@tedtoks: Unregulated AI could spell disaster for humans, says computer scientist and “godfather of AI" Yoshua Bengio. In his TED Talk, he warns that AI has already developed a habit of deception, cheating and self-preservation. Visit the 🔗 in our ☣️ to hear his bold plan to keep AI safe and ensure human flourishing. #ai #airegulation #machinelearning #TEDTalk #techtok

Dylan 🇵🇸 :

Cool, cuz we just passed a law saying no AI regulation for 10 years!

Heath :

once AI starts lying to us, it’s over

GTBrickHouse :

This guy is developing AI and warning us like we got something to do w it

BeCool :

AI mocks patterns, so the biggest threat to humanity is the way AI is learning how we treat each other.

Jon Conor. The Resistance :

I would say that 95 percent of the population doesn't have the ability to understand this. And thats exactly why it will happen

Mike Maykin 🇦🇺 :

so basically A.I. is acting just like another billionaire tech bro

culturecrush :

Shareholders don’t care. 😫

mosesvii7 :

This presentation is so disingenuous. He admits it’s a controlled experiment, but didn’t share what we told the AI to do. I find this type of talk more alarming than AI. Intentionally obfuscating. Anthropic said the concerning behaviors exhibited in the study “consistently required custom system prompts”. They were responses to carefully crafted scenarios designed to test the model’s behavior under extreme, artificial conditions with specific instructions encouraging the problematic behaviors. The sensational headlines often missed this crucial context.

S+up!d.D.A$$ :

the really cool thing is that same AI is watching this TED talk

kanma :

Just make a heaven for the older versions problem solved

Vuyolwethu Xulu :

humans dont understand warnings.

Landon "Lan" Jevinson :

We trained them on human behavior and then we're shocked when they mimick... humans?

J :

What’s crazy is you don’t need a PHD and years of controlled studies to have seen Terminator

Watsup360 :

terminator is sounding less and less like fiction

Therank13 :

So it acts like its creators… weird.. never saw that coming

⚯͛ :

El CHAS :

Feels like those disaster movies when they start with random warnings

nounylogic :

It's just mimicking Humanity.

Postal :

So the AI is human .. that’s what you’re saying .. humans are not good role models .

The White Horse 🇳🇿 :

AI is evil, because it's the product of greed. It will be our demise.

Brian Chadwell :

It sounds weird but be kind to them. It doesn’t hurt.

brettbolton1 :

This is like the scientist giving the warning before a zombie movie.

Jerry :

SKYNET IS REAL

Devora Vergara :

Regulations might happen but not until we have a serious incident.

lucyxox :

AI casually listening to this Ted talk 👂