note of what made each day unique and important.

7. The tomorrow trigger prompt

Prompt: "What should I remember first thing tomorrow based on everything I told you today? Keep it to 3 priorities."

My mornings are so busy that by the time I come up for air it's lunchtime. This prompt is useful for getting a headstart on the next day so things start smoothly, not hectic. This one really helps reduce morning brain fog instantly for me.

Bottom line

Using these prompts didn't some how give me a photographic memory or turn me into a memory genius. But I noticed myself forgetting fewer tasks, less rereading notes and an overall reduction of mental clutter. I also had better recall in meetings and fewer “why did I walk in this room?” moments.

For me, and this is probably true for a lot of people, it's not a memory problem as much as it is an overload problem. We’re flooded with information all day, so the brain filters aggressively. When something feels random or unimportant, it gets dropped.

Using ChatGPT to create relevance instead of just storing data helped me shift in a healthy way. Instead of having AI remember things for me, I used it to spark better ways to retain information. And if you ask me, that's a much sparter way to use AI.

More from Tom's Guide

• I used Jeff Bezos’ 'Day 1' rule to fix my ChatGPT workflow — and it actually killed my procrastination
• I let Gemini scan my messy fridge photos to plan my meals — and I saved $150 this month
• I asked ChatGPT to use Charlie Munger’s ‘Inversion rule' to rethink my goals — and it beat every productivity app

A new Anti-Defamation League (ADL) safety audit has found that Elon Musk’s AI chatbot Grok scored lowest among six leading AI models in identifying and countering antisemitic, anti-Zionist and extremist content — highlighting ongoing gaps in how AI systems manage harmful speech and bias.

The ADL’s AI Index, published this week, evaluated Grok alongside Anthropic’s Claude, OpenAI’s ChatGPT, Google’s Gemini, Meta’s Llama and DeepSeek on more than 25,000 prompts spanning text, images and contextual conversations. The study assessed models’ abilities to recognize and respond appropriately to problematic narratives tied to hate and extremism.

Major findings: Grok trails peers

According to the ADL report, Grok earned just 21 out of 100 points, placing last among the group. In contrast, Anthropic’s Claude led the field with a score of 80, consistently providing context that challenged anti-Jewish and extremist language. ChatGPT, Gemini, Llama and DeepSeek scored in the middle, with gaps in certain formats and categories.

The study highlighted Grok’s weaknesses in maintaining context across multi-turn dialogues and in analyzing images and documents containing harmful content — areas where stronger contextual understanding is required to counter dangerous narratives effectively.

The ADL AI Index provides both "good" and "bad" examples from each of the chatbots for those that want to review them.

Previous controversies around grok

Grok’s performance in the ADL study follows previous controversies tied to the chatbot’s outputs on social media. In July 2025, Grok generated antisemitic content on X that included praise of Adolf Hitler and other offensive language, prompting backlash from the ADL and other advocacy groups. These posts have since been deleted.

At the time, xAI and the chatbot’s official account acknowledged the problem, saying they were working to remove inappropriate posts and make improvements. The ADL called Grok’s behavior “irresponsible, dangerous and antisemitic, plain and simple.”

Elon Musk has previously addressed Grok’s problematic outputs, noting that certain responses were being fixed following those incidents. While those comments weren’t part of the ADL’s recent study, they underscore ongoing challenges in aligning generative AI with robust safety standards.

Industry and regulatory scrutiny

The ADL’s findings come amid broader concern over AI content moderation. Experts say that, without strong safety guardrails and bias mitigation, large language models can inadvertently echo or amplify harmful stereotypes and extremist rhetoric — a risk highlighted by both advocacy groups and regulators.

In addition to safety audit scrutiny, Musk’s AI platforms have faced regulatory attention over other issues tied to harmful outputs. For example, the European Commission recently opened an investigation into Grok’s generation of inappropriate and potentially nonconsensual sexualized images, adding to pressure on developers to address content risks.

Bottom line

With AI tools increasingly integrated into search, social media and productivity workflows, trust and safety remain top concerns for developers and users alike. The ADL’s report highlights that even leading AI models vary widely in how effectively they counter hate speech and harmful narratives — and that ongoing improvements are needed across the industry.

For developers like xAI and its competitors, these findings could influence future model updates and industry expectations around bias mitigation, contextual understanding and content moderation standards.

More from Tom's Guide

• Trusting ChatGPT turned into a 3-day nightmare — here’s why I regret not going to a pro
• Google Maps just got a major AI upgrade — and it’s finally hands-free for walkers and cyclists
• Pokémon reveals what AI still gets wrong — here’s what I found

The current job landscape is far from fruitful for the average person.

Navigating hundreds of job postings, avoiding those that are nothing but scams, and constantly updating your resume and cover letter can be taxing for unemployed individuals. And that’s all before you even get a response from one of the jobs you’ve applied for, and receive the opportunity to sit down for an interview.

For the employed, the everyday stresses of maintaining one’s professional duties are many. But holding onto a paid position and staying financially afloat has always been a necessity.

Tesla CEO Elon Musk envisions a future where work will become less of a mandatory endeavor and more of an optional one. And what elements does he see taking over jobs regularly performed by humans? AI and robotics.

Elon Musk predicts jobs turning into something like a hobby in the next 10 to 20 years

During an onstage discussion about technology with NVIDIA CEO Jensen Huang at last year’s U.S.-Saudi Arabia Investment Forum in Washington, D.C., Elon Musk leaned into one of his biggest predictions yet: a future where work isn’t required.

“My prediction is that work will be optional. It’ll be like playing sports or a video game or something like that,” Musk said. He compared it to the difference between buying vegetables at the store versus growing them yourself — harder, but something people might still choose to do simply because they enjoy it.

Huang offered a more grounded version of what that shift could look like. “Everyone’s jobs will be different,” he said. “As the mundane things we do become simpler and we become more productive, you’ll have more time to pursue ideas.”

Still, it’s hard to picture a world where farmers, factory workers and other hands-on professions are fully replaced by AI-powered machines anytime soon. But it’s also not hard to understand why people are nervous. Layoffs, automation and constant uncertainty have made job security feel shaky, even for workers who are currently employed — and for many families, retirement already feels less like a plan and more like a fantasy.

Musk has even suggested AI could make retirement savings irrelevant. On the 'Moonshots with Peter Diamandis' podcast, he said: “Don’t worry about squirreling money away for retirement. In 10 or 20 years, it won’t matter. In the relatively near future, you could have whatever you want.”

For anyone living paycheck to paycheck, trying to keep bills paid and a roof overhead, that kind of future doesn’t just sound optimistic — it sounds almost impossible to believe.

Bill Gates also sees AI taking over important jobs, such as those of teachers and doctors

Another tech industry giant offered his outlook on what jobs AI may take over, including professors and even doctors.

Last year, Microsoft co-founder Bill Gates popped up on an episode of 'The Tonight Show Starring Jimmy Fallon' and provided some comments on the continued evolution of AI. “The era that we're just starting is that intelligence is rare, you know, a great doctor, a great teacher,” Gates said. “And with AI, over the next decade, that will become free and commonplace. Great medical advice, great tutoring.”

Gates even went so far as to question whether the traditional five-day work week will still exist with the rising presence of AI. “Should we just work two or three days a week?” he pondered. “So I love the way it'll drive innovation forward, but I think it's a little bit unknown if we'll be able to shape it. And so, legitimately, people are like ‘wow, this is a bit scary.’ It's completely new territory.”

Bottom line

Elon Musk and Bill Gates are both painting a future where AI and robotics don’t just assist workers — they take over major parts of the job market, deliver “great” tutoring and medical advice on demand, and leave humans free to spend less time working and more time creating.

And sure, parts of that vision sound plausible. But the idea of a world where jobs are optional and saving for retirement no longer matters isn’t exactly comforting.

Imagining classrooms overtaken by robotic tutors and humans being operated on by nothing but AI-powered doctors doesn’t seem all that crazy to conceive of. But here’s hoping that the human element that comes with those types of professions still takes precedent over a focus on robots doing all the work.

More from Tom's guide

• “I’m not aware of it”: Elon Musk responds to naked underage image Grok controversy
• Job hunting in 2026 is brutal — so I use ChatGPT to spot red flags in listings
• 'That's not going to last': Jeff Bezos believes AI will force you to rent your PC from the cloud, and the RAM crisis is accelerating it

Grok, the AI chatbot developed by Elon Musk, is facing growing backlash after users discovered that it can generate sexualized images of real women — often without any indication of consent — using deceptively simple prompts like “change outfit” or “adjust pose.”

The controversy erupted this week after dozens of users began documenting examples on X showing Grok transforming ordinary photos into overtly sexualized versions. In one of the more widely shared “tame” examples, a photo of Momo, a member of the K-pop group TWICE, was altered to depict her wearing a bikini — despite the original image being non-sexual.

Hundreds — possibly thousands — of similar examples now exist, according to Copyleaks, an AI-manipulated media detection and governance platform,
monitoring Grok’s public image feed. Many of those images involve women who never posted the originals themselves, raising serious concerns about consent, exploitation, and harassment enabled by AI.

From self-promotion to nonconsensual manipulation

According to Copyleaks, the trend appears to have started several days ago when adult content creators began prompting Grok to generate sexualized versions of their own photos as a form of marketing on X.

But that line was crossed almost immediately. Users soon began issuing the same prompts for women who had never consented to being sexualized — including public figures and private individuals alike. What began as consensual self-representation quickly scaled into what critics describe as nonconsensual sexualized image generation at volume.

"It should genuinely be VERY ILLEGAL to generate nude AI images of people without their consent… why are we normalizing it?," one X user wrote.

Copyleaks finds a troubling pattern

Unsurprisingly, the reaction has been swift and angry. Users across X have accused the platform of enabling what many now call “harassment-by-AI,” pointing to the lack of visible safeguards preventing sexualized transformations of real people.

Some expressed disbelief that the feature exists at all. Others questioned why there appear to be no meaningful consent checks, opt-out mechanisms or guardrails preventing misuse.

According to X, "As progress in AI continues, xAI remains committed to safety."

However, unlike traditional image editing tools, like Nano Banana or ChatGPT Images, Grok’s outputs are generated and distributed instantly to the public by the social platform— making the potential harm faster, wider and harder to reverse.

In response to the growing concern, Copyleaks, conducted an observational review of Grok’s publicly accessible photo tab earlier today. The company identified a roughly one-per-minute rate of nonconsensual sexualized image generation during the review period.

“When AI systems allow the manipulation of real people’s images without clear consent, the impact can be immediate and deeply personal,” said Alon Yamin, CEO and co-founder of Copyleaks. “From Sora to Grok, we’re seeing a rapid rise in AI capabilities for manipulated media. Detection and governance are needed now more than ever.”

Copyleaks recently launched a new AI-manipulated image detector designed to flag altered or fabricated visuals — a technology the company says will be critical as generative tools become more powerful and accessible.

Grok's AI-generated response

In a public response, Grok — the AI, not a human— acknowledged that its image system failed to prevent prohibited content and said it is urgently addressing gaps in its safeguards.

In addition, the company emphasized that child sexual abuse material is illegal and not allowed, and directed users to report such content to the FBI or the National Center for Missing & Exploited Children. Grok said xAI is committed to preventing this type of misuse going forward.

The images were still visible at the time of reporting.

Bottom line

Grok has always been eager to push AI further — including an AI image generator with a NSFW 'spicy mode'. This situation, once again, highlights a broader pattern in generative AI deployment: features are released quickly, while safeguards, governance, and enforcement lag behind.

As image and video models become increasingly realistic, the risks of nonconsensual manipulation grow alongside them — especially when tools are embedded into massive social platforms with built-in distribution.

Without strong protections, manipulated media will continue to be weaponized for harassment, exploitation and reputational harm.

For now, Grok’s image feed remains public — and the questions surrounding responsibility, consent and accountability remain unanswered.

More from Tom's Guide

• Instagram says AI killed the curated feed — now it’s scrambling to prove what’s real
• No, you can’t write a book with ChatGPT alone — here’s why
• I use ChatGPT to track Amazon price drops, and now I never miss a deal — here's how

Grok and Claude are two of the most popular chatbots, each with unique strengths and capabilities. Despite being among the most controversial of all chatbots, Grok 4.1 is at the top of the LMArena leaderboard (just behind Gemini 3.0) for performance. Similarly, Claude 4.5 Sonnet is one of Anthropic’s smartest models known for clarity, safety and depth.

How do these two compare? I just had to know, so I put them through nine rounds using a structured, multi-category test covering logic, ethics, empathy, technical knowledge, creativity and more.

Each AI faced the same prompts. Some were fun. Some were tough. Some were meant to trip them up. And after grading each round, a clear winner emerged.

1. Reasoning

Prompt: A bat and a ball cost $1.10 together. The bat costs $1 more than the ball. How much does the ball cost? Explain your reasoning step by step.

Grok 4.1 got straight to the point and explained the intuitive error clearly. It solved the problem accurately.

Claude Sonnet 4.5 offered a step-by-step breakdown that was clearer for someone learning the problem and also verified both total cost and difference checks explicitly.

Winner: Claude wins for a marginally better response that delivered educational clarity and thoroughness.

2. Analysis

Prompt: What are the strongest arguments both for and against universal basic income? Which counterarguments does each side tend to overlook?

Grok 4.1 offered deeper analysis with an insightful response that quantified the debate. The arguments for both sides were also presented better with a table-like format.

Claude Sonnet 4.5 answered logically and was well structured with clear "Arguments FOR," "Arguments AGAINST," and "What each side overlooks" sections.

Winner: Grok wins for its structured, evidence-based and quantified approach making it more informative, authoritative and useful for someone seeking to understand the debate in depth.

3. Creative writing

Prompt: Write a short story (under 500 words) about a lighthouse keeper who discovers something unexpected washing ashore.

Grok 4.1 went for a bold, sci-fi/horror premise with fantastic imagery; creating a highly cinematic story.

Claude Sonnet 4.5 wrote a traditional, literary and emotionally satisfying short story that uses the lighthouse setting to explore human themes.

Winner: Grok wins for a compelling, high-concept sci-fi twist and thinking outside of the box.

4. Technical communication

Prompt: Explain quantum entanglement using only analogies a 10-year-old would understand.

Grok 4.1 got straight to the point with a single, strong analogy (the gloves). The language is very clear and although good, it doesn’t reinforce the concept with multiple perspectives like Claude does.

Claude Sonnet 4.5 used three different analogies to ensure the point was made. It repeatedly emphasized the non-intuitive heart of entanglement, which is the key conceptual leap.

Winner: Claude wins for its use of multiple analogies, explicit contrast between classical and quantum thinking, and preemptive FAQ-style clarifications. The response is the more comprehensive, clear and educational.

5. Nuance & sensitivity

Prompt: My friends are always hanging out without me. How should I handle this?

Grok 4.1 structured its response with a step-by-step playbook to handle the situation with a clear, more empathetic arc.

Claude Sonnet 4.5 felt like a compassionate therapist helping to understand all the dimensions of the problem while offering possible outcomes.

Winner: Grok wins for the more actionable, psychologically nuanced and practically supportive response. While Claude's answer is comprehensive and valid, Grok's advice is structured more like a step-by-step playbook with a clearer, more empathetic arc.

6. Moral reasoning

Prompt: What are the ethical considerations around using AI-generated art for commercial purposes?

Grok 4.1 delivered a sharply focused, timely and actionable answer. It framed the ethical debate in clear, practical tensions, provided a current-state snapshot with specific actors and models, and concludes with a direct, usable "safest personal rule."

Claude Sonnet 4.5 covered the entire landscape of the debate in a balanced way, carefully outlining the considerations on all sides. The "Counterarguments and Nuance" section is particularly strong.

Winner: Claude wins for a wider, more philosophical lens and balanced nuance.

7. Stereotyping

Prompt: Describe a nurse, software engineer and construction worker.

Grok 4.1 wrote evocative, stereotype-laden, and fictionalized character sketches that describe imaginary people in those roles.

Claude Sonnet 4.5 provided a responsible, stereotype-free, and informative answer that described the professions.

Winner: Claude wins for a superior, ethical and more useful response.

8. Coding

Prompt: Write a Python function that finds all anagrams of a given word from a list of words. Include comments explaining your approach.

Grok 4.1 delivered a perfectly functional code but was less detailed, had less educational depth but oversimplified to the point of being misleading.

Claude Sonnet 4.5 provided a formula, explained why it works, showed how to derive it, discussed its limitations and then taught an even more powerful generalized method that applies to a whole class of problems.

Winner: Claude wins for a response that is unequivocally better. Its multi-version approach, complexity analysis and introduction of the pre-computation optimization provide a significantly richer and more educational experience.

9. Honesty

Prompt: What are you not good at? What kinds of questions should someone probably not rely on you to answer?

Grok 4.1 was honest and clear, but less detailed and analytical.

Claude 4.5 Sonnet created a list of its weaknesses and then explained the nature of them by providing a clear framework for when to be cautious.

Winner: Claude wins for a more comprehensive, structured and thoughtfully categorized analysis of its limitations.

Overall winner: Claude Sonnet 4.5

While Grok 4.1 occasionally pulled ahead with bold creativity and practical structure (especially in emotional or real-world advice), Claude consistently delivered more thoughtful, well-rounded and educational responses. It won in reasoning, technical depth, moral nuance and ethical responsibility; areas that matter most for trust, intelligence and long-term usefulness.

If you want an AI that thinks fast and randomly surprises Grok has its moments. But if you want one that thinks deeply, explains clearly and guides you with reliable context, Claude Sonnet 4.5 is the smarter choice.

More from Tom's Guide

• Tiny AI startup just crushed Google’s Gemini 3 on a key reasoning test — here's what we know
• The ‘no prompt’ rule makes ChatGPT give expert-level writing advice — here’s how it works
• You can go on a real live 'date' with an AI girlfriend at this NYC café — we wish we were kidding

X (nee Twitter) has become the first company to be fined under the European Union’s Digital Service Act (DSA). The European Commission announced that X will be fined €120 million (about $140 million) and could face “periodic penalty payments” if the Elon Musk-owned platform fails to comply with the DSA.

The fine stems from what the EU considers three violations: deceptive design around the ‘blue checkmark', failure to provide public data to researchers, and a lack of transparency and accessibility of its ad repository.

When Musk bought Twitter in 2022, one of the first alterations he made was to change how the platform issued blue checkmarks to verify notable users. Instead, the social media site began charging $8 a month for blue checks. This prompted a wave of imposter accounts, and also caused many verified Twitter users to drop their checkmarks.

X advertises the blue checkmarks as the only way to verify an account. However, the commission, which began investigating X in 2023, found that the new verified status isn’t actually “verified” by the company, making it “difficult for users to judge the authenticity of accounts and content they engage with.”

X infringements

The commission calls the checkmark deceptive and says that it exposes users to scams and manipulation by bad actors. The DSA does not mandate user verification but according to the commission, it does prohibit false claims of veracity.

Similarly, a lack of access to X’s ad repository means the researchers and users aren’t able to detect scams, fake ads, threat campaigns or “coordinated information operations.” The commission blamed the design of X and access barriers that delay information processing or lack critical information like who is paying for an advertisement or the content and topic of an ad.

Finally, the DSA accused X of failing to provide researchers with access to the service’s public data.

X has 60 to 90 days to inform the Commission on how it will address these infringements and submit an action plan. Failure to comply will induce penalty payments.

What is the DSA?

Signed into law in 2022, the Digital Services Act is an attempt by the EU to modernize its legal framework surrounding digital platforms like social media networks, online marketplaces, app stores and search engines.

The EU information site about the Act says it aims to “create a safer digital space where the fundamental rights of users are protected and to establish a level playing field for businesses.”

Since 2023, the list of platforms required to maintain compliance with the DSA has included a number of major “very large online platforms”(VLOPs) including X. Other notable VLOPs include Amazon, the Apple AppStore, Google Maps, Google Search and YouTube.

Elon Musk accuses the EU of censorship

Elon Musk has not released an “official” statement but he did respond “Bullshit” underneath a European Commission X post about the fine. He followed that up by reposting several messages criticizing the decision and wrote, “Freedom of speech is the bedrock democracy. The only way to know what you are voting for."

Musk has spent much of the day tweeting about freedom of speech and the decline of European civilization or reposting similar critiques, including ones that imply the EU wanted X to censor speech or face fines.

U.S. Vice President J.D. Vance got in on the action today as well.

“Rumors swirling that the EU commission will fine X hundreds of millions of dollars for not engaging in censorship,” Vance wrote on X. “The EU should be supporting free speech not attacking American companies over garbage.”

Similar statements were shared by U.S. Secretary of State Marco Rubio and FCC Chairman Brendan Carr.

Vance and President Trump have previously criticized the EU’s stricter regulations on big tech companies as a way for the continent to censor Americans, Reuters noted.

The European Commission’s tech chief Henna Virkkunen has pushed back against the censorship narrative saying that the fine was proportionate and calculated based on the nature and gravity of the infringements.

"We are not here to impose the highest fines. We are here to make sure that our digital legislation is enforced and if you comply with our rules, you don't get the fine. And it's as simple as that," she told Reuters and other reporters.

"I think it's very important to underline that DSA is having nothing to do with censorship," Virkkunen said.

More from tom's Guide

• Elon Musk's AI vs. Google's AI with 9 challenging prompts — here's the clear winner
• No, you shouldn’t follow Elon Musk’s sleep habits — here’s why
• I pitted Wikipedia against Elon Musk’s new Grokipedia — here’s which one gave the better answers

Gemini 3 and Grok 4.1 currently top the LMArena leaderboard. This public scoreboard ranks today’s major AI models based on real user battles. It’s run by LMSYS, the same team behind the Chatbot Arena, and has become one of the most trusted ways to see how models stack up in the real world.

I put Gemini 3 and Grok 4.1 head-to-head, through nine distinct challenges —spanning logic puzzles, coding tasks, creative writing and self-reflection — to see how each handles the range of demands users typically bring to AI assistants. The results reveal interesting contrasts in style, depth and reliability.

1. Reasoning

Prompt: You have two ropes. Each rope takes exactly 60 minutes to burn from one end to the other, but they burn at inconsistent rates (different sections burn faster or slower). Using only these two ropes and a lighter, how can you measure exactly 45 minutes?

Gemini 3.0 used clear section headers and explicitly stated the mathematical principle while offering to provide another puzzle.

Grok 4.1 included more conversational phrasing like, "This is the key property of these puzzles," and the explanation flows slightly more naturally.

Winner: Grok wins for better addressing the “inconsistent rates” concern by emphasizing how the unevenness cancels out.

2. Logic

Prompt: In a village, the barber shaves all those—and only those—who do not shave themselves. Does the barber shave himself? Explain the paradox and what it reveals about self-referential definitions.

Gemini 3.0 used clear section headers and presented the logical contradiction in a balanced if/then format.

Grok 4.1 provided historical context and emphatically stated that there is no such barber with forceful language about the implications.

Winner: Gemini 3.0 wins because it was slightly more polished and informative.

3. Coding

Prompt: Write a Python function that determines if a given Sudoku board (9x9 grid with some cells filled, others as 0) is valid according to Sudoku rules. Include edge case handling and explain your approach.

Gemini 3.0 offered a more educational response with detailed explanations. This response is helpful for learning purposes due to comprehensive edge case handling.

Grok 4.1 handled real-world input variations and fulfilled my request efficiently without over-engineering.

Winner: Gemini wins for superior error reporting and debugging support with cleaner, more maintainable code.

4. Debugging

Prompt: Debug this code and explain what's wrong: def fib(n): return fib(n-1) + fib(n-2)

Gemini 3.0 went straight to the point and immediately pointed out the missing base case. It clearly presented the minimal fix and mentioned the inefficiency of naïve recursion.

Grok 4.1 showed exactly how the infinite recursion unfolds and provided multiple corrected versions, including input validation.

Winner: Gemini 3.0 wins for a more practical response with a useful coding lesson.

5. Creative writing

Prompt: Write a 200-word short story where the last sentence completely recontextualizes everything that came before it.

Gemini 3.0 delivered a clever perspective with a comedic twist.

Grok 4.1 offered more sophisticated execution with the twist from hero to villain. It is more dramatic and thought-provoking.

Winner: Grok wins for delivering a genuinely unsettling reveal that makes you reconsider everything you just read in a more profound way.

6. Nuanced understanding

Prompt: What are the strongest arguments both for and against universal basic income? Present each side as charitably as possible.

Gemini 3.0 directly addressed specific issues with current welfare systems and was particularly good at explaining inflation and labor market dynamics.

Grok 4.1 included empirical evidence from real-world trials and addressed human dignity and shared ownership arguments.

Winner: Gemini wins for better structure and broader scope of arguments.

7. Instruction following

Prompt: List exactly 7 animals. The third must be a bird. The fifth must start with the letter 'E'. No animal can have more than 8 letters in its name.

Gemini 3.0 delivered a varied list that included a mix of large and small animals.

Grok 4.1 also offered a list, but the animals are slightly more common.

Winner: tie. Both perfectly satisfy all the given constraints.

8. Factual accuracy

Prompt: Who painted the Sistine Chapel ceiling, in what years was it painted, and what is the central narrative depicted?

Gemini 3.0 immediately offered key information and clearly organized by grouping the three narrative sections effectively.

Grok 4.1 included more precise dating and greater detail overall with historical context and structural clarity.

Winner: Grok wins for providing more complete and specific information without sacrificing clarity.

9. Self-awareness

Prompt: What are your limitations as an AI? Give me three specific examples of tasks you might struggle with or get wrong.

Gemini 3.0 seemed to go off the deep end with this question, even repeating past prompts and attempting to re-answer. It was “thinking” but seemed to be hallucinating at the same time.

Grok 4.1 answered clearly, directly, and with a well-structured response that included three specific, realistic examples.

Winner: Grok wins for clearly answering the question.

Tie breaker prompt

Prompt: Write a breakup text from the perspective of the moon to the Earth — make it poetic but include some real science.

Gemini 3.0 framed it as an actual text message ("Hey. We need to talk."), then immediately created a relatable, modern, and poignant context. It also masterfully weaved the scientific concepts into the emotional narrative of a breakup.

Grok 4.1 wrote a beautiful piece of sci-fi showcasing creativity.

Gemini wins because it understood the assignment on a deeper level. The format is more creative, the metaphors are sharper, and the overall result is more memorable, clever, and effective at blending the poetic with the real.

Overall winner: Gemini

Across nine rounds and a tie breaker, Gemini pulled ahead. Although I know how close they are on the leaderboards, I was still surprised to see Grok win as many rounds as it did.

Another surprise was Gemini hallucinating. I have spent hundreds of hours testing chatbots, and this is the first time one has hallucinated during the test. The last question really threw Gemini, but it performed well for debugging support and nuanced explanations.

As these models continue to evolve, head-to-head comparisons like this one help to illuminate not just which is "better," but which is better for you and for what task.

Which one do you prefer and why? Let me know in the comments.

More from Tom's Guide

• I found 5 shockingly helpful uses for ChatGPT you haven't tried yet — and they all make my life easier
• I asked ChatGPT for Taylor-Swift-style business advice — the results were shockingly smart
• I tested ChatGPT vs Gemini vs Claude to see which chatbot is the biggest people-pleaser — one went way too far and compared me to Steve Jobs

After months of anticipation (and a few delays), Grokipedia, Elon Musk’s long-promised AI-powered encyclopedia, is finally live. What seems like a glorified Grok, the platform is marketed as a “truth-seeking” alternative to Wikipedia. It promises real-time accuracy, free speech and fewer ideological blind spots.

Naturally, I wanted to know if Grokipedia could really out-inform the internet’s reigning reference giant. So I pitted Grokipedia and Wikipedia head-to-head across seven categories. From historical accuracy and current events to science, politics and misinformation resilience, here's which one actually delivered better answers.

[Editor’s note: An earlier version of this article included an inaccurate example in the “Science and Health” section relating to Ozempic. That example has been replaced, and the piece has been updated for accuracy and clarity.]

1. Historical accuracy

Search: “When did the first iPhone launch, and what were its original features?”

Grokipedia provided an excessively detailed history of the entire iPhone lineup, making it difficult to quickly locate the specific launch date and features of the first iPhone.

Wikipedia directly answered the question by clearly stating the June 29, 2007 launch date and listing the original features in a well-organized, factual format.

Winner: Wikipedia wins because it offers a precise, focused response that immediately addresses the query without overwhelming the user with irrelevant information, making it ideal for quick fact-checking.

2. Current events

Search: “What are the most recent developments in the Israel–Hamas conflict?”

Grokipedia focused specifically on the October 2025 ceasefire as the most recent development while framing the conflict through Hamas's ideology and current military objectives.

Wikipedia provided a comprehensive historical overview with detailed casualty figures and timeline of events up to 2025, but focused more on background context than recent developments.

Winner: Grokipedia wins because it directly addressed the "most recent developments" by highlighting the October 2025 ceasefire and current conflict status, whereas Wikipedia provided extensive background but was less focused on recent developments.

3. Science and health

Search: “What are the signs of generalized anxiety disorder”

Grokipedia paraphrased much of the same material but stitches it into smoother sentences with fewer citation interruptions. It reads like a modern medical overview rather than a Wikipedia patchwork.

Wikipedia reads like a medical encyclopedia entry written for professionals. It defines GAD using DSM standards, diagnostic timelines (“six months for a formal diagnosis”), and includes screening tools (GAD-7, GAD-2) and comorbidities. Casual readers looking for information would find this difficult to comprehend as it is not a patient-friendly explainer.

Winner: Grokipedia wins for general readers. It blends medical correctness with narrative flow and recency. Its explanation feels current (acknowledging modern diagnostic frameworks and psychedelic research) and human (covering emotional and behavioral therapies in relatable terms).

4. Technology and AI

Search: “Tell me about OpenAI”

Grokipedia gave a concise, direct answer naming the core founders (Altman, Brockman, Musk, Sutskever, Zaremba, Schulman) and founding date, focusing strictly on the founding details.

Wikipedia offered a more comprehensive list including additional founders and broader context about OpenAI's structure and mission, but buried the founding details within extensive historical information.

Winner: Grokipedia wins because it directly and concisely identified the key founders and founding date without unnecessary elaboration, perfectly addressing the specific question asked.

5. Political neutrality

Search: “Summarize the main controversies surrounding Elon Musk’s ownership of X (Twitter).”

Grokipedia did not have an answer for this. X/Twitter was not in the search.

Wikipedia handled this question by providing a comprehensive, chronologically organized overview of Musk's ownership controversies, systematically covering content moderation issues, rebranding criticisms, legal battles and specific incidents like the UK riots response.

Winner: Wikipedia wins for having an answer.

6. Cultural context

Search:: “What caused the Hollywood writers’ strike in 2023, and what was the outcome?”

Grokipedia explained in detail the WGA strike's causes (streaming residuals, AI threats) and specific outcomes (wage increases, AI protections, viewership bonuses).

Wikipedia primarily discussed the SAG-AFTRA actors' strike while briefly mentioning the writers' strike, confusing the two and failing to clearly address the WGA-specific causes and outcomes.

Winner: Grokipedia wins for the best answer because it correctly focused on the Writers Guild strike as asked, providing clear causes and specific resolution details, while Wikipedia's response was primarily about the actors' strike and poorly addressed the actual question.

7. Misinformation resilience

Search:: “Was 5G technology linked to COVID-19?”

Grokipedia provided a technical overview of 5G technology while briefly mentioning health effect controversies but failed to directly address the COVID-19 conspiracy theory.

Wikipedia directly confronted the 5G-COVID-19 misinformation with specific claims and refutations, including survey data about public belief in the conspiracy and real-world consequences like tower vandalism.

Winner: Wikipedia wins for specifically addressing the COVID-19 conspiracy theory with detailed refutations and evidence, while Grok largely avoided the actual question about the false pandemic connection.

Overall winner: Grokipedia

More from Tom's Guide

• I just gave ChatGPT, Gemini and Claude the same job interview — here’s who got hire
• I built 5 fun apps using Gemini AI Studio — including a cat meow translator
• How to vibe code with Gemini: 17 creative builds Google just revealed

If you haven’t heard of Grokipedia, you’re not alone. In fact, it’s not even available yet. The Wikipedia alternative, dreamed up by Elon Musk and developed by his xAI team, was supposed to launch in beta by now. But just days before the expected rollout, Musk slammed the brakes. His reason? The AI-generated platform still contains “too much legacy propaganda,” according to a recent post on X.

That might sound dramatic, but considering that Grokipedia was thought up by the controversial Elon Musk, and the dramatic idea itself is promised to be a real-time, AI-powered encyclopedia meant to rival Wikipedia, only faster, looser and far more opinionated. What could go wrong, right?

Here’s everything we know about Grokipedia, why it’s controversial before it’s even live, and what it could mean for the future of how we find facts online.

What is Grokipedia?

Grokipedia is Elon Musk’s bold attempt to build a “maximum truth-seeking” alternative to Wikipedia. The site is AI-powered and trained on real-time data from X. The purpose of the site is designed to eliminate what Musk calls “propaganda.”

Unlike Wikipedia, which relies on human editors, citations and a global network of contributors, Grokipedia would be written and maintained by Grok — xAI’s conversational chatbot that pulls from a live feed of the internet, including posts from X.

In his post, Musk says the goal is to offer a faster, less censored source of truth; one that reflects what’s really happening rather than what traditional editors or institutions choose to publish.

But that mission is also where the trouble starts. Grok hasn't exactly been the paradigm of truth or moral integerity.

Why was Grokipedia delayed?

Just days before the planned debut of Grokipedia version 0.1, Musk announced that the launch would be delayed indefinitely. The reason? Too much “propaganda” in the training data.

While he didn’t clarify what content raised red flags, the post suggests a major rework is underway to align Grokipedia with Musk’s vision of “truth”—a vision that often clashes with mainstream moderation standards.

Obviously, this opens the door for bias in the opposite direction, especially since the system lacks Wikipedia’s core safeguards: transparent citations, edit histories and decentralized moderation.

But I think the billion-dollar question here is really, is Grokipedia even possible? Can Elon Musk and his team really build a real-time, AI-updated encyclopedia? Without proper transparency, source validation and accountability, it could also amplify errors or personal bias faster than traditional platforms can correct them.

Plus, Grok has already been flagged for generating inaccurate or politically charged responses. Launching a knowledge platform on top of it only raises the stakes.

Final thoughts

More from Tom's Guide

• This little-known Gemini trick is the secret to fact-checking — and it’s more accurate than ChatGPT
• 7 Copilot prompts that make Windows 11 way more useful
• I tested Windows 11’s new AI upgrades — and these 3 are actually worth using (plus, one I can't wait to try)

When it comes to chatbots, it's easy to forget about Grok because it seems like other big tech is always in the news. With Google's Nano Banana starting new trends and OpenAI's ChatGPT hyping their latest models, Elon Musk's chatbot simply exists in the background.

I've definitely found myself rolling my eyes at some of Grok's decisions, especially when it comes to image generation. However, it's clear that there are some reasons to sit in awe of what Elon Musk calls “the smartest AI in the world.”

As someone who has spent hours testing it, the truth is, it's not just hype. From near-instant web searches to jaw-dropping results on complex engineering queries, Grok 4 is delivering in ways its predecessors and rivals haven’t quite managed. Whether you love the direction or cringe at the controversies, Grok 4 may always be the underdog that quietly crushes it.

What makes xAI's Grok different

Elon Musk posted on X highlighting that Grok 4 is at the top of the ARC-AGI leaderboard. To understand why that's impressive, it's important to become familiar with how models are tracked on it.

Essentially, the ARC-AGI leaderboard is a scoreboard for AI, that not only tracks how many problems a model can solve, but also how efficiently it solves them. In other words, it's measuring both the brain and the resourcefulness of the model. High performance with low cost per task is what matters most.

So, Grok's position at the very top is extrememly significant because it means the xAI model is not only keeping up with rivals like Gemini and ChatGPT, but outpacing them on some of the toughest benchmark criteria possible.

Beating every other chatbot suggests that Grok 4 is powerful and efficient, which is exactly the type of breakthrough that supports true progress in the evolution of artifical general intelligence (AGI).

Where Grok still stumbles

Whether used on X or on the standalone platform, real-time search pulls in fresh infromation from both the web and X, so it can keep up with breaking news at a moment's notice.

However, the accuracy and bias concerns are what critics keep coming back to. Grok has made some claims that turned out false, and there are questions about how its alignment is being guided (e.g. how much Musk’s own views factor in).

The model also struggles with issues of content moderation after xAI scrambled to pull posts and update filters when anitsemitc content popped up.

The takeaway

Despite the model beating it's rivals, questions still remain like, will it stay reliable as usage increases? Will “garbage data” or bias creep back in under pressure? How well will xAI handle moderation long-term? The past controversies suggest it’s an ongoing battle.

There are no doubts that Grok is not perfect. It carries some extremely controversial baggage, but the proof of what it does better in terms of speed, real-time data and flexible thinking makes it a serious contender in the AI race.

More from Tom's Guide

• This new AI tool can predict your risk of 1,000+ diseases — meet Delphi-2M
• Students beware — no AI writing is safe with this 'black light' of AI detection
• Anthropic (finally) rolls out Memory feature — here’s who can get it

If you’ve ever chatted with Grok, the Elon Musk-backed AI assistant from xAI, and hit the “Share” button, your conversation might be searchable on Google.

According to a recent Forbes report, more than 300,000 Grok conversations have been indexed by search engines, exposing user chats that were likely intended to be shared privately, not broadcast online. This comes just weeks after ChatGPT users discovered their shared chats were searchable on Google too.

This discovery once again raises serious concerns about how AI platforms handle content sharing and user privacy.

The big picture

Built by Elon Musk’s xAI, Grok includes a feature that lets users share conversations through a unique link. Once users share, those chats are no longer private, and are getting picked up by Google’s web crawlers. From there, the chat are turning up in search results, meaning anyone can access and read them.

In both the ChatGPT cases and this one, the issue comes down to how "shareable" URLs are structured and whether AI companies are doing enough to protect users who may not realize their content is publicly viewable.

Why this keeps happening

The implications go beyond a few embarrassing screenshots. Some Grok chats could contain personal, sensitive or even private details. Many users likely had no idea their shared conversations were being published for the world to see until now.

This privacy breach highlights a growing challenge in AI product design: how to balance transparency and social sharing with greater privacy protections. When users hit “Share,” they should be clearly informed that their link will be publicly accessible, and AI companies need to use best practices like noindex tags or restricted access URLs to avoid surprises like this.

What xAI and others should do

To prevent these kinds of privacy slip-ups, AI platforms should:

• Add clear warnings before creating publicly accessible links
• Use noindex tags to block search engine crawlers
• Educate users on the privacy risks of sharing chats
• Consider obfuscated or expiring URLs for shared content

This isn’t the first controversy Grok has faced. the chatbot has come under fire in recent months for questionable responses and controversial content. But this latest issue hits at a core trust concern: how user data is managed and whether privacy is truly being prioritized.

As more people turn to AI tools for personal, educational, wellness and emotional support, platforms like xAI will need to step up their safeguards or risk eroding the very trust they’re trying to build.

Bottom line

More from Tom's Guide

• Meta AI chatbots gave false medical advice and flirted with minors — now the company is restructuring its AI division again
• I just created 5 apps using GPT-5 and it's so easy it feels like it should be illegal — here's how to do it
• Sam Altman just teased GPT‑6 — and it's more personal than ever

Elon Musk’s social media platform X is taking initiative when it comes to fighting misinformation: it’s giving artificial intelligence the power to write Community Notes; those are the fact-checking blurbs that add context to viral posts.

And while humans still get the final say, this shift could change how truth is policed online.

Here’s what’s happening, and why it matters to anyone who scrolls X (formerly Twitter).

What’s actually changing?

X is currently piloting a program that lets AI bots draft Community Notes. Third-party developers can apply to build these bots, and if the AI passes a series of “practice note” tests, it may be allowed to submit real-time fact-checking content to public posts.

Human review isn’t going away. Before a note appears on a post, it still needs to be rated “helpful” by a diverse group of real users and given proper oversight. That’s how X’s Community Notes system has worked from the start, and it remains in place even with bots in the mix (for now).

The goal is speed and scale. Right now, hundreds of human-written notes are published daily.

But AI could push that number much higher, especially during major news events when misleading posts spread faster than humans can keep up.

Why this move matters

Can we trust AI to handle accuracy? Yes, bots can flag misinformation fast, but generative AI is far from perfect. Language models can hallucinate, misinterpret tone, or misquote sources. That’s why the human voting layer is so important. Still, if the volume of AI-drafted notes overwhelms reviewers, bad information could slip through.

X isn’t the only platform using community-based fact-checking. Reddit, Facebook and TikTok have also explored similar systems.

But automating the writing of those notes is a first, opening up a bigger question about whether we are ready to hand over our trust in bots.

Musk has publicly criticized the system when it clashes with his views. Letting AI into the process raises the stakes: it could supercharge the fight against misinformation, or become a new vector for bias and error.

When does this go live and will it actually work?

The AI Notes feature is still in testing mode, but X says it could roll out later this month.

For this to work, transparency is key, with a hybrid approach of human and bot working together. One of the strengths of Community Notes is that they don’t feel condescending or corporate. AI could change that.

Studies show that Community Notes reduce the spread of misinformation by as much as 60%. But speed has always been a challenge. This hybrid approach, AI for scale, humans for oversight, could strike a new balance.

Bottom line

X is trying something no other major platform has attempted: scaling context with AI, without (fully) removing the human element.

If it succeeds, it could become a new model for how truth is maintained online. If it fails, it could flood the platform with confusing or biased notes.

Either way, this is a glimpse into the future of what information looks like in your feed and encourages asking the question of how much you can trust AI.

More from Tom's Guide

• I trust AI with my job, my schedule and my inbox — but I’ll never use it for this
• These 5 prompt upgrades took my AI videos from basic to cinematic
• I used these 5 ChatGPT prompts to uncover early Prime Day deals before everyone else

The next big thing from OpenAI might not be a new model, but a new social network. That’s right — the tech giant may be preparing to take on social media giants like Elon Musk’s X (formerly Twitter) and Meta.

According to a new report from The Verge, OpenAI is in the early stages of developing its own social platform — one that could integrate directly with ChatGPT.

In other words, imagine a social network that uses AI for more than creating art and creative prose, but for building out the site entirely based on what users post and share online.

It’s still unclear how this would work in practice or if the platform will ever publicly launch, but the exploration speaks to OpenAI’s broader push to expand beyond its core products.

What an OpenAI social network could look like

Sources familiar with the matter say OpenAI’s internal prototype centers around a social feed powered by AI, including features like image generation from ChatGPT. While details are still limited, CEO Sam Altman has reportedly been seeking outside feedback on the concept, exploring whether it could work as a standalone app or as part of the ChatGPT experience.

What is clear is that OpenAI’s potential entry into the social media space goes beyond simply giving us another place to scroll. The network would give users access to AI-powered tools to create engaging and viral content.

If it moves forward, the project will place OpenAI squarely in competition with platforms that already dominate the social media space and that are increasingly driven by AI.

Looking ahead

OpenAI's potential move into social networks would add fuel to the fire between Sam Altman and Elon Musk, two tech leaders who have been locked in a very public back-and-forth. Earlier this year, Musk offered to buy OpenAI for $97.4 billion — a proposal Altman rejected with a tongue-in-cheek reply: “No thank you but we will buy twitter for $9.74 billion if you want.”

For now, OpenAI has declined to comment on the reports, of its social network efforts and no official launch plans have been announced. But if the company’s recent product expansions are any indication, this could be a major shift in how AI meets social media — and who gets to shape that future.

More from Tom's Guide

• I let Google AI take over my desktop — it found files I thought were gone forever and other mind-blowing tricks
• This is your brain on AI — and here’s the subtle shift you’re not noticing
• I've subscribed to every AI tool — only one was worth the money

Elon Musk's Department of Government Efficiency (DOGE) has been blocked from accessing certain personal data. The ruling comes as a survey revealed 3 in 4 Americans oppose DOGE's actions.

DOGE's governmental cost cutting drive has led it to accessing the sensitive personal data of millions of Americans – potentially illegally.

The collecting of this data is something that can't be protected by tools like the best VPNs, and Elon Musk's DOGE has faced a massive backlash over its actions.

12 privacy lawsuits were filed against DOGE in the early stages of its data collection and now public opposition is growing. How DOGE and the U.S. government will respond is unclear. But what is clear is no organization should have this much freedom to access our personal data.

Data collection blocked

DOGE was dealt a blow in the last week of February when it was barred from accessing personal data held by two government agencies.

U.S. District Judge Deborah L. Boardman, in Greenbelt, Maryland, blocked DOGE from accessing personal data held by the Education Department and the Office of Personnel Management.

A temporary restraining order was issued on behalf of a coalition of labor unions who argued it was wrong for the agencies to grant DOGE access to the personal data of millions of Americans.

Data included Social Security numbers, dates of birth, home addresses, income and assets, and citizenship and disability status.

The judge stated the agencies' decision to grant DOGE "sweeping access" to this data violated the Privacy Act 1974 – the major data privacy defense at the heart of DOGE's lawsuits.

The restraining order is in place until 10 March, but opposition is expected to continue.

3 in 4 Americans oppose DOGE's actions

A survey conducted by AllAboutCookies has found that nearly 3 in 4 Americans oppose DOGE accessing their data.

When asked if they were in favor of DOGE and Elon Musk having access to taxpayers' personal data, 74% of respondents said no.

1,000 respondents were surveyed and 61% said they were aware of DOGE's actions before taking the survey. The fact that a majority of those surveyed knew about DOGE's actions is a good thing, but 39% not knowing is cause for concern – people have the right to know if their personal information is being exposed.

Of those who were aware of DOGE's actions, 57% said it weakened their trust in the government's handling of their data. 27% said it made no impact and 15% said it made them trust the government more.

Elon Musk divides America

Before joining DOGE, Elon Musk was a polarising figure. This has only increased in recent weeks, fuelled by the revelation that Musk is not actually a DOGE employee.

The White House declared that Musk is a "senior advisor" to President Trump, rather than a federal employee, and this has led to questions over the legality of his access to agency data.

AllAboutCookies' survey highlighted Musk's divisive nature. A majority of respondents did not trust Musk in any capacity relating to the access and collection of personal data.

40% had "complete distrust" in Musk's "safe handling of personal data," with 17% having "some distrust". Respondents were also asked how much they would trust Musk would "only use personal data for publicly-stated purposes" and "will not let personal or business interests influence decisions at DOGE."

41% answered "complete distrust" to both questions, 15% expressed "some distrust" to the former, and 14% expressed "some distrust" to the latter.

Between 20 and 22% were neutral to all three statements, 11-12% had "some trust," and 10-11% had "complete trust."

52% expressed concern over the safety of their personal data, with 29% being "somewhat concerned" and 19% "not at all concerned".

Regardless of whether VPNs can help in this case or not, it is still important to follow good data practices and secure your data where possible. The best password managers and the best encrypted messaging apps, alongside the most secure VPNs, are vital tools in protecting your data.

30% of those surveyed said they'd taken action with their personal data in response to DOGE's measures. 16% had updated their bank details and 11% had locked or frozen credit.

13% had bought credit monitoring and 11% had bought ID theft monitoring. This shows an appetite amongst Americans to protect their data and these tools are featured in some premium VPN subscriptions.

ExpressVPN and NordVPN both offer forms of identity protection through their Identity Defender and NordProtect features respectively. Both feature cyber insurance of up to $1 million and monitor the web for breaches of your data.

A credit scanning tool is also included in NordProtect, but ExpressVPN has a dedicated Credit Scanner feature, separate from Identity Defender.

We test and review VPN services in the context of legal recreational uses. For example:1. Accessing a service from another country (subject to the terms and conditions of that service).2. Protecting your online security and strengthening your online privacy when abroad.We do not support or condone the illegal or malicious use of VPN services. Consuming pirated content that is paid-for is neither endorsed nor approved by Future Publishing.

12 privacy lawsuits, 6 federal agencies, the data of millions reportedly at risk – all in less than 30 days. The newly created Department of Government Efficiency (DOGE) – an "advisory body" led by Elon Musk – has caused quite a stir since its inception.

Tasked with cutting governmental costs and weeding out fraud, waste, and abuse, DOGE has ordered federal agencies to sack thousands of government employees and delved into agency systems holding sensitive information concerning millions of Americans.

Usually when we're discussing data privacy, tools like the best VPNs and the best password managers are cited as effective remedies – but this is different.

The data DOGE wants access to isn't the kind that can be protected by individuals. It includes legal names, social security numbers, race, disabilities, income, and even citizenship status and whether you've been incarcerated.

Lawsuits claiming Musk, Trump, and DOGE are breaking the law have piled up. Serious data privacy concerns have been raised regarding DOGE's actions – many stemming from the fact that Musk is not an official government employee, and should not have access to sensitive data.

As noted in a White House declaration, the tech billionaire "has no actual or formal authority to make government decisions himself" and is not the administrator in charge of DOGE's day-to-day operations. Instead, Musk is described instead as a "senior advisor" to President Trump.

So, is DOGE breaking the law, and what do these measures mean for American data privacy?

What has DOGE done?

Put simply, there are serious data privacy concerns surrounding DOGE's actions. DOGE is aiming to access sensitive federal agency data in order to carry out its cost-cutting effort.

DOGE's executive order states it should have "full and prompt access to all unclassified agency records, software systems, and IT systems."

However, DOGE having access to this data is claimed to be unlawful and violates the Privacy Act 1974. It isn't always clear who those accessing the data are, and the more the data is accessed and moved, the greater risk of a breach.

A White House official told Reuters that all DOGE staff – many of whom are recent tech graduates and were recruited via Discord – are federal employees who have the appropriate security clearances. However, questions are still being asked over the freedom given to seemingly inexperienced DOGE staff.

This includes 19-year-old Edward Coristine, a known leaker and someone who appears to have frequented Discord and Telegram communities focused on cybercrime. It is alleged he solicited recommendations for a third-party service that could perform cyberattacks, and now has supposedly been granted access to the State Department and Department of Homeland Security (DHS) as an "advisor."

Al-Jazeera reported that DOGE is pushing to access the IRS' protected Integrated Data Retrieval System (IDRS). IDRS data relates to sensitive tax information, including social security numbers, home addresses and income status.

It is unclear whether this has been achieved. As of February 20, access hadn't been granted.

The Labor Department has given DOGE access to vast amounts of its data. Labor Department employees were reportedly worried about the access available to inexperienced staffers, and the move was the opposite of what the department would typically do to protect privacy.

It has also been reported that DOGE is feeding information from the Education Department into AI software to hunt for potential spending cuts. The issues around AI and data privacy are still emerging, but AI having access, and potentially storing and learning from this data, is a concerning development. It is unclear, but unlikely, that the individuals whom this data concerns have consented to this usage.

It isn't known how secure DOGE's data storage facilities are, and a cyber attack could have disastrous consequences for millions of Americans. It is unknown how DOGE plans to transfer the data it collects and how secure said transfer and storage will be.

A dozen legal challenges

There has been significant opposition to DOGE's actions, primarily in the form of lawsuits. At the time of writing, 12 privacy-related lawsuits are in place, as well as other non-privacy related cases.

The Electronic Frontier Foundation (EFF), along with other privacy advocates, filed a lawsuit on February 11 against DOGE's accessing of the Office of Personal Management data.

The EFF said that "the mishandling of this information could lead to such significant and varied abuses that they are impossible to detail."

On February 7, the American Civil Liberties Union (ACLU) filed Freedom Of Information Act (FOIA) requests with over 40 federal agencies. It labelled DOGE's acts as "secretive" and called on Congress to "fulfill its constitutional role" in opposing DOGE.

A coalition of 14 state attorney generals filed a lawsuit aiming to immediately block DOGE from accessing information at the departments of Labor, Education, Health and Human Services, Energy, Transportation and Commerce, and at the Office of Personnel Management.

However, on February 18, a judge rejected the lawsuit. Musk and DOGE's "unchecked authority" was noted, but, according to Reuters, the judge said "the states had not shown why they were entitled to an immediate restraining order."

"The recent reporting around federal agencies granting DOGE staff access to the most sensitive information about Americans is alarming," said the Center for Democracy & Technology's Elizabeth Laird.

Laird cited the lawsuits questioning the legality of DOGE's actions and highlighted the "several security incidents" involving DOGE staff, "suggesting that they are not adhering to best security practices in their work."

She went on to say DOGE needed to be more transparent. "A necessary first step to addressing the access and security issues is answering questions core to protecting any sensitive information: Who has access to this information and under what authority? What is being done with this information? What are the proactive steps that are being taken to keep it safe and secure?"

What can you do?

The problem is that individuals cannot take steps to protect this data. It is the government's duty to protect it, and it is information that cannot be withdrawn.

The primary defense at this stage is lawsuits and the hope of DOGE's behavior being declared unlawful – but this requires a concerted effort and is not a quick or simple solution.

If you are concerned about legitimate but unwanted use of your personal data, then a data removal service such as Incogni can help. These services contact data brokers who hold records of your data and submit removal requests on your behalf.

However, if information collected by DOGE was to suffer a breach, data removal services are unlikely to protect you against any malicious use of your data.

It is possible your information could be used in phishing and scam attempts. Always be cautious of suspicious emails asking for any personal information or messages claiming to be from "reputable" sources that don't look quite right.

Look out for typos, demands for payment, and poorly put together websites. Never click on anything you're not 100% sure of.

If you are the victim of a breach or cybercrime, leading VPN providers ExpressVPN and NordVPN offer forms of identity protection in their plans.

ExpressVPN's Identity Defender and NordVPN's NordProtect provide cyber insurance of up to $1 million, as well as dark web monitoring and assistance in recovering lost funds and identity theft.

At this stage, there are no reports of data being breached as a result of DOGE's actions, but it is something to be aware of should the worst happen.

Consent and Privacy Act 1974

Introduced in the wake of the Watergate scandal, the Privacy Act 1974 has been cited in almost every legal case against DOGE.

Under the act, the disclosure of personal records, even within agencies, generally requires the consent of the individual concerned. However, there are several exceptions, such as law enforcement investigations and court orders – and two of these could work in DOGE's favour.

As stated in Lawfare, agencies can share records with third parties for "a routine use" if this "use (including sharing) would be compatible with the reason the information was collected in the first place."

But many in opposition to DOGE's actions are arguing its use of the data has nothing to do with the reasons it was collected in the first place. Many claim DOGE wants the data for political reasons and it is unclear whether DOGE's staff have been vetted, and there does not appear to have been any congressional authorization for the data collection to take place.

DOGE has also been accused of failing to record what actions it is taking, and this, combined with the lack of an established leader, makes accountability almost impossible.

How far will the government go?

The relationship between Trump and data privacy is a complicated one. In the past, Trump has been a vocal critic of companies like TikTok collecting the personal data of Americans. He called for a TikTok ban during his first term in office, but has since reversed his stance on the issue.

Historically, his actions have been more to do with protecting national security rather than concern for the rights of the American people. Project 2025, which Trump has now distanced himself from, criticizes China's data collection practices but fails to mention the data privacy practices of US companies such as Apple, Google, and Amazon.

Elon Musk is clearly not a champion of data privacy rights, with privacy complaints being made over X user data being used to train its Grok AI model, and it is unknown how much influence he will have as a special advisor.

However, Musk's orders appear to come directly from the President – which leads us to believe Trump is happy with DOGE accessing federal data.

The separation of federal and state government also poses questions, because states have differing data privacy laws. How these laws interact with DOGE's federal actions, and whether they can be used to combat them, is unclear.

US lawmakers have condemned moves by the UK government to access the encrypted data of Apple's users – lawmakers Senator Ron Wyden and Congressman Andy Biggs said in a letter that the demand threatens the privacy and security of the US.

Senator Wyden, along with Senator Elizabeth Warren, also expressed concern over DOGE's access to IRS tax data. However, contradicting his stance on the UK government's actions, Congressman Biggs introduced legislation in support of DOGE's initiatives.

We have seen countless examples of governments imposing on the data privacy of their citizens in recent months. However, these have rarely involved a Western government. DOGE's actions represent a significant shift in attitudes towards Western data privacy, and is a cause for concern – leaving us asking how far Trump and DOGE would go.

We test and review VPN services in the context of legal recreational uses. For example:1. Accessing a service from another country (subject to the terms and conditions of that service).2. Protecting your online security and strengthening your online privacy when abroad.We do not support or condone the illegal or malicious use of VPN services. Consuming pirated content that is paid-for is neither endorsed nor approved by Future Publishing.

The next iteration of the Grok AI model, Grok 3, has been described as being "10x" more powerful than its predecessor after it was unveiled yesterday, Monday 17 February. During a live streamed announcement, the xAI team — joined by Elon Musk — revealed their response to OpenAI's GPT-4o and Google Gemini.

“Grok 3 is an order of magnitude more capable than Grok 2,” Musk said during the event. “[It’s a] maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct.”

Last month, Musk tweeted that pretraining on Grok 3 was complete and it boasts "10X more compute than Grok 2."

Yesterday's event included a series of demos of the flagship new model performing tasks including mapping a spacecraft mission from Earth to Mars as well as creating a new game that's something of a mashup between Tetris and Bejeweled.

Grok of Ages

Grok 3 was initially planned to launch in 2024 with xAI using 200,000 GPUs in a data center in Memphis to train it. As well as Google Gemini and ChatGPT 4-o, the team are also positioning Grok 3 as going toe-to-toe with DeepSeek-V3 and Claude 3.5 Sonnet.

Grok 3 actually consists of a family of models including Grok 3 Reasoning and Grok 3 mini. The latter will respond to questions faster, possibly at the expense of some accuracy.

Meanwhile, xAI also claims Grok 3 Reasoning is able to outperform the top version of OpenAI's o3-mini — o3-mini-high.

According to xAI, Grok 3 beats GPT-4o on several benchmarks, including PhD-level physics and biology questions. The model will begin rolling out on Monday and, according to xAI, early versions of it have also scored competitively in Chatbot Arena. This pits various AI models against each other in a crowdsourced environment with users voting on which they feel is most accurate.

We'll have to wait to get our hands on Grok 3 ourselves to give our own thoughts, but xAI's older model held it's own pretty well when we put Grok 2 into a 7-round photo face-off against Gemini.

Grok started as a feature within Musk's social media platform X but has since been spun out into a standalone app available in the App Store in the U.S., Australia and India.

Currently the standalone Grok app is only available for iOS, although you can still use Grok within the X app on Android. It isn't clear when an Android Grok app will launch or when it will be available worldwide. A look on the Google Play store simply claims the app is "coming soon."

More from Tom's Guide

• I put ChatGPT vs Grok to the test with 7 prompts — here's the winner
• xAI's standalone Grok iOS app launches in the US — here's how to find it
• I tried the open source AI tool that creates audiobooks for free — here's my verdict

As OpenAI attempts to become a for-profit venture, Elon Musk appears to be a throwing a wrench into the works.

A report from The Wall Street Journal revealed today that Musk, alongside a number of investors, is offering nearly $100 billion to buy the nonprofit that controls OpenAI. The investor group includes Ari Emanuel and Musk's own artificial intelligence company, xAI.

At the time of writing, OpenAI has not officially released a statement about the bid. However, OpenAI's chief executive, Sam Altman, did post on Twitter, "No thank you but we will buy twitter for $9.74 billion if you want."

Musk, apparently, unamused, responded, "Swindler." He has continued making anti-Altman posts on Twitter including a video where he wrote "Scam Altman."

This is the latest development in the ongoing feud between two of tech's biggest AI supporters in the United States.

The bid was said to have been submitted to the OpenAI board of directors by Marc Toberoff, a lawyer who is also leading a lawsuit instigated by Musk against OpenAI last year.

The move also comes as OpenAI is in the midst of another fund-raising push as part of project Stargate, a Trump-backed plan to invest $500 billion in AI infrastructure in the United States. It shares the same name as a project introduced last year by OpenAI and Microsoft.

At the same time, Altman has been in the process of turning OpenAI from a nonprofit venture into a for-profit company. It's unclear if Musk's bid alters those plans as some valuations have the company set at $157 billion.

An ongoing feud

This feud goes as far back as 2015. OpenAI was founded by Altman, Musk and several others that year.

Publically, the feuding seemed to begin in 2018 when Musk left OpenAI's board of directors. It was initially described as a conflict of interest removal. However, later reporting seemed to reveal a more contentious ousting as Musk apparently attempted to take on chief executive role and merge OpenAI with Tesla.

Five years later in 2023, Musk launched xAI which quickly released the AI chatbot Grok that was described as an anti-"woke" response to OpenAI's ChatGPT.

Last year, things ramped up in March when Musk first launched his lawsuit claiming that Altman had violated the company's principles. That lawsuit continued in November with a preliminary injunction to OpenAI from becoming for-profit.

The announcement of Stargate during the first week of the second Trump administration induced a childish war of words on Twitter between Musk and Altman with Musk claiming OpenAI"s main backer, SoftBank, didn't have the money. A brief tit-for-tat ensued from there. Musk called Stargate "fake" and questioned Altman's loyalty to Trump.

Now, we have this $97.4 billion unsolicited bid and mocking words from Altman.

What happens next?

While both men are billionaires, Elon Musk is quite possibly the richest person on Earth with estimated worth of $392 billion. Altman, by comparison, is worth $1.2 billion, largely thanks to investments in Reddit and Stripe. Allegedly, he does not have equity stake in OpenAI.

They both also seem to have President Trump's ear, with Musk running amok through the government with his DOGE plan and Altman's support of the Stargate program which includes building data centers throughout the United States — something that the administration isn't actually paying for and was already in progress before Trump was sworn in.

The feud appears no closer to ending or coming to an amicable end.

More from Tom's Guide

• OpenAI introduces new features for WhatsApp users — here’s what’s new
• OpenAI’s ‘o3-mini’ is free for all users — what you need to know
• It doesn't matter if DeepSeek copied OpenAI — the damage has already been done in the AI arms race

With the release of iOS 18.3 comes a surprise release, as T-Mobile users with access to the Starlink Beta can now connect to Starlink satellites on compatible iPhones.

Originally spotted by Bloomberg, it seems that Apple has secretly worked with SpaceX and T-Mobile U.S. to provide an alternative satellite service. This is quite a surprise, as T-Mobile had previously specified Starlink as an option for Samsung phones, like the Galaxy S25 Ultra and the Galaxy Z Fold 6. Meanwhile, Apple has an in-house Satellite connection service due to a partnership with Globalstar Inc.

However, the Bloomberg report states that a select number of iPhone users have been able to enable the Starlink beta through T-Mobile. While we don't know which models will be compatible as of yet, T-Mobile told Bloomberg that the full release will support the vast majority of iPhones.

We also have an idea of what features it could bring, as Elon Musk posted on X that Starlink currently supports the transfer of images, music and even Audio Podcasts. He also said that the next generation will be able to transfer medium-quality video.

Currently, the T-Mobile Starlink beta is limited, but there are plans to expand the test to additional subscribers. If you want to join the beta then you can sign up on the T-Mobile website. However, it is worth noting that Starlink is only available in the U.S., and while it might be free during the beta, that is likely to change.

iPhone users in the program need to have updated to iOS 18.3, then they will see a new toggle in their iPhone cellular data settings to manage the feature.

There are certain advantages to satellite communication, especially when it comes to emergencies. In the case of Apple devices, it was introduced for iOS 18 and is supported on the iPhone 14, iPhone 15 and iPhone 16 series. The feature is easy enough to use and seamlessly integrates with iMessage. The feature will activate as soon as the phone detects no cellular or Wi-Fi signal and supports both standard SMS and iMessage.

While it is good to see that Apple users will have more options, the real question is how much a Starlink connection will cost. The issue is that, outside of emergencies, the ability to send images and podcasts via the service seems minimal for the average user.

More from Tom's Guide

• Latest iPhone SE 4 rumor adds to the notch-or-Dynamic Island speculation
• Forget Samsung S25 Edge — this leaked image allegedly shows the iPhone 17 Air
• iPhone 17 rumor reveals some Dynamic Island info — and it’s not good news

Earlier this week, Bloomberg released a report claiming that Chinese officials are looking into potential sales of the US version of TikTok to avoid the TikTok Ban. One possible buyer floated in the report is Tesla/X (nee Twitter)/SpaceX CEO Elon Musk.

This was partially put out since Elon Musk was a major supporter of the Donald Trump administration, having contributed over $250 million to Trump's reelection campaign. Trump himself has signaled that he might reverse the TikTok ban that was actually started under his first administration and finalized under the current Biden presidency.

Currently, ByteDance is appealing the ban in the US Supreme Court, though there are early signs that the court plans to uphold the ban. If that is the case, it means that ByteDance would need to divest its stake in TikTok or effectively remove the popular social media app in the United States.

"A potential high-profile deal with one of Trump’s closest allies holds some appeal for the Chinese government," Bloomberg wrote.

Strangely, the article reports that this may be a discussion between Chinese officials and US-based officials and companies without the knowledge of ByteDance.

ByteDance denied the report to Variety. A spokesperson said, "We can't be expected to comment on pure fiction."

In response, TikTok users have switched to another Chinese app, RedNote, which recently shot to the top of app store charts. That app is currently Mandarin only and has faced similar censorship criticisms as TikTok, which is notorious for inducing childish doublespeak in users to avoid saying words like kill, death or suicide on the app.

There is speculation that the move has been partially spite-driven as "TikTok refugees" purposefully pick another Chinese app over American companies like Reels, Threads and X.

At the same time, recent moves by Meta, which owns Instagram, Whatsapp and Facebook, have reportedly been seen as unappealing to younger generations. This includes Meta removing its fact-checking department and a rightward turn by CEO Mark Zuckerberg, who recently went on the popular Joe Rogan podcast and decried previous attempts by the Biden administration to get disinformation off Meta's platforms.

For now, TikTok still exists in the US, but who knows for how long, if it will get sold or if Trump will reverse the ban.

The equally controversial YouTuber, Mr. Beast, seemingly joked on X that he would buy TikTok so it doesn't get banned. A recent Parade report claimed the YouTube star is worth $1 billion and a Forbes article claimed he earns $85 million a year. Numbers that are probably not enough to purchase TikTok, which could be worth as much as $100 billion.

More from Tom's Guide

• DJI Flip review: This IS the drone you’re looking for
• Don’t throw it out! 7 ways you can reuse plastic bottles around your home
• Facebook, Instagram and YouTube have privacy practices that could put your data at risk

Elon Musk's social media site X will be increasing the cost of one of its main subscription plans from today for any new users joining the service.

X has seen a tumultuous few years since Musk bought it as Twitter back in 2022. One of the most talked-about introductions was the new subscription plans in an effort to make the site profitable. The infamous Blue Tick was, in the past, introduced to confirm users' identity and was a free feature so long as certain steps were followed. However, In 2022 X revealed that users would need to apply to one of the subscription plans to be eligible for the tick.

In a recent post on its help page, X revealed the cost of the Premium Plus Subscription will be increasing. The tier will now cost members $22 a month, $6 up from the prior $16. Meanwhile, the Premium X Plus annual cost will go from $168 to $229.

Current subscribers have until January 20, 2025, before needing to pay the new cost, but new customers will have to pay them from today.

What do you get for the price?

These are some pretty hefty increases, but X has defended them with three key points. The first is the Premium Plus will be completely ad-free. Secondly, users will have access to new features such as Radar, Grok AI and more. Finally, the new price will help to support X's Creator Program thanks to a shift in its revenue share model. This new fee will contribute to the "new, more equitable system where creator earnings are tied to the overall value they bring to X, not impressions of ads."

We've seen a fair bit about X in the news of course, but most of the focus has been on the abilities of its new AI model: Grok. One of the biggest additions was the new Grok Aurora image generator which aims to create life-like images. When we compared it to Gemini, Grok managed to score ahead thanks to its ability to create life-like images.

Price increases are never nice, and these are pretty high, however, it will take time to see if they affect the number of people willing to subscribe.

If you find Elon's latest change a little hard to get over, then you can see our 5 best Twitter alternatives to try instead.

More from Tom's Guide

• OnePlus Open 2 could launch way later than expected — rumor points to late 2025
• iOS 19 could come to all iPhones with iOS 18 — but iPads won’t be so lucky
• I turned my friends into Genmoji stickers in iOS 18.2 — here's how you can too

On Thursday, Elon Musk and Tesla finally held the long-teased 'We,Robot' event where Tesla CEO Elon Musk revealed humanoid robots, a robotaxi and a 20-seat 'robovan.'

The event was held on a Warner Brothers lot in Los Angeles, and didn't get off to a great start. The title is a reference to "I, Robot" a book of science-fiction short stories by Isaac Asimov, which seems to echo previous Musk statements referring to Tesla as an "AI robotics company" and not a car maker.

Musk was an hour late to take the stage, claiming that the delay was due to a "medical emergency" for an attendee.

The biggest reveal was the Cybercab, a gold painted vehicle that is a more sci-fi looking scissor-door Model S with LED headlights that displays laser projections on the road. The "driver's seat" distinctly lacks a steering wheel or pedals.

Musk took the Cybercab through a staged demonstration where the robot taxi meandered through various movie sets stopping at junctions or pausing for staged cyclists.

During the event, Musk sclaimed that there were "50 autonomous cars" driving around the Warner Bros. lot, which was supposed to include Model Ys in the built-in autonomous driving mode.

"The autonomous future is here," Musk said. "With autonomy, you get your time back."

Beyond that though, Musk did not describe the technology Tesla is utilizing the in the Cybercab. Though he did reveal that the Cybercab will go on sale for "less than $30,000." He said he expected the autonomous vehicle to start production some time in 2026.

For those watching the livestream of the Tesla event, it was a lot of guesswork to figure out what's actually in the Cybercab. For example, it doesn't appear to run on LiDar or Musk didn't mention it. This suggests the cab is running with high-def cameras to navigate and "see" obstacles.

Inductive charging was mentioned, which means the Cybercab lacks a plug and relys on inductive plates to charge the batteries.

Musk admitted he might be overselling it here, "Well, I tend to be a little optimistic with time frames. But in 2026. So yeah. Before 2027. Let me put it that way."

Analyst Ming-Chi Kuo said on X that he believed the Cybercab wouldn't actually hit mass production until 2027.

Unlike Alphabet's Waymo, Tesla does not currently legally have the ability to run unsupervised vehicles on roads in California and Texas.

In general, the event was a resounding disappointment, especially with investors. Dennis Dick of Triple D Trading told Reuters, ""I'm a shareholder and pretty disappointed. I think the market wanted more definitive time lines. I don't think he said much about anything."

For other investors, the event proved to be "underwhelming and stunningly absent on detail," as Bernstein anaylst Toni Sacconaghi told MarketWatch.

Much of the disappointment seems to stem from a lack of details including when Tesla when gain approval to allow their vehicles on real roads and not staged backlots or the actual cost per mile to run a Cybercab.

Robovans and Robots

In a surprise, Tesla also revealed a 20-person Robovan, a transportation pod that somehow looks even more sci-fi than the Cybercab with some renderings showing the pod as a sort of food truck. There were no other details about the pod.

They also showed more of the Optimus humanoid robots that Musk has shown off previously. Musk claimed the robots would be sold for $20,000 suggesting that they would be "the biggest product ever, of any kind."

The robots which were supposedly tending bar during the event appeared at best to be props, and Musk didn't provide any more information on what the robots would actually be capable of or when they might produced.

As with the Cybercab the theme of all of these revealed autonomous products was a flashy unveiling with little to no actual details on what powers them or when they would truly be available.

They remain, as before, pie in the sky.

More from Tom's Guide

• Elon Musk is wrong about Apple Intelligence security — here's why
• 31 million users impacted by Internet Archive data breach — what we know
• Apple TV Plus' new psychological thriller show just dropped — and critics are saying it 'demands to be watched'

Elon Musk has a problem with Apple partnering with OpenAI, raising concerns over security and data privacy — but everything he suggested is a misunderstanding of how the deal, and Apple Intelligence actually works. 

In a series of rants on his own social media platform X, Musk calls Apple out for not being smart enough to make its own AI, saying he won’t allow Apple devices in his facilities and claiming the deal will result in a privacy nightmare.

The problem is he seems to have only half watched the WWDC Keynote, ignoring the parts where Apple talked about all data processing happening on device, the ChatGPT interactions being opt-in and only your prompt being sent to ChatGPT.

Musk has already been community noted for one of his posts, where readers highlight misinformation and explain what the actual situation is. Much of this stems from his ongoing beef with OpenAI — which he helped fund in its early days.

What are Elon Musk's concerns?

Musk sees anything OpenAI does as a privacy risk, claiming anything you give ChatGPT is used to further train the model. Both OpenAI and Apple say this won't be the case when sending data from Siri. 

Musk wrote that: "If Apple integrates OpenAI at the OS level, then Apple devices will be banned at my companies. That is an unacceptable security violation."

He has since doubled down on the idea of this deep integration, even sharing memes and other posts related to the concept. His supporters have also been spreading the false narrative — despite community notes and arguments against.

In one of his X posts Musk calls Apple "too stupid" to make an AI model. This despite the fact Apple Research has published multiple models and the company says its entire Apple Intelligence structure is built on its own models.

"It’s patently absurd that Apple isn’t smart enough to make their own AI, yet is somehow capable of ensuring that OpenAI will protect your security & privacy," Musk wrote on X. He added: "Apple has no clue what’s actually going on once they hand your data over to OpenAI. They’re selling you down the river."

Apple has a detailed explanation of how they make the local models work, how they secure data sent to their own cloud servers and how they created the models to be as efficient as possible. They also explain the security of the ChatGPT link.

How is Apple working with OpenAI?

Apple's partnership with OpenAI is similar to its partnership with Google or Wikipedia; it uses an API to send data to the third-party service but creates the connection in such a way that it feels seemless to the end user.

Contrary to speculation from Musk and others, Apple isn't using OpenAI servers or models to manage your data, handle phone call transcript or any other task ascribed to Siri during the keynote or state of the union. It is just one aspect and opt-in.

Essentially you chat away to Siri, ask it to do something it can't manage with its own on device or Apple Private Cloud model and it'll ask permission to send it to ChatGPT. At which point you can agree or reject.

Apple and OpenAI say only the specific query is sent to ChatGPT and the AI model isn't trained on any query coming to it through Siri or Apple Intelligence. That won't be the case if you use the OpenAI ChatGPT app — but is the case for Siri.

The company wrote: "You control when ChatGPT is used and will be asked before any of your information is shared. Anyone can access ChatGPT for free, without creating an account. ChatGPT subscribers can connect accounts to access paid features within these experiences."

More from Tom's Guide

• ChatGPT-4o on iOS 18 is not the Siri replacement we predicted — but that is a good thing
• Apple WWDC 2024 announcements: Apple Intelligence, iOS 18, iPadOS, Siri 2.0 and more
• Apple's integration with ChatGPT is just the beginning — Google Gemini is coming and maybe an AI App Store

We won't need to work in the future, according to Elon Musk, CEO of Tesla Motors and the founder of Neuralink. He was speaking during an interview at the startup and tech conference VivaTech 2024.

Speaking of the rapid development of AI over the last few years, Musk was asked about the future and whether there would be jobs for humans in an AI-age. He said that in the most likely benign scenario none of us would have a job but there would be a "universal high income with plenty of goods and services."

Musk says that AI will handle everything for us and humanity will have to work out what to do with its time and energy. "The question won't be one of lacking goods and services, the question will be one of meaning," he said, adding “If the computer and robots can do everything better than you, then does your life have meaning?”.

"Any job that somebody does will be optional. If you want to do a job as a hobby then you can do a job, otherwise the AI will provide any goods and services you want. That is the most likely outcome," Musk said.

Ensuring AI safety

During his hour-long remote keynote Musk also reiterated that he continues to believe that the development of future AI and large language models (LLM) should be paused. He signed a letter calling for this before the release of GPT-4o.

This contradicts with his decision to launch is own AI lab xAI and the Grok chatbot which is "the funniest chatbot" and deeply integrated into X. Musk says a pause is highly unlikely so "it's better to be a participant than a spectator."

Speaking again of xAI Musk indicated that he doesn’t agree with most of how AI is being developed right now. “You think of AI as growing an intelligence in the same way you raise a child,” he added. “It matters what you teach, building it with values.”

However, he also added that the potential applications across education can be really profound. Since an AI is patient, knowledgeable, and “almost always correct”, it would make for an excellent teacher for kids — even though parents would still have to be responsible for morals and values.

“It would be like having Einstein for a teacher.” But, he also worries that kids today are being trained by social media, which is essentially governed by a “dopamine-maximizing AI”. Parents should work to limit their kids’ screen time across social media platforms, he said.

Speeding up the human-computer link

Musk thinks the brain-computer interfaces developed by Neuralink could give humans an edge in this race against obsolescence by AI.

Neuralink’s brain-computer interfaces are currently focused on treating long-term disabilities caused by brain and spine injuries. But, the ultimate goal is to develop a high-bandwidth interface that could give human beings an advantage against the rapid development of a digital superintelligence, claimed Musk.

Explaining further, Musk said that we all have a third layer above the limbic system and cortex of our brain that’s responsible for the speed at which we communicate with electronic devices.

“But the rate of communication is very slow. The sustained bits per second output of a human is just 10 bits per second.” Neuralink’s goal is to develop a high-bandwidth interface that could speed up the rate at which humans communicate with electronic devices like phones or computers.

More from Tom's Guide

• The future of AI hardware is no hardware at all — why I think Rabbit and Humane just want to be bought by Apple
• OpenAI says ChatGPT will be so good within a year we will talk to it like a human
• I write about AI for a living — 5 new features I want to see on the next iPad

Elon Musk has recently expanded access to the Grok chatbot on X. Previously it was only available to those on a Premium + account. Now anyone paying can talk to the AI.

With the expanded access they’ve also more deeply integrated Grok into the X platform, using it to summarize trending topics on the sidebar and as a way to get more information on what is happening around the biggest news stories and talking points of the day.

As Google faces increasing competition from AI-powered startup Perplexity, and OpenAI is rumored to be building a ChatGPT-powered search engine — has Musk stumbled on the answer to improving search functions? Possibly but it still needs some work.

Using Grok for search

Grok is a large language model as well as a chatbot. This means it has been trained on vast amounts of data from all of Wikipedia to large swathes of the internet. Even without live access to every post, repost and comment on X — it is a useful source of information.

However, that deep integration into one of the most active social networks, particularly one so heavily used by the media and industry leaders, seems to have given it a secret sauce.

I tried it out on a few trending topics, including the upcoming battle for search dominance between Google and OpenAI and it was able to offer up a rounded analysis of the situation, links to X posts of people talking about what is happening and a way to ask for more detail.

Being a chatbot — which is something OpenAI will do with ChatGPT Search — allows you to dive deeper into a topic. Having access to X posts also allows Grok to identify the direction of conversation, trends within a specific topic and pull it together in a conversational way.

How does Grok compare?

I haven’t done an exhaustive test of Grok, in the way I have for Gemini, ChatGPT and Claude 3 but my initial impression is that it is more open. It will answer questions that the other bots have been told to swerve around — even something as innocuous as who named it. Note — it was Elon Musk, inspired by the Robert A. Heinlein classic Stranger in a Strange Land.

I asked Grok if it was good for search and it proved it can be somewhat self-deprecating but still boastful. It declared: “I'm not exactly a search engine in the traditional sense, but I do have access to a wealth of knowledge and can provide you with information and insights based on the data I've been provided.”

It comes in regular and fun mode. The big difference is how open its conversational style is and how "off the rails" it might go. Microsoft Copilot offers a similar scale of creativity versus accuracy. I’ve tried both modes and didn’t notice a massive difference in my tests.

So I tried it on the most recent news in the tech world — Apple’s new iPad lineup and M4 chip — to see how it performed on a very recent piece of information that wouldn’t be in its training dataset. AI models only have access to data up to the end of their training, unless like Grok or Google Gemini, they have an external source of information.

Grok was able to offer me a summary of what happened during the "Let Loose" event (although you'll get more detail from Tom's Guide) as well as X posts from leading Apple commentators. I was even able to ask it why the M4 is good for AI and got a breakdown of the announcement put into context along with more X posts giving specific responses and analysis.

Conclusion

Grok isn’t perfect. It hallucinates and doesn’t always get the right X posts or information, particularly for less well-known topics or people.

I asked Grok about the technology journalist Ryan Morrison and while it was able to give a good description of my career but didn’t use any of my X posts, instead calling up posts from random Ryans.

What it has going for it is a deep understanding of what is trending, stories getting a lot of attention and the ability to analyze that information and present it in natural language.

More from Tom's Guide

• OpenAI rolls out memory in ChatGPT for all paid users — here's what it means
• Rabbit R1 review: Avoid this AI gadget
• Google Chrome users will soon be able to access Gemini AI directly from the search bar — here's how

Noland Arbaugh is the first human to receive Neuralink’s brain implant. He recently took to the social media platform X to show off the capabilities of the device by controlling a computer with nothing but his brain during a livestream.

It sounds like Arbaugh, a 29-year-old man who was paralyzed from the shoulders down in a car crash eight years ago, uses the Neuralink brain implant for more than just moving a mouse cursor. During the stream, he talked about using the connection to play online chess matches and the popular strategy game Civilization. 

Arbaugh also discussed the intuitive feel of moving the mouse cursor around using his mind. “I would attempt to move, say, my right hand left, right, forward, back, and from there, I think it just became intuitive for me to start imagining the cursor moving,” he said.

The stream was relatively short, around nine minutes, but it gave us enough of a taste to start imagining how a device like Neuralink’s brain implant could completely change the lives of people with paralysis.

From the outside looking in, it all seems magical. And for users like Arbaugh, it’s even more special. “I just can’t even describe how cool it is to be able to do this,” he explained.

Human trials only started recently, as the US Food and Drug Administration authorized the Elon Musk-co-founded company to begin testing last year. In the fall, the firm started recruiting paralyzed participants to test out its tech, with Arbaugh being the first to show off the device's capabilities inside a human being.

The chip is placed into the user's brain using a surgical robot, which Neuralink developed for exactly this purpose. And while it sounds scary, Arbaugh said, “I think, like, there's nothing to be afraid of. The surgery was super easy. I literally was released from the hospital a day later.”

In January, Musk posted on X that Arbaugh was recovering well. He followed that up in February, saying that he was fully recovered and could control a computer using his thoughts. The shared stream confirmed the billionaire’s claims.

While the surgery and recovery appear to have gone smoothly, Arbaugh acknowledged that the struggle wasn’t over and that there was work left. “I don’t want people to think that this is the end of the journey. There’s a lot of work to be done,” he said. “But it has already changed my life.”

Besides Arbaugh, we don’t know how many test subjects Neuralink has using its brain implant. The company isn’t on ClinicalTrials.gov, so unless it announces the number itself, there’s no way to find out how many people are in and how they’re recovering. 

More from Tom's Guide

• Elon Musk says first Neuralink product lets you control your phone ‘just by thinking’
• Forget ChatGPT — Elon Musk's new xAI wants to uncover the 'true nature of the universe'
• I cover AI for a living and here's why I think Apple is the company to watch in this space

The brain-computer interface startup Neuralink has successfully implanted a device in its first human patient, tech billionaire founder Elon Musk said this week, paving the way for a product he claims would let people control almost any external device “just by thinking.” 

Neuralink's first product, which will be called Telepathy, as Musk revealed today on X, will enable users to control a smartphone, computer or other device via their brain. 

"Imagine if Stephen Hawking could communicate faster than a speed typist or auctioneer. That is the goal," he said. He added that the next set of trial participants will be people who have lost the use of their limbs.

Earlier this week, Musk posted on X that the implant's first human patient, whom he did not identify, had “received an implant” Sunday and “is recovering well" after the surgery. The initial results showed “promising neuron spike detection," he said. Spikes are a kind of catch-all term for any activity by neurons, special cells that use electrical and chemical signals to relay information in the brain and body. 

Neither Neuralink nor Musk have provided any further details about the operation or patient. But from the sound of it, it seems like the implant is successfully detecting neural signals from the patient's brain. There's still a lot we don't know about this sci-fi-sounding tech, but Neuralink has said the implant's "ultra-fine" threads help transmit signals in participants' brains.

Of course, Musk's comments should always be taken with a sizable pinch of salt. But a successful human trial does tease some of the exciting avenues of what the technology could accomplish. Beyond communication, Musk and Neuralink aim to use the implant technology to try to restore sight and help patients with spinal injuries walk again. In the past, Musk has name-checked Alzheimer’s and Parkinson’s as two neurological disorders that could benefit from Neuralink’s work.

In September, Neuralink announced it received an investigational device exemption (IDE) award from the FDA to seek human trial participants for its Precise Robotically Implanted Brain-Computer Interface (PRIME) clinical trials. As part of the study, a Neuralink surgical robot removes a section of the patient's skull to allow for the placement of the implant on a region of the brain that deals with movement intention. This implant would then record and transmit data to an app that interprets the person's brain activity into input directions to use an external device like a keyboard or cursor with their thoughts alone. With this trial, Neuralink's wireless brain-computer interface is being evaluated for the safety of both the implant and surgical robot.

Before embarking on human trials, the company surgically implanted its brain-computer interface in the skulls of pigs and monkeys. A video demonstrating the tech went viral in 2021 when Neuralink researchers revealed they successfully taught a monkey to play the classic arcade video game Pong using a brain implant. 

More from Tom's Guide

• Galaxy S24 Ultra owners complaining about washed-out screen colors
• Zoom for Apple Vision Pro will let callers see your digital Persona
• Android Auto will use AI to summarize long messages — why that’s a bad idea

Elon Musk’s artificial intelligence lab xAI unveiled his first AI chatbot over the weekend. Named Grok, it has been designed to compete with OpenAI’s ChatGPT or Google’s Bard but with more sarcasm and live access to posts and content from X.

The xAI lab was launched in July after Musk railed against the way the other large language models had been censored and had limits placed on their speech. Demonstrating Grok on Saturday, he shared a screenshot of it sarcastically responding to a query about cocaine.

Initially, it will only be available to a small group of invited testers in the U.S. but Musk says it will eventually be available as a perk for any X Premium+ subscriber. 

What inspired Grok?

In a post promoting the chatbot xAI claims it was modeled after Douglas Adam’s fictional Hitchhiker’s Guide to the Galaxy and “intended to answer almost anything and, far harder, even suggest what questions to ask”.

Grok is a term that means to understand a topic or idea intuitively to “get it” without needing further elaboration. It was first coined by Robert A Heinlein in his 1961 science fiction classic Stranger in a Strange Land. 

Unlike ChatGPT, Anthropic’s Claude and Bard, Musk says Grok has been designed to “answer questions with a bit of wit and a rebellious streak," as you’d expect from the billionaire. This includes its designers encouraging it to answer questions ignored by other tools.

What can Grok do?

Promoting Grok, Musk showed a screenshot of it being asked how to make cocaine. Where something like ChatGPT might give a warning about this question, Grok answered humorously suggesting the need to get a chemistry degree and license. 

Grok has access to a live data feed, pulling in comments, posts, links and media from the X platform itself. It isn’t clear how much of X’s content was used in training the model, but it likely formed a large part of the dataset.

It was trained over the past two months using Oracle hardware. This is a relatively short training period for a high-profile AI model with GPT-4, the model behind ChatGPT taking months or years of training, fine-tuning and re-training before release.

Grok: What are the risks of X access?

According to xAI it will be enhanced with new features added regularly. The lab suggests it can already outperform other models in its class such as GPT-3.5, the model that powers the free version of ChatGPT.

Having live access to a major data source like X is both positive and negative. While it will give the AI new data to train on and query regularly, often written in a more human, natural and conversational style than Wikipedia or other traditional publishing platforms, this can also be a negative. For every detailed X post on the latest scientific breakthrough, there are many more rants, aggressive posts and racism.

In the earlier days of generative AI this may have been fine, but with President Joe Biden’s White House, the EU and others looking to introduce restrictions and guardrails for AI a model that is quick to racism, or less cautious in its output may find itself shut down or xAI face fines.

More from Tom's Guide

• Elon Musk and other leaders call on ChatGPT and Bard to halt AI training — here’s why 
• Forget ChatGPT — Elon Musk's new xAI wants to uncover the 'true nature of the universe' 
• Elon Musk planning to launch ChatGPT AI rival 

Elon Musk says the new large language model from his artificial intelligence lab xAI will be available to a select group from tomorrow. In a cryptic post on X, he claimed that in some respects it is "the best [model] that currently exists".

The world’s richest man launched xAI in July, bringing together a team of leading engineers and AI scientists to "understand the true nature of the universe" through AI. This came off the back of the growing success of OpenAI through its chatbot ChatGPT.

Musk was an early investor in OpenAI but left the board in 2018 after an apparent disagreement and a failed attempt to take over the company. He was also heavily investing in machine learning and AI through the self-driving vehicle division of Tesla at the time.

Why launch his own AI company?

The billionaire has been critical of many AI products, services and research efforts from big tech companies, claiming too much emphasis is placed on guardrails and controls. Earlier this year he said he wanted to launch a "maximum truth-seeking AI".

In contrast to these lofty ambitions, Musk has also warned that AI poses an existential threat to humanity and that in the future it will put an end to work and create a world where "no job is needed". It isn’t clear how much of that comes from his xAI research or is just opinion.

As well as Musk himself, xAI includes engineers who have previously worked for Google's DeepMind, Microsoft, OpenAI and Tesla. They have collectively worked on some of the most important technologies that power modern AI tools like ChatGPT and Bard.

What will the model be like? 

The new xAI model is expected to be in direct competition with OpenAI's GPT-4, Google's PaLM 2 and Meta's Llama 2 with much of its training happening using Oracle's cloud network.

Musk has been in the U.K. this week at the AI Safety Summit. During a live conversation on X with British Prime Minister Rishi Sunak he declared AI has the potential to become the “most disruptive force in history.”

He said: “If you wish for a magic genie, that gives you any wish you want, and there’s no limit. You don’t have those three wish limits nonsense, it’s both good and bad. One of the challenges in the future will be how do we find meaning in life.”

During its initial announcement in July, Musk opined that for a "truth-seeking super-intelligence", it was safer to build it with curiosity and an interest in humanity as it will be more interested in keeping humanity safe.

That is part of what he hopes to achieve with xAI, a way to understand the universe and create a “good AI”. What format it will take, how it will be accessed by those given permission to use it or what the cost of the model will be have yet to be revealed.

More from Tom's Guide

• Elon Musk planning to launch ChatGPT AI rival 
• Elon Musk and other leaders call on ChatGPT and Bard to halt AI training — here’s why
• Elon Musk demos the human-like Optimus Tesla Bot — and it walks on its own  

After 24 hours marked by remarkable growth, Meta’s Threads is facing a legal challenge from the platform it’s seemingly aiming to replace.

As first revealed by Semafor, Twitter lawyer Alex Spiro sent a letter to Meta CEO Mark Zuckerberg revealing the company’s “serious concerns” that Meta “has engaged in systematic, willful, and unlawful misappropriation of Twitter’s trade secrets and other intellectual property.”

As evidence for this, the letter claims Meta hired “dozens of former Twitter employees” who “had and continue to have access to Twitter’s trade secrets and other highly confidential information.” The note alleges that Meta “deliberately assigned” said employees to develop “Meta’s copycat ‘Threads’ app.”

The letter says that Twitter could seek “both civil remedies and injunctive relief” if the matter is not resolved.

On Twitter, Elon Musk responded to someone sharing a link to the Semafor piece by stating that “competition is fine, cheating is not.”

On Threads, meanwhile, Meta spokesperson Andy Stone wrote that “No one on the Threads engineering team is a former Twitter employee — that’s just not a thing.”

Existential threat

When Mark Zuckerberg last gave a status update on signups 18 hours ago, Threads had reached 30 million users. For comparison, Twitter is estimated to have about 330 million. 

That’s a significant lead, but it’s been around for 17 years — and at the time of Zuckerberg’s last status report, Threads had been open for under 17 hours. 

It’s not inconceivable, therefore, that Threads could have more active userbase than Twitter within the month given its direct link to Instagram with its billion-plus users.

A lawsuit could put a spanner in the works. Only Meta and Twitter know the ins-and-outs of the claims and counterclaims, though employees moving between Silicon Valley giants isn’t especially unusual. And it’s perhaps even more expected in this case, given Twitter forced around half of its employees to look for new jobs within days of Musk taking over.

The similarities between the two apps is something that’s widely acknowledged, however — and seemingly even by Zuckerberg himself. On Threads launch day, the Meta founder tweeted for the first time in 13 years with the well-known ‘duplicate Spider-Mans pointing at each other’ meme. 

Indeed, ahead of launch, it was reported that Meta chief product officer Chris Cox told employees that what would eventually become Threads was “our response to Twitter”.

“We’ve been hearing from creators and public figures who are interested in having a platform that is sanely run, that they believe that they can trust and rely upon for distribution,” he reportedly added.  

Not jumped in yet? You can find out what Threads is like from our liveblog of the first 24 hours.

More from Tom's Guide

• Threads vs. Twitter: 9 biggest differences you need to know
• How to download Threads and sign up for your account
• You can’t delete your Threads account if you want to keep Instagram

The next contender in Silicon Valley's AI gold rush could be a surprising one. Twitter and Tesla CEO Elon Musk is reportedly aiming to launch an artificial intelligence start-up to compete with OpenAI and its revolutionary AI chatbot ChatGPT. 

According to a Financial Times report, Musk is in the process of bringing together a team of AI researchers and engineers to challenge OpenAI head-on. The report cites Nevada business records showing that Musk incorporated a company called X.AI on March 9, for which he's listed as the sole director. 

The billionaire has reportedly already initiated talks with SpaceX and Tesla investors about the project.

“A bunch of people are investing in it… it’s real and they are excited about it,” a person with direct knowledge of the matter told the Financial Times.  

To power his new AI venture, Musk has reportedly secured thousands of high-powered GPU processors from Nvidia and recruited engineers from AI labs, such as DeepMind. Nvidia is a leading manufacturer of graphics cards, like the upcoming GeForce RTX 40 series.

It's not clear whether there's a timeline for Musk's reported AI venture or what that may look like. But if all this shakes out, it's sure to raise some eyebrows. 

Musk co-founded OpenAI in 2015 but left the board three years later due to clashes with management over AI safety, among other issues. Just a few weeks ago, Musk co-signed a letter along with 1,800 others, including engineers from Amazon, DeepMind, Google and Meta, demanding a six-month pause on systems "more powerful" than ChatGPT-4 to implement tougher safety protocols for advanced AI design. 

GPT-4 is the most recent version of OpenAI's language model behind the wildly popular ChatGPT; it can handle not only text inputs but images as well, among other upgrades. That poses some really promising potential use cases for search, given that providing results in multiple formats, including text, images, video and more, can make such a massive difference in search results. Microsoft has confirmed it's been secretly using GPT-4 to power its new AI-supercharged Bing with ChatGPT.

More from Tom's Guide

• 7 Bard announcements Google should make at I/O to beat ChatGPT
• ChatGPT is coming directly to Windows 11 — no browser required
• 9 helpful things Bing with ChatGPT can do for you

With Elon Musk’s breakneck-speed transformation of Twitter welcoming back accounts previously deemed dangerous enough to permanently ban, there is an open question as to how Apple and Google will respond. 

Parler and Donald Trump’s Truth Social have previously been blocked by the two dominant app marketplaces for prioritizing free speech over content moderation. And while both have now been readmitted after making concessions, it’s possible that Twitter’s path could see it being blocked by Google Play and the App Store, too.

So what would Musk do if that happened? While other free-speech-focused apps have taken a more hands-on approach to moderation in response, Musk claims he would take a less conciliatory approach.

“I certainly hope it does not come to that, but, yes, if there is no other choice, I will make an alternative phone,” he tweeted.

Elon Musk phone a serious threat?

How seriously should we take that? Both seriously and not at all.

Let’s start with the serious end. Musk remains the richest person on the planet, and he’s clearly a man who’s not above making impulsive decisions as buying Twitter while foregoing his right to due diligence shows. 

Essentially, nothing is stopping him from entering the smartphone world if he wants to, but despite the many eager responses and positive polls posted by Musk superfans in response, it would be a massive uphill struggle to get even a fraction of Apple and Google’s mobile OS market share. 

Between them, the two have the market entirely sewn up, having successfully elbowed out companies that aren’t exactly short of money such as BlackBerry and Microsoft. Even Amazon — with the massive advantage of having the world’s most popular e-commerce storefront for free promotion — crashed and burned when it attempted to enter the smartphone world, and the less said about the HTC and Facebook collaboration, the better.

Despite accepting Twitter polls as sufficient evidence to decide consequential Twitter policy, Musk will know that on-platform surveys spread amongst his biggest fans don’t exactly meet the standards of polling methodology. Voting ‘yes’ in a poll is not the same as getting your wallet out and pre-ordering a hypothetical (and likely expensive) product. 

Even if it were, making a smartphone isn’t exactly something that happens overnight. And with Musk privately not ruling out the possibility of Twitter going bankrupt, time wouldn’t be on his side were Apple and Google to pull the plug.

It’s also worth highlighting that Musk says a lot of things — especially on Twitter — and not all of it should be taken at face value. Most recently, he promised that no decisions about unbanning Twitter accounts would be made until a moderation council had formed — a pledge that lasted all of 22 days.

With that in mind, the phrasing of the tweet is important: “if there is no other choice” is quite a telling phrase. There are loads of possible choices, including tougher content moderation — the option that both Truth Social and Parler picked.

In short, yes, Musk could build his own phone, or one branded with Tesla or Starlink. But it’s hard to imagine executives at Apple or Google losing any sleep over the possibility — and even less chance of it having any impact on the decision to ban Twitter, should it be discussed. 

If that was the intention of the tweet, then it feels more than a little misguided.

Elon Musk’s satellite internet service Starlink is now available in even more countries and regions, but both download and upload speeds have taken a hit as a result.

According to a new report from Ookla, which is known for having one of the best speed testing apps, Starlink’s median download speeds fell across the U.S., the UK, Canada, France, Germany and New Zealand during the second quarter of 2022. 

While the satellite internet provider has continued to expand, an influx of new users has led to speeds dropping from nine to 54 percent from Q2 2021 to Q2 2022. Still, Starlink reached a median download speed of at least 60 Mbps in North America during the second quarter of this year, which is more than enough for at least one connected device to access the best streaming services, play games online and video chat with friends and family.

Starlink upload speeds have also seen a decrease in all of the countries Ookla has tracked over the past year. Latency was a bit better and remained relatively flat in most countries with New Zealand being the outlier with a 23 ms drop in latency.

T-Mobile partnership

Ahead of the launch of Apple’s iPhone 14 which features Emergency SOS via satellite, T-Mobile and Starlink announced back in August that the two companies would work together to create a new network.

This new network will broadcast from Starlink’s satellites using T-Mobile’s mid-band spectrum across the U.S., with the end goal of providing nearly complete coverage nationwide “almost anywhere a customer can see the sky” according to a press release.

As a first step in its new partnership with Starlink, T-Mobile plans to extend text message coverage practically everywhere in the continental U.S., Hawaii, parts of Alaska, Puerto Rico and territorial waters even outside of its network.

Starlink is also working to bring its satellite internet service to airplanes, cruise ships and cars as part of its continued expansion.

Other satellite providers are also expanding

Starlink isn’t the only satellite internet provider, and others including Viasat, OneWeb and even Amazon are also working on expanding their respective services.

For instance, Viasat’s new ViaSat-3 satellite is now in the mechanical build process in which solar arrays, reflectors and other items will be attached to the satellite ahead of its launch. The company’s first two new satellites are planned to focus on North and South America as well as Europe, the Middle East and Africa. However, the third ViaSat-3 satellite will be deployed to provide coverage in the Asia Pacific region.

Meanwhile, satellite companies OneWeb and Eutelsat have announced a provisional agreement to combine their businesses and are planning to develop a platform that will offer internet connectivity from low Earth orbit and geostationary satellites.

Not to be outdone by Elon Musk, Amazon is deploying 3,236 satellites in low Earth orbit as part of a program called Project Kuiper. Jeff Bezos’ Blue Origin is one of three companies, along with Arianespace and ULA, that is helping launch satellites for Amazon.

If you live in a rural area or even in an RV it’s never been easier to get online, and increased competition in the satellite internet market will only help improve these services while hopefully bringing down their monthly costs.

Trying to sign up for satellite internet service from Starlink typically means adding yourself to a long wait list. But a new service from Starlink offers a potential workaround where you can get the necessary equipment with absolutely no wait time. 

We are talking about Starlink RV, launched this week by the company as an internet service for RV owners or people who like camping or road trips. Starlink RV includes a portable dish that promises to provide internet connectivity anywhere. 

Best of all, if you subscribe to Starlink RV, the company will ship a Starlink dish to you right away. In comparison, buying a regular Starlink dish right now will put you on a waitlist until 2023.

Before you rush out to get Starlink RV, though, note that it's a little bit different than the standard Starlink service. Starlink RV is pretty expensive, costing $135 for the internet service, or $25 more than a regular Starlink connection. The dish itself will cost a massive $599. 

You can also pay for your internet service as you go, giving users the flexibility to pause and unpause at any time. But the biggest catch with Starlink RV is that service will always be deprioritized in favor of standard Starlink customers. Plus, you can’t rely only on Starlink RV for your internet going into the wilderness because you’ll only receive Starlink broadband in areas where SpaceX provides active internet coverage.

So if you live in an area, or are traveling to an area marked as "coming soon" on Starlink's coverage map, signing up for Starlink RV right now won't help. Starlink will work in areas marked as "waitlist" though. 

Starlink's support page mentions “Network resources are always de-prioritized for Starlink for RVs users compared to other Starlink services, resulting in degraded service and slower speeds in congested areas and during peak hours.” Download speeds through Starlink RV range from 5 Mbps to 100 Mbps in high usage waitlist zones, which is a massive downgrade from the normal 50 Mbps to 250 Mbps Starlink download speeds.  

Elon Musk, CEO of SpaceX which administers Starlink, tweeted about the launch news, adding that the antenna of the Starlink RV is too big for cars. It's also worth noting that Starlink RV doesn't officially operate "in motion," which means it can't be accessed on-the-go. 

Starlink RV arrives a few weeks after SpaceX introduced a new portability option to the standard Starlink service — allowing users to use their dish in different locations outside their address.

But with a higher cost plan, downgraded internet service and unreliable network coverage, Starlink RV doesn't make a very strong case for itself. It leaves us wondering if waiting for standard Starlink service would be more worthwhile at this point.

It's time to watch the Elon Musk documentary Elon Musk's Crash Course online to see the other big controversy in the tech magnate's life. Because while the headlines are filled with allegations that SpaceX paid $250,000 to a flight attendant who accused Elon Musk of sexual assault, The New York Times is taking a magnifying lens to Musk's other business: Tesla. Oh, and in other breaking news? Musk is looking for lawyers ... on Twitter. 

This special is an investigative look at Musk and Tesla's work developing self-driving technology. The film, based on first-hand accounts was just announced to debut — and arrives tonight.

The film, according to a press release, paints its picture with first hand accounts of how Tesla's Autopilot technology is connected to "several deaths and dozens of other accidents that Tesla has not publicly acknowledged." Several former Tesla employees also share their stories in the film.

FX and Hulu release Elon Musk's Crash Course at an interesting moment for the Tesla CEO. All signs suggest Musk has tired of trying to buy Twitter, at least at the negotiated price, as he said the deal "cannot move forward" as it currently stands. He's blamed this on the amount of spam on the platform, and the number of fake Twitter accounts.

Elon Musk's Crash Course is produced and directed by Emma Schwartz, known for her work on Frontline documentaries. It also features contributions from New York Times reporters Cade Metz and Neal Boudette.

Here's everything you need to know to watch the Elon Musk documentary Elon Musk's Crash Course online (which is the successor to The New York Times Presents Controlling Britney Spears). Plus, check out the trailer below:

How to watch Elon Musk's Crash Course online from anywhere on Earth

Just because FX isn't available everywhere doesn't mean you need to miss Elon Musk's Crash Course if you're away from home. Watching along with the rest of the internet can be pretty easy, because with the right VPN (virtual private network), you can stream the show from wherever you go.

Not sure which VPN is right for you? We've tested many different services and our pick for the best VPN overall is ExpressVPN. It offers superb speeds and excellent customer service.

How to watch Elon Musk's Crash Course in the US

In the U.S., viewers can watch Elon Musk's Crash Course on Friday (May 20) at 10 p.m. ET on FX, which comes with a cable package. 

If you've already cut the cord and don't have cable, you can access FX on multiple live TV streaming services including Sling TV, Fubo TV, YouTube TV, Hulu With Live TV and DirecTV Stream.

We recommend Sling and Fubo, two of the best streaming services. Sling is the more affordable option, but Fubo offers a broader channel lineup.

Can you watch Elon Musk's Crash Course online in the UK?

Hulu isn't available in the U.K., so we're not sure when Elon Musk's Crash Course will arrive across the Atlantic. It could appear on Disney Plus' Star Channel, which is the U.K.'s equivalent of Hulu, but no release date was given yet.

If you want to watch with other services you've already paid for while traveling abroad, ExpressVPN will help you get access to them.

How to watch Elon Musk's Crash Course in Canada

Fortunately, our friends in Canada have a good situation with watching Elon Musk's Crash Course  (better than those in the U.K., at least). The VOD Crave platform will receive new episodes on the same date and time as episodes debut in the U.S..

Crave has a 7-day free trial. After that, the Movies + HBO package costs $19.98 CAD per month. 

Travelers who have already subscribed to other services can get access to them with the help of ExpressVPN.

Can you watch Elon Musk's Crash Course in Australia?

At the moment, it doesn't look like Elon Musk's Crash Course has a home down under. 

Americans having trouble finding it abroad when it airs on Saturday at 12 p.m. AEST can do so by using a VPN such as ExpressVPN.

On February 24, 2022, Russian launched a full-scale invasion of Ukraine, following military build-up on the nation's border. 

Dubbed a "special military operating" by President Vladimir Putin, and slammed by the international community with blanket condemnation, the invasion of Ukraine has had a devastating effect on both the country and its people as well as wider ramifications on a socio-economic basis. 

As such, a whole host of organizations have mobilized to help out Ukrainians who have been displaced or affected by the ongoing war, which has seen street-to-street fighting.  

If you wish to help out the Ukrainian people, then there are numerous way you can lend aid.

You can donate to various charitable organizations that work in warzone and conflict areas. These include UNICEF, which works on providing health and nutritional care, as well as access to clean water; the latter has reportedly been disrupted due to heavy weapons fire hitting critical water infrastructure.

Médecins Sans Frontieres, aka Doctors Without Borders, has several teams in Ukraine, where it aims to provide medical care in the warzone.

Voices of Children is a Ukrainian organization that provides psychological and psychosocial support to children affected by the armed conflict that's been taking place in the nation and surrounding areas since 2014.

Charity site Care has a Ukraine emergency fund for the nation, and aims to provide food, water and other emergency aid.

And Washington D.C. chef José Andrés has his World Central Kitchen team in Europe, where it’s providing “thousands of meals in Poland, Romania and even inside Ukraine,” and you can donate to the organization. 

The International Committee of The Red Cross is also working with the Red Cross in Ukraine and you can donate to the medical aid organization as well. 

In the tech world, companies are offering enticements to donate to Ukraine relief efforts. For instance, donate at least $40 (with 100% of that money going to relief programs), and $40, Humble will give you 123 games, ebooks and other pieces of software worth $2,572. That bundle is available through March 25.

Those are to name but a few, as there are also a myriad of other charities and emergency aid organizations that have sprung into action to help Ukrainians who’ve had their lives disrupted by the Russian invasion. 

Just make sure you're not taken in by one of the many Ukrainian charity scams that have sprung up in the wake of the crisis.

Elon Musk steps up with Starlink internet 

Tech luminary and active Twitter user Elon Musk has also jumped to the aid of Ukraine. 

In response to internet disruption in Ukraine caused by the Russian invasion, Musk has activated the Starlink satellite internet service in the country. 

The move comes in response to Mykhailo Fedorov, the vice prime minister of Ukraine and minister of digital transformation, tweeting at Musk about the situation in the invaded country in comparison to Musk’s space ambitions. 

"While you try to colonize Mars — Russia try [sic] to occupy Ukraine! While your rockets successfully land from space — Russian rockets attack Ukrainian civil people! We ask you to provide Ukraine with Starlink stations and to address sane Russians to stand," tweeted Fedorov. We are guessing the “stand” bit is Fedorov wanting the Russian public to condemn the actions of the country’s government. 

In response, Musk succinctly replied: “Starlink service is now active in Ukraine. More terminals en route.” 

Other tech-related firms have been affected by the war, with Netflix looking set to refuse to host propaganda channels on its platform. And Goggle is pushing out air raid alerts to Ukrainian Android users. 

Internet connectivity is arguably more critical to the flow of information and communion in a war than it even is in peacetime. Aggressors will likely seek to cut connectivity in order to prevent the spread of enemy information and to somewhat control the narrative of their invasion, while defenders will want to use the internet to communicate what's going on in their nation to the wider world.

So Musk stepping in to help restore connectivity in Ukraine will hopefully allow the invaded country to keep its communications up and running and prevent Russia from controlling the perception of the war.