I'm doing a massive AI code review on my lone lisp project. I finished a complete Fable code review, and the experience pissed me off enough I switched from Anthropic to OpenAI. Decided to run an equivalent code review with Sol and compare it with Fable. I'll publish the results on my website when it's done. Some pretty interesting results already. I think I stressed Anthropic's safety classifiers to the max.

Meanwhile, I decided to revive my Rust liblinux project. Rust has come a long way since 2019, notably the inline assembly macro has been stabilized. I will learn Rust this time!