He might want to talk to his boss about this.

Chatbot Mania

Facebook's AI guru and machine learning pioneer Yann Lecun is coming for the artificial intelligence chatbot craze — and it may put him at odds with his own employer.

As Fortune reports, Meta's chief AI scientist Yann Lecun admitted during a talk in Paris this week that he's not exactly a fan of the current spate of chatbots and the large language models (LLMs) they're built on.

"A lot of people are imagining all kinds of catastrophe scenarios because of AI, and it’s because they have in mind these auto-regressive LLMs that kind of spew nonsense sometimes," he told the Meta Innovation Press Day crowd. "They say it’s not safe. They are right. It’s not. But it’s also not the future."

He went on to argue that today's LLMs, such as OpenAI's GPT-3 and GPT-4 which undergirds ChatGPT, will eventually be replaced by better and more robust algorithms — a statement that's extra-interesting, because Meta CEO Mark Zuckerberg seems to be all-in on making sure his company doesn't miss the craze.

Zuck Moves

Indeed, Zuckerberg announced in an all-hands meeting just last week that the company formerly known as Facebook intends to jam AI into "every single one" of its products.

Axios, which broke the news of the initiative, reported that part of the plan includes allowing users to "use a text prompt to modify their own photos and share them in Instagram Stories" and to insert chatbots with different personalities into Facebook Messenger. Unless Facebook is sitting on some amazing and secret tech, that'll almost certainly be powered by LLMs.

To be fair, Zuckerberg and LeCun have very different roles and may, in fact, agree in private about what the future of AI holds, and what value today's chatbots could bring to the company — but it's fascinating, to say the least, to hear one of the most respected names in AI trashing LLMs right after his boss announced that they're the root of the company's next pivot.

More on Facebook — er, Meta: Facebook Has Crowbarred Open the Pandora's Box of AI, Experts Warn


Share This Article