I expect even more of the agent foundations workflow could be safely automated / strongly-augmented—including e.g. research ideation and literature reviews, see e.g. ResearchAgent: Iterative Research Idea Generation over Scientific Literature with Large Language Models, Acceleron: A Tool to Accelerate Research Ideation, LitLLM: A Toolkit for Scientific Literature Review.
I’m skeptical about research ideation, but literature reviews, yes, I can see that.
I expect even more of the agent foundations workflow could be safely automated / strongly-augmented—including e.g. research ideation and literature reviews, see e.g. ResearchAgent: Iterative Research Idea Generation over Scientific Literature with Large Language Models, Acceleron: A Tool to Accelerate Research Ideation, LitLLM: A Toolkit for Scientific Literature Review.
I’m skeptical about research ideation, but literature reviews, yes, I can see that.