The end of LLM in AI?

[Mik]

https://www.msn.com/en-us/news/technology/he-s-been-right-about-ai-for-40-years-now-he-thinks-everyone-is-wrong/ar-AA1QtEKt

[Horace]

Interesting, but not much insight into how or why this world model is better than the language models. If Lecun differs with nearly every other expert in the field on the issue, then my money is on nearly other expert in the field. As is, obviously, the world's money, at this point.

[Wim]

LeCun is not the only one:

https://www.theguardian.com/news/ng-interactive/2025/sep/16/song-chun-zhu-why-one-of-the-worlds-most-brilliant-ai-scientists-left-the-us-for-china

You might reconsider your investment strategy. China did.

[Mik]

Fascinating article. Thanks, Wim.

[Axtremus]

To the extent that “language” is just a small part of the “world,” it has always been obvious to me that AI cannot stop at “language” models. It’s just a matter of when and where we realize the limits of “language models” to be.

[Horace]

We will know we have achieved artificial general intelligence when the model communicates not with words, but with interpretive dance.

[Doctor Phibes]

@Horace said in The end of LLM in AI?:

We will know we have achieved artificial general intelligence when the model communicates not with words, but with interpretive dance.

Link to video

[89th]

Sure, changes the meaning of when I do “the robot” on the dance floor.