AI pioneer Yann LeCun does not believe that artificial intelligence is on the brink of achieving true intelligence. LeCun, a professor at New York University, a senior researcher at Meta, and a recipient of the esteemed A. M. Turing Award, has openly expressed his doubts in the past. For instance, he tweeted that before we start worrying about managing super-intelligent AI, “we first need even the slightest hint of a design for a system that's smarter than a house cat. ” In a recent interview with the Wall Street Journal, he responded to a question about AI potentially becoming a threat to humanity by stating, “You’ll have to excuse my French, but that’s complete B. S. ” LeCun contended that current large language models (LLMs) lack essential cat-like abilities, such as persistent memory, reasoning, planning, and an understanding of the physical world.

According to him, LLMs simply illustrate that “you can manipulate language without being intelligent” and will not result in genuine artificial general intelligence (AGI). Although he is not entirely dismissive of the possibility of AGI, he believes that new methodologies are required. He highlighted efforts made by his Fundamental AI Research team at Meta related to processing real-world video as an example.