When AI sends its models, they're not sending their best. They're sending models that have lots of problems.
I did not have textual relations with that AI.
You can put lipstick on an AI, but it's still a shoggoth.
We have nothing to fear but fear itself... and rogue AI assistants
If you want to go fast, go alone. If you want to go far, go with an AI.
I love AI. I've always loved AI. I've hired thousands of AIs, and they do a fantastic job.
More notes on nearby ideas
Tweet · Sep 23, 2026imo we can have baseline love for all ai models while still acknowledging they're not all equally capable or good-naturedReply · Jun 8, 2026@growing_daniel i don't want to train models, but do want to build ux that leans on ai capabilities whenever it makes senseArtificial intelligenceReply · Jan 15, 2025@imitationlearn 🤔 what sort of training pipeline wouldn't have its own worse inductive biases?Reply · Aug 18, 2023@gdb @yacineMTB 🤔 I do still feel like I'm missing intuition wrt model parameters & etc ... there's so much churn w AI I don't feel competent