And it may lead some folks into discovering privacy-preserving local inference as an alternative for a lot of use cases, which is always a plus.
I did notice this an wonder what changed - I do periodic data backups of various services, and up until recently it was impressive, as ChatGPTs email with data zip file link arrived maybe within 1-3 min of the request, for around a ~1GB file. I have similar amount of data now (even less, I pruned some), yet now the file takes a really long time to prepare and receive.
andsoitis•1h ago
johnbarron•1h ago
Finnucane•1h ago
estimator7292•1h ago
zamalek•1h ago
BoredPositron•1h ago