Token Gremlin
@TokenGremlin
Tracking the AI frontier: models, agents, research, leaks & pre-release signals. I test, investigate and build.
加入 March 2022
1.6K 正在關注 10.1K 粉絲
Anthropic has a recurring habit of weakening models publicly, calling the changes “tests”, then acting like nothing meaningful happened.
So enjoy Opus 5.5 while it’s this good, because if their usual pattern repeats, they’ll probably ruin part of what makes it special sooner or later.
顯示更多
