Meta’s Top AI Scientist Says AI Systems’ Intelligence Isn’t At Par With Dogs As It Can’t Comprehend The Reality Of The Real World

The leading AI scientist at Meta claims Generative AI might be trending for obvious reasons in the tech world but it’s yet to attain the same level of intelligence as dogs.

Yann LeCun adds how AI technology-based systems cannot comprehend that reality arising in the real world. And that’s a huge lacking when you come to think of it. But that does not mean we won’t be seeing it ramp up and supersede the brain power and intellect of humans, one day.

But that day is yet to come. However, he refused to call it a threat to humans and wants people to calm down in this regard.

The news comes as LeCun was seen speaking in Paris on Thursday at the much-revered Viva Tech conference. Moreover, he shed light on how the problems linked to AI have to do with its a great many limitations and how it’s being trained solely on text models and nothing else.

In addition to that, he claims that most of the knowledge that the human mind possesses does not have anything to do with a specific native language. Hence, that’s why a huge chunk of the entire experience of humans isn’t taken on board by artificial intelligence.

In the same way, he claims that generative AI has the capability of passing exams in America like the Bar Exam. But with that being said, a test was necessary if you wished to go all the way to become a lawyer in the country so it’s not that simple.

So it’s not just that the technology needs to be at par with humans, it’s got a long way to go to be at par with animals like dogs.

He also emphasized on another striking example of how AI is yet to be compared to the intellect of humans. LeCun gave the example that a young baby who is a couple of months old says that their intellect might see something floating and not say a lot about it. But a few months down the line, when their intellect is further developed, they’ll soon realize that perhaps that particular object shouldn’t be floating in the first place. Hence, their reaction would be one of sheer surprise.

He similarly admitted how researchers have zero clue about how they can best copy the human intellect into AI but that does not mean they’re done trying. But as far as reaching the level of intelligence of humans and dogs is concerned, he doesn’t see that happening too soon.

We know that ever since OpenAI launched its much-hyped ChatGPT tool to the masses, there was plenty of fear across the board about whether or not it might lead to human extinction and people getting fired as their jobs would be replaced by AI.

But LeCun says that he does not believe in any of this.

He sees a future where you’ll find people having their own personal assistants based on AI technology. They just might be smarter than the owner and would help in carrying out daily duties in their lives.

Moreover, he similarly sees the world of generative AI being curated in a manner where it’s not out of hand but functions in a very controllable manner where it would be obedient to humans.

Hence, anyone that feels that AI would take complete control of the world and become over-smart would really need to relax, he concluded.


Read next: Google Domains Unexpectedly Announces Shut Down As Squarespace Ready To Take Over Business
Previous Post Next Post