SEATTLE — If artificial intelligence hasn’t sounded all that promising over the past couple of weeks, you may be thinking of privacy issues. Like, say, Facebook talking about using AI to delete hateful and terrorist content. Or Google using it to pick out your face and voice. Those may be useful cases, but they're associated with fake news, privacy and data scandals. Instead, it would be nice to see AI put to use altruistically. So it was almost a breath of fresh air when Microsoft CEO Satya Nadella announced some of his company’s plans for artificial intelligence.
Sure, there are lots of developer tools, neural net processing, chatbots, data tools and a new, non-consumer version of Kinect that will help drones fly with Azure cloud services. But what was refreshing was when Nadella unveiled Microsoft’s AI for Accessibility program. It will use $25 million over five years to fund AI projects by developers looking to assist people with disabilities.
Microsoft claims that only one in 10 people with disabilities has access to assistive technologies. The program will provide seed grants to developers, non-governmental organizations, inventors and universities, identify candidates and include their AI innovations in its own services.
This isn’t Microsoft’s first foray into AI for social good. Late last year Microsoft president announced a similar initiative to help the environment.
“We need to ask ourselves not only what computers can do, but what computers should do,” Nadella said on stage.
Some of this, is, admittedly, great for publicity. Microsoft looks really good by virtue of throwing large bags of money at good causes. But it’s also putting a positive light on AI. This isn’t something designed to better target ads or replace workers. It’s just to help some of the people who could really use it the most.
Sure, Microsoft has done some of this already with tech like Seeing AI, which helps narrate the world for those with impaired vision. And others, like Google have put together a big collection of creative apps that help people understand what neural networks are capable of.
But what Nadella had going for him on stage at Build was timing. When Facebook is holding up AI as a defense against people taking advantage of its platform and Google’s AI approach is viewed, by some, as a potential invasion of privacy, funding these project makes Microsoft look like a charity in comparison.
Nadella also discussed the importance of fair, unbiased AI. Artificial intelligence, he said, is built with the bias of creators. Microsoft intends to use computing to “de-bias” AI to make it fair and ethical.
While he was at it, Nadella also took the chance to discuss privacy. He called privacy a “human right,” echoing Apple CEO Tim Cook, and called GDPR, European regulation which prevents certain tracking by ads, “a sound, good regulation.”
Microsoft did have some privacy stumbles early-on with Windows 10, so let’s hope the company learned from those.
Of course, Microsoft is still a company, and it will make money off of AI. AI a big part of the future of Microsoft, and a main pillar of the company after its most recent reorganization. But if Nadella and Microsoft plan on making AI a primary focus going forward, at least they’re making the whole enterprise look like a worthy endeavor. AI badly needs this makeover.