Tried twice and didn't get usable results. The first reduced the tokens by 80% and the results were better than I would've expected. The second try reduced the

Tried twice and didn't get usable results. The first reduced the tokens by 80% and the results were better than I would've expected. The second try reduced the same prompt by only 20% and the results were worse than the first try. Both missed a key detail about using a specific term in the response. Interesting idea. Not sure how useful it will be with costs dropping fast. But if you can get it to work better, it might be useful for recursive prompt-chains like Auto-GPT or babyAGI.