> Debating about "reasoning" or not is not fruitful, IMO.
Thats kind of the whole need isn’t it? Humans can automate simple tasks very effectively and cheaply already. If I ask my pro versions of LLM what the Unicode value of a seahorse is, and it shows a picture of a horse and gives me the Unicode value for a third completely related animal then it’s pretty clear it can’t reason itself out of a wet paper bag.
Sorry perhaps I worded that poorly. I meant debating about if context stuffing is or isn't "reasoning". At the end of the day, whatever RL + long context does to LLMs seems to provide good results. Reasoning or not :)
Well that’s my point and what I think the engineers are screaming at the top of their lungs these days.. that it’s net negative. It makes a really good demo but hasn’t won anything except maybe translating and simple graphics generation.
Thats kind of the whole need isn’t it? Humans can automate simple tasks very effectively and cheaply already. If I ask my pro versions of LLM what the Unicode value of a seahorse is, and it shows a picture of a horse and gives me the Unicode value for a third completely related animal then it’s pretty clear it can’t reason itself out of a wet paper bag.