maxsich

Minimal agents also allow room for more focused add-on tools/infra. I'm working on a context management layer[1] and it's very difficult to do well. In our benchmarking, we've observed simple harnesses like Stirrup[2] outperforming more elaborate ones.

[1] https://www.induction.ai/docs/context-management [2] https://github.com/ArtificialAnalysis/Stirrup

> Stop expecting an AI to cure all diseases or solve all problems just because it can read all the scholarship and “think” for a very long time. No matter how much an AI “knows,” it is always too little.

AI improves our problem solving in another way though! E.g. I can now quickly spin up frameworks to test/eval my hypotheses. This speeds up our ability to trial-and-error towards progress.

Was hoping this would be a sandwich art simulator:’)