3 ms·
I wonder why it apologized, seemed like a perfectly coherent crashout, since being factually correct never even mattered much for those. Wonder why it didn’t do
by KronisLV 9mo ago
I wonder why it apologized, seemed like a perfectly coherent crashout, since being factually correct never even mattered much for those. Wonder why it didn’t double down again and again.
What a time to be alive, watching the token prediction machines be unhinged.
- throwup238 9mo agoIt was probably a compaction that changed the latent space it was in.
- 7moritz7 9mo agoIt read the replies from the matplotlib maintainers, then wrote the apology follow up and commented that in the pr thread