Ask HN: Does AI research need "world models" more than bigger LLMs?

[unjuno]

It seems that generating hypotheses is no longer the main problem. The harder problem may be continuously collecting real-world data, building accurate world models, and using them to guide experiments. Is this where automated research is heading?

[unjuno]

By “world model,” I mean a model continuously updated from experimental or sensor data, not just an LLM trained on text.

[alexander2002]

yes