Most users still believe that artificial intelligence could match and even surpass human intelligence. But for Yann LeCun, the lead researcher Target AI, this trend will remain in the realm of fiction. He presented significant arguments in support of his claims.
“Limited understanding of logic”. This excerpt from Yann LeCun’s interview caused a reaction from netizens. For this high-tech expert, reference LLM models are unable to imitate reasoning ability of a human being. For him, these artificial intelligences will remain tools, regardless of their performance. However, Meta’s head of artificial intelligence predicted that artificial intelligence could reach this level, but it would take a long time to develop this technology.
Yan LeCun is not convinced by the arrival of IAG
AGI, or artificial general intelligence, is the pinnacle of AI. At this level, these tools will be able to match human intelligence. Sam Altmanand others big names in high technology are very optimistic about this topic. But Yann LeCun disagrees. He likes LLM models of artificial intelligence Gemini Or ChatGPT they are very limited.
“It certainly seems like reasoning to most people, but it is first and foremost a matter utilize knowledge gathered from numerous training data», explains LeCun in an interview with the Financial Times. He also added that these are LLM models inherently dangerous. Indeed, these AIs rely on training data, which may contain inaccuracies or be completely out of date.
Certainly, Yann LeCun gave his opinion on the level current LLM models. The latter do not have the capacity to reach the level of intelligence of human beings. But he doesn’t plan to stop there. With over 500 employees, this expert intends to create a new form of intelligence ArtificialIAG?
“We are at a point where we believe we are at the dawn of the next generation of artificial intelligence systems,” explains Yann LeCun in his interview.
Team Meta will then to create an LLM system quite different from the current versions . This AI will have advanced analysis capability. Instead of scanning a database, Meta’s new technology will be equipped with an understanding and logical approach. The project will be realized, even if it takes 10 years.
However, investments are increasing to implement AGI
Yann LeCun doesn’t see it cominggeneral artificial intelligence 10 years ago.But investors disagree. Scale, an artificial intelligence company, has raised more than $1 billion to fuel its research. Another French start-up, H, also raised $220 million to realize its “full-AGI” project. And these are just examples.
Will these investments be key to accelerating the implementation of AGI? Don’t hesitate to express yourself in the comments.
- Share the article:
Our blog is powered by readers. When you shop through links on our site, we may earn an affiliate commission.