[emp17344]

It’s laughable to suggest AI is improving exponentially. Recent models, like o3, hallucinate more than their precursors and are therefore less useful.

[energy123]

Reasoning models didn't even publicly exist before September 2024.