[SecretDreams]

> then it'll try to remove the process that is using it and promptly commit suicide.

Not unlike a child trying to take the safety cover off a plug so that they can stick a fork into it.

LLMs need that "world model" view that most people have acquired by their 20s where they (hopefully) stop to ask "why" before they "do".

[MichaelZuo]

That is a pretty good analogy. Like exceedingly smart 5 year olds.

Or whatever the age is before children typically develop object permanence, a theory of mind, and so on.

[fapjacks]

Not to sound like a codger, but we even said in the 90s that computers are just very fast idiots.

[moffkalast]

And they've been getting faster! Still idiots though.

[ulbu]

or pain perception

[wunderlotus]

> LLMs need that "world model" view that most people have acquired by their 20s where they (hopefully) stop to ask "why" before they "do".

The next evolution of multi agent orchestration / “advisor strategy” [1] will be branded in humanized language like this. Less about tokens and capability, more about wisdom and knowledge to guide a “younger” (less capable) model. Somebody will make a billion dollars by selling it as paired programming for LLMs.

[1] https://platform.claude.com/docs/en/agents-and-tools/tool-us...