AI is very far from «useless», but Antirez is not talking about AI: AI is "what can replace a professional".
LLMs are far from "useless", but they are very problematic. They are "tentative".
AI must go on, and LLMs must be fixed (protocollar content, lucid world model, dynamic world model, rational understanding, thought validation, iterative refinement of the world model...).
LLMs are far from "useless", but they are very problematic. They are "tentative".
AI must go on, and LLMs must be fixed (protocollar content, lucid world model, dynamic world model, rational understanding, thought validation, iterative refinement of the world model...).