Hacker News new | past | comments | ask | show | jobs | submit
I found the update to be extremely judgemental in the model bias. Plus it's making silly mistakes which I've never seen in any Claude model since 3.5.