Is Claude based?

IMO there is no moral difference between burning ants with a magnifying glass and sending annoying prompts to ChatGPT

Product managers in control

Yesterday I read the SAM2 paper with some friends. I wrote some notes comparing it to the original SAM paper, others reading the paper may find useful: docs.google.com/presentation/d…

New Gemini image editing is good, definitely filters less (I tried some race/hate speech stuff) But the quality of the final image is still too low for basic editing tasks. It garbles the text in this image, while ChatGPT leaves it pristine. Neither result is useable, btw

POV you're about to waste your Sunday afternoon

UPDATE: This also seems to happen with some new keys. Maybe OpenAI is running an experiment that breaks a random subset of keys?

8647 would be a beautiful name for a baby boy

I guess it's just clown car prompting

It's not affirmative action because it's not zero sum. Adding foreign students to US universities * decreases cost for Americans (foreigners pay more) * increases education quality (higher avg IQ of peers) * increases signaling value (avg degree holder is more intelligent)