Post

I’m more careful since I read thinking traces where the LLM was like “hmm by covrrsge the human must have mean coverage”, and I’m pretty sure if Slack somehow tapped into the thinking traces of my colleagues and I had to painfully watch them mentally fix my typos I’d be carful there too

I waste way too much time trying to word things correctly when I honestly should just spit out whatever comes to mind and let the robot figure it out. They are surprisingly good at making sense of bizarre run on sentences.

i like that argument i'm going to adopt it