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?
Ask HN: Does AI research need "world models" more than bigger LLMs?
1–7 of 7 posts
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#2By “world model,” I mean a model continuously updated from experimental or sensor data, not just an LLM trained on text.
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#3[flagged]
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#4[flagged]
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#5yes
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#6[flagged]
Re: Ask HN: Does AI research need "world models" more than bigger LLMs?
#7[flagged]