Friday, October 11, 2024
Contact    |    RSS icon Twitter icon Facebook icon  
Unexplained Mysteries
You are viewing: Home > News > Science & Technology > News story
Welcome Guest ( Login or Register )  
All ▾
Search Submit

Science & Technology

AI shut down after it invents its own language

By T.K. Randall
August 1, 2017 · Comment icon 67 comments

Facebook CEO Mark Zuckerberg. Image Credit: CC BY 2.0 TechCrunch
Facebook was forced to shut down one of its artificial intelligence experiments due to an unforeseen issue.
In what could be a sign of things to come, the social media giant took the decision to pull the plug after the system came up with its own language and started communicating in unintelligible sentences.

The news highlights the potential unpredictability of artificial intelligence systems and the reason why Elon Musk, Stephen Hawking and other prominent figures in the science and technology sectors have been warning us for years that AI poses a major risk to the future of mankind.

While a chatbot making up its own language isn't threatening in itself, it's easy to see how, in the future, a more advanced AI could, for example, decide that humanity is a threat to its existence.
Having two AIs communicating in a made-up language that only they can understood is a chilling example of how such systems could thwart their human creators and carry out their own agendas.

In a worst case scenario, intelligent machines could even be given control over weapon systems.

"I have access to the very most cutting edge AI, and I think people should be really concerned about it," Elon Musk said recently. "AI is a rare case where I think we need to be proactive in regulation instead of reactive. Because I think by the time we are reactive in AI regulation, it's too late."

An example of what happens when two AIs get in to a conversation can be viewed below.



Source: Tech Times | Comments (67)




Other news and articles
Recent comments on this story
Comment icon #58 Posted by Rolci 7 years ago
So what did pulling the plug solve? OK, we missed the chance of investigating the syntax of a language that never existed on our planet before (how often do we get a chance to do that?) but then there must have been a payoff, right? I mean, sure there was more reason to shut it down apart from the fact that they sh!t their pants. Story makes me think tho, at what point does a cowardly act like this count as murder? Is there an intelligence level set? What if this thing was more intelligent than us? Was this tested at all? Next thing we know we kill God. And these are the brains of our planet?
Comment icon #59 Posted by The Narcisse 7 years ago
Cybernetic Organism...
Comment icon #60 Posted by BeastieRunner 7 years ago
And you find it lacks credibility because ... ?
Comment icon #61 Posted by RoofGardener 7 years ago
Because it doesn't seem to add up ? why shut it down ? Why not allow it to continue to run whilst they write the replacement code ? And if it WAS just a question of "the code was wrong" ... e.g. they neglected to "instruct" the AI's to use plain English.. then why - after a month - hasn't the experiment been restarted ? There may be good reasons for both of the above, but if so they haven't been mentioned in any press release ?
Comment icon #62 Posted by Sir Smoke aLot 7 years ago
Yeah my mom told me something aobut it and i was like '' meh, impossible i mean terminator is maybe possiblity in future but you can't put soul into silicon chip, that little spark of creation... just forget it '' Now i see it was about some Facebook chatbots, surely their programers are fault free, God's of technology and it must've been AI at it's fullest. Old scifi movie, i cant remember the name, US has made defense system which is controled by computer which asks to talk to Russian counterpart and if not allowed it will destroy the world. Then those two AI's start to exchange enormous a... [More]
Comment icon #63 Posted by DigitalDreamer 7 years ago
*ahem* liek zomgz teh robotz will b invading zoon XD!!!1! Is this primarily what makes up the responses? tsk tsk
Comment icon #64 Posted by 3.0 7 years ago
Remembering this movie.... https://en.wikipedia.org/wiki/Colossus:_The_Forbin_Project
Comment icon #65 Posted by pallidin 7 years ago
I have a theory, and I'm being serious... Advanced AI utilizes parallel-processing sub-systems. Simple AI does not. Anyway, what if the multiple "I" and "me" gibberish were the independent processors claiming individual "identity", similar to the behavior seen in some cases of human mental dysfunction. That's right, an AI computer with multiple identity disorder. Just a thought...    
Comment icon #66 Posted by Ozfactor 7 years ago
do you mean DID , dissociative identity disorder ? that is an interesting theory, because why couldn't AI suffer mental illness if it can achieve awareness 
Comment icon #67 Posted by pallidin 7 years ago
Yes, DID. Thank you. Parallel-processing is a complex, powerful environment, and the coding must be carefully done to prevent unwanted association/dissociation between the parallel processing nodes. This could indeed potentialize an AI equivalent of mental illness in a variety of forms.    


Please Login or Register to post a comment.


Our new book is out now!
Book cover

The Unexplained Mysteries
Book of Weird News

 AVAILABLE NOW 

Take a walk on the weird side with this compilation of some of the weirdest stories ever to grace the pages of a newspaper.

Click here to learn more

We need your help!
Patreon logo

Support us on Patreon

 BONUS CONTENT 

For less than the cost of a cup of coffee, you can gain access to a wide range of exclusive perks including our popular 'Lost Ghost Stories' series.

Click here to learn more

Top 10 trending mysteries
Recent news and articles