By far the most important lesson I took out of this game is that whenever there’s behavior that needs to be repeated around to multiple types of entities, it’s better to default to copypasting it than to abstracting/generalizing it too early.
This is a very very hard thing to do in practice. As programmers we’re sort of wired to see repetition and want to get rid of it as fast as possible, but I’ve found that that impulse generally creates more problems than it solves. The main problem it creates is that early generalizations are often wrong, and when a generalization is wrong it ossifies the structure of the code around it in a way that is harder to fix and change than if it wasn’t there in the first place.
- Understanding GPT tokenizers - 8th June 2023
- Weeknotes: Parquet in Datasette Lite, various talks, more LLM hacking - 4th June 2023
- It's infuriatingly hard to understand how closed models train on their input - 4th June 2023
- ChatGPT should include inline tips - 30th May 2023
- Lawyer cites fake cases invented by ChatGPT, judge is not amused - 27th May 2023
- llm, ttok and strip-tags - CLI tools for working with ChatGPT and other LLMs - 18th May 2023