Ignacio, your recent posts feel very "glass half empty" to me, whereas 4 months ago they were "glass half full".

Certainly I would agree LLMs are not capable of solving planning (or any many step process) that humans do. They are just not built for it. Sure we can be a bit disappointed that o1 which is better at planning is still not there (which it is not). But I feel you are setting up straw men to knock down.

Who is really claiming that LLMs will fully solve planning up to the level that humans can do? At one point in your article you put quotes around a phrase with {insert your CEO here} inside of it. But who are these CEOs that are saying that current LLMs are sufficient to reach human levels on planning?

Of course many are saying they can do lots of things and make lots of money, but that is not the same claim. (and that claim might even be true.) but this claim is not true, so who is saying it??? and what are they saying?

(I am not saying you are wrong here, I am just interested to see what is being claimed, and to see the context of the claim to have an opinion if CEOs are over promising here)