This is why I’m in AI interface work. It’s literally alignment with human intentions, or from another angle, it’s augmenting human intelligence
But I think — _especially_ in the rapidly accelerating NLP and generative AI space — good, intuitive, understandable interfaces aren't just nice-to-haves, it's a part of the alignment/safety problem.
— Linus (@thesephist) August 2, 2022
New interface materials might get us there!
🔗 https://t.co/lb81aJcmAp pic.twitter.com/0octPQV9O7
