Download the
Tom's Guide App from the AppsStore
News and trends on internet
/ mobile / "sound & picture" / IT
Yes No

Scientists Find Networks Can Get Schizophrenic

- By - Source : University of Texas at Austin

Computer networks may have more human characteristics than we know. Researchers at the university of Texas found that computer networks can show a kind of virtual schizophrenia, if they can't forget fast enough.

The project was based on a virtual computer model designed to simulate the excessive release of dopamine in the human brain. The network was able to learn a natural language was used to investigate what happens to language as the result of eight different types of neurological dysfunction. The results of the simulations were compared by Ralph Hoffman, professor of psychiatry at the Yale School of Medicine, to what he saw when studying human schizophrenics.

The information provided to the network very much in the same way a human brain stores information-not as distinct units, but as statistical relationships of words, sentences, scripts and stories, the researchers said. “With neural networks, you basically train them by showing them examples, over and over and over again,” said Uli Grasemann, one of the project leaders. “Every time you show it an example, you say, if this is the input, then this should be your output, and if this is the input, then that should be your output. You do it again and again thousands of times, and every time it adjusts a little bit more towards doing what you want. In the end, if you do it enough, the network has learned.”

Next, the scientists reran the data input process, but simulated an excessive release of dopamine by increasing the system’s learning rate-essentially telling it to stop "forgetting" so much.  “It’s an important mechanism to be able to ignore things,” said Grasemann. “What we found is that if you crank up the learning rate in [the virtual computer model] high enough, it produces language abnormalities that suggest schizophrenia.” When retrained with the faster learning request, the network "began putting itself at the center of fantastical, delusional stories that incorporated elements from other stories it had been told to recall," the researchers said. In one example, the system "claimed responsibility for a terrorist bombing."

The scientists concluded that there are parallels between virtual neural networks and the human brain. The hop is that such computer systems can aid in clinical research in the future.

Share:
20
Comments
X

Comments

joytech22 05/06/2011 11:03 PM
Hide
-4+

That's interesting but creepy at the same time?

Fokissed 05/06/2011 11:04 PM
Hide
-1+

"The Ihop is that such computer systems can aid in clinical research in the future."
Fixed.

nebun 05/06/2011 11:17 PM
Hide
-0+

amazing...we are getting close and closer to cyber machines

Anonymous 05/06/2011 11:25 PM
Hide
-1+

So, it shouldn't be "Hope" ?

Zingam 05/06/2011 11:32 PM
Hide
--1+

poorly written article

dalethepcman 05/06/2011 11:43 PM
Hide
-1+

This is horrible news. We have just proven that once someone actually manages to program a self learning AI that at some point of learning everything about everything its going to go insane. Hello skynet.

fir_ser 05/07/2011 12:07 PM
Hide
-2+

So this system claimed responsibility for a terrorist bombing, what about the world financial crisis and global warming is this system responsible for these too?

terastac 05/07/2011 8:15 AM
Hide
--2+

so if we teach a schizo brain and do this a different way like teach it new stuff everytime rather than repetative moron learning it will eventualy become einstein becuase it will learn massive amounts of information and eventually become sentient. Is it not "funny" that people at university cramming for exams become schizo and when we find out and every doctor I have heard say its always the smart ones that are really good at remembering things that end up this way that should be taught faster and maybe a new einstein emerges I know been in a repetative loop my self does lead to some wierd stuff if you cant work your way out of it

eddieroolz 05/07/2011 11:20 AM
Hide
--1+

Only matter of time until we discover they can be diagnosed with Alzheimer's too!

Anonymous 05/07/2011 3:51 PM
Hide
--1+

the "hop" lol, this is not a well written article.

The Greater Good 05/07/2011 10:40 PM
Hide
-0+

Do you ******* proof-read? By, "you" I mean the people at Tom's Hardware... and by "*******" I mean men that like other men in a sexual nature.

shanky887614 05/07/2011 11:57 PM
Hide
--1+

terminator anyone?

lathe26 05/08/2011 4:10 AM
Hide
-4+

Computer networks != neural networks

The scientists were not talking about networks like ethernet or Wifi. Neural networks are a computer science topic, similar to how data structures and algoritms are computer science topic.

sipmarton 05/08/2011 9:04 AM
Hide
-2+

A terrible choice of title, I wonder if the author knows anything about neural networks. If so, why talk about them as computer networks, when the two have very little in common, like lathe26 pointed out.

bit_user 05/09/2011 12:34 PM
Hide
-0+

lathe26 and sipmarton are correct. This is referring to Neural Networks, as stated in a quote buried in article. A neural network is a machine learning technique that's loosely patterned after the way neurons collaborate to process information and learn.

From what it says, I think all they did was build a neural net to do text processing & recall. Then, they tinkered with some constants and observed effects analogous to what happens when equivalent constants are changed in a human brain. Nowhere does it say that the neural net was sophisticated enough to have any degree of comprehension of the text.

This work is interesting simply because it suggests that we may be able to use computer models to analyze certain psychological conditions. That's it. It really isn't telling us anything about machine intelligence. Nor did Watson, while I'm on the subject.

bit_user 05/09/2011 12:38 PM
Hide
-0+

P.S. you guys really need to do a little more homework, when you report on these kinds of things. The potential for confusion and misunderstanding is so huge that you're better off not covering them than doing it poorly.

cookoy 05/09/2011 4:21 PM
Hide
-0+

my brain hurts just reading this article. Don't ask if i understood it.

gm0n3y 05/09/2011 6:46 PM
Hide
-0+

Two comments, first of all this article is very badly written. Secondly, the title should note that these are NEURAL networks. These have little relation to regular computer networks.

drinking12many 05/09/2011 8:54 PM
Hide
-0+

Did they threaten it with waterboarding before it talked? :)

jrnyfan 05/10/2011 1:22 PM
Hide
-0+

What an awful article. In all my many years coming here, this is by far the worst written article. A monkey throwing poo against a keyboard could make a more coherent string of sentences.