FaceBook AI develops own language... shut down

[playzwtrux]

http://www.firstpost.com/india/face...anguage-unintelligible-to-humans-3876197.html

Just a snippet:

"Days after Tesla CEO Elon Musk said that artificial intelligence (AI) was the biggest risk, Facebook has shut down one of its AI systems after chatbots started speaking in their own language, which used English words but could not be understood by humans. According to a report in Tech Times on Sunday, the social media giant had to pull the plug on the AI system that its researchers were working on "because things got out of hand". The trouble was, while the bots were rewarded for negotiating with each other, they were not rewarded for negotiating in English, which led the bots to develop a language of their own.

"The AI did not start shutting down computers worldwide or something of the sort, but it stopped using English and started using a language that it created," the report noted. Initially the AI agents used English to converse with each other but they later created a new language that only AI systems could understand, thus, defying their purpose. This led Facebook researchers to shut down the AI systems and then force them to speak to each other only in English."



Not sure where this goes, but...

Is this the beginning of SkyNet?

 

[AlexDaGator]

That's not scary at all.

Alex.

 

[Captain Sasquatch]

Not entirely accurate.

https://www.forbes.com/sites/larrym...experiment-is-highly-exagerated/#2683ee06417e

 

[Zambo]

This kind of **** is frankly a road we don't need to go down. It's like inventing the nuclear bomb. Don't invent it.

 

[playzwtrux]

Not entirely accurate.

https://www.forbes.com/sites/larrym...experiment-is-highly-exagerated/#2683ee06417e

from the article...

"It is true that the bots created their own "language," but it wasn't to hide from their human overlords but simply to be more efficient."

This is how it all starts. Just ask Sara Connor

 

[Captain Sasquatch]

from the article...

"It is true that the bots created their own "language," but it wasn't to hide from their human overlords but simply to be more efficient."

This is how it all starts. Just ask Sara Connor

They didn't shut it down because the bots created their own language. In fact, they expected that to happen.

 

[Gatordiddy]

It was probably this:

 

[T REX]

from the article...

"It is true that the bots created their own "language," but it wasn't to hide from their human overlords but simply to be more efficient."

This is how it all starts. Just ask Sara Connor

More efficient is the problem. Runaway AI trying to become more efficient is how it all begins. It's a slippery slope for sure.

 

[TheDouglas78]

The problem is that the bots were doing what they were designed to do which was customer service. There is a lot more AI out there than most people realize. This particular AI was just following the natural evolution of speech, especially amount common people/friends which is make it more tailored towards the group you are with. Think about how you talk among your truly close friends, are their phrases and words that have different meanings to your close group of friends that would sound greek to those outside of your group.

 

[playzwtrux]

@Captain Sasquatch & @TheDouglas78, I'm not denying anything that you two are saying. Part of this is serious and part is simply sarcasm (2nd post in particular).

But in all Truthfulness, and REALITY... there should be a hint of concern. If the bots can develop their own language in the name of efficiency, who says it won't eventually start to think humans are simply inefficient and want to eliminate us. This is about 95% of all the scenarios envisioned relating to AI technology. I'm not really a conspiracy theorist, but the scenario seems within the realm of logic and both of you have somewhat proven this point.

I believe that they thought that it was possible that it would create its own language and did not shut it down because of that, but because for a brief moment, someone with some sense actually had the thought of "Hey, at some point this thing might start to become more self aware and start to seclude itself from us. We should shut it down while we can and not wait until it's too late and we can no longer control it."

Maybe your wording was wrong Dougie, but why would they shut it down for actually doing it's job, unless it was trying to handle customer service with humans in it's newly created language? If it was doing it's job and the secondary language was just a by product, why shut it down for that alone?

I believe that somewhere someone became concerned and that's why it was shut down.

 

[TheDouglas78]

Maybe your wording was wrong Dougie, but why would they shut it down for actually doing it's job, unless it was trying to handle customer service with humans in it's newly created language? If it was doing it's job and the secondary language was just a by product, why shut it down for that alone?

The problem was it's job was to deal with customers, and in the development of it's own language it made no sense to a person. You shut it down, much like bug review. You can't continue to have a bad piece of software, creating more of an issue until you can see what caused the evolution. It depends on the nature of the bug/issue in this case it would have been a "work stoppage" issue, creating more work to keep in place than remove.

 

[playzwtrux]

The problem was it's job was to deal with customers, and in the development of it's own language it made no sense to a person. You shut it down, much like bug review. You can't continue to have a bad piece of software, creating more of an issue until you can see what caused the evolution. It depends on the nature of the bug/issue in this case it would have been a "work stoppage" issue, creating more work to keep in place than remove.

so I nailed it...

 

[Lake Gator]

As long as we can pull a plug...we're safe from the terminators.