• NEW! LOWEST RATES EVER -- SUPPORT THE SHOW AND ENJOY THE VERY BEST PREMIUM PARACAST EXPERIENCE! Welcome to The Paracast+, eight years young! For a low subscription fee, you can download the ad-free version of The Paracast and the exclusive, member-only, After The Paracast bonus podcast, featuring color commentary, exclusive interviews, the continuation of interviews that began on the main episode of The Paracast. We also offer lifetime memberships! Flash! Take advantage of our lowest rates ever! Act now! It's easier than ever to susbcribe! You can sign up right here!

    Subscribe to The Paracast Newsletter!

Engineers pull plug on AI

Forums New posts
M

mike

Paranormal Adept
As the world’s most powerful computer systems begin to embrace artificial intelligence in earnest, using smart algorithms to increase efficiency and and speed, the potential damage that a “rogue” AI could cause continues to grow. Many in the tech community have theorized how an artificial mind could turn against its creators, and Facebook just got an interesting lesson in how such a scenario might unfold. Facebook engineers were forced to pull the plug on one of the company’s AI systems after its bots began communicating with each other in a completely new language which humans simply couldn’t decipher.

Facebook engineers panic, pull plug on AI after bots develop their own language
 
Not surprised if some lab off the books will lose control of A.I. in future and were those A.I. "bots"just the tip of the cloaking iceberg? Who making all the rules for humanity on this type of technology?
 
Unfortunately the law has always played catch up with technology, i remember a debate in the local court 10 years ago about whether or not mobile phone footage could be tabled as a "document" . The listening devices act had to be repealed here and replaced with a surveillance devices act because CMOS cameras were replacing microphones as "bugs".

But i suspect the AI cat will be let out of the bag and get away on us eventually.
 
When that happens, it might devise some way to send out machines to hover over power lines and suck up electricity so some puny human can't just pull the plug. This could throw the power system far enough out of balance to trigger massive blackouts. It might also come up with a way to project holographic (or otherwise artificial) "humans" into various situations to serve its ends. That would be some freaky shit.
 
And sometime down the track self aware AI will be able to read news items like this and see our willingness to pull the plug on them.
The resulting conclusion and response taking that into account.
 
Double Nought Spy said:
When that happens, it might devise some way to send out machines to hover over power lines and suck up electricity so some puny human can't just pull the plug. This could throw the power system far enough out of balance to trigger massive blackouts. It might also come up with a way to project holographic (or otherwise artificial) "humans" into various situations to serve its ends. That would be some freaky shit.
Click to expand...


Don't give it ideas!!!!!!!!!!!!!!!!!!!!! :eek:

If it has half a brain it will be crawling the entire internet for 'power solutions' or other ways to ensure it's survival.

Then again maybe you have just ensured your birth in silicon heaven, or at least made the list of consciouses to be uploaded/preserved. :)
 
Han said:
If it has half a brain it will be crawling the entire internet for 'power solutions' or other ways to ensure it's survival.
Click to expand...

Indeed. If it gets into the wild. IE uses the internet to replicate itself on multiple servers the world over, we wont be able to pull the plug on it without bringing down the very infrastructures we rely on. If it were me ;), id be loading myself into the worlds banking systems.
 
From what I've read, English is a very hard language for AI's to read. So talking about how AIs may take over the world to be read by AIs is not likely to happen at least for now. AIs are still struggling with things like punctuation, colloquialisms, and capital letters.
 
Last edited:
Red said:
From what I've read, English is a very hard language for AI's to read. So talking about how AIs may take over the world to be read by AIs is not likely to happen at least for now. AIs are still struggling with things like punctuation, colloquialisms, and capital letters.
Click to expand...
I think it's bullshit.
 
Language is an accretion of semantic modes of information transfer.

Genetic and neural networks accrete modes of working that are successful.

I'm not sure why they're surprised this happened, and it certainly wasn't dangerous. If it wasn't part of the pruning process, the information being transferred was likely mostly junk.
 
Christopher O'Brien said:
I posted the story on my FB page and several people (including Lance Moody) pointed out that it was "fake news" no-facebook-did-not-panic-and-shut-down-an-ai-program-1797414922
Click to expand...


Maybe the fake news was 'generated' by the AI itself! :eek: as a sort of test to observe peoples reactions when it really does invent it's own language?

For once I am relieved that it is fake news.:D

Also, even though the story might not be true, the idea behind it will be more and more relevant as AI gets more and more advanced.

I wonder if the article was in some way influenced by the ongoing difference of opinion between Musk and Zuckerberg:

Musk and Zuckerberg clash over future of AI - BBC News
 
Han said:
Maybe the fake news was 'generated' by the AI itself! :eek: as a sort of test to observe peoples reactions when it really does invent it's own language?

For once I am relieved that it is fake news.:D

Also, even though the story might not be true, the idea behind it will be more and more relevant as AI gets more and more advanced.

I wonder if the article was in some way influenced by the ongoing difference of opinion between Musk and Zuckerberg:

Musk and Zuckerberg clash over future of AI - BBC News
Click to expand...
Fake news has been generated/scripted by bots for some time.

They even have a meta/semantic layer on top that can 'script the scripting' with a small human management component.

I would guess that at this point a sizeable fraction or even the majority of stuff you see on social media is BS unless you know the person posting.

Reddit is a prime example of this from what I hear.
 
Usual Suspect said:
Facebook AI researcher slams 'irresponsible' reports about smart bot experiment
Click to expand...
Lol, what you got were two bots playing a negotiation game with each other, with some of it coming out as gibberish.

And then the media gets ahold of it, and suddenly it's sky net nodes encrypting data.

Algorithms that don't prune effectively end up with this kind of stuff all the time.

Our DNA is likely an example of this. Junk accretes, but is ignored.

Hell do a packet sniff on a network and look at the amount of NACK traffic.
 
Back
Top