@mrsiipa There was a good one a while ago about using reflections and progressive scoring to mimic the “reasoning” behavior with o1 - I find this general approach works great https://t.co/ffwsoumqQs
@mrsiipa There was a good one a while ago about using reflections and progressive scoring to mimic the “reasoning” behavior with o1 - I find this general approach works great https://t.co/ffwsoumqQs
In my experience, the most useful approaches to AI are:
- Ask clearly bounded questions, where you can easily inspect the results.
- Don’t let AI make decisions for you. Instead, challenge it to broaden your options.
- Take advantage of the fact that it doesn’t have feelings, and use its honesty to get useful feedback.
ChatGPT is dumber than it looks
A while ago I got interested in being more agentic. Emmett Shear tweeted that it was teachable. He would train people by prompting them with certain questions. I gathered those questions and put them into a flow chart that I go through to problem-solve: https://t.co/kM0MID8bT5... See more