AI pioneer Yann LeCun doesn't think artificial intelligence is actually becoming intelligent.
LeCun (professor at New York University, senior research fellow at Meta, and recipient of the prestigious AM Turning Award) has been outspoken about his skepticism, for example, before worrying about controlling superintelligent AI, saying, must do the following,” he tweeted. This is the beginning of tips for designing systems that are smarter than your cat. ”
He elaborated on his opinion in an interview with the Wall Street Journal, responding to a question about AI becoming smart enough to pose a threat to humanity: ”
LeCun argued that today's large-scale language models lack several important cat-level functions, such as persistent memory, reasoning, planning, and understanding the physical world. In his view, LLM only demonstrates that “you can manipulate language but you're not smart,” and will never lead to true artificial general intelligence (AGI).
That's not to say he's completely skeptical of AGI. However, he said a new approach would be needed. For example, he pointed to efforts by Meta's foundational AI research team to avoid digesting real-world videos.