Pinned Tweet
bitget lost 352 million dollars tonight and the ceo goes live
Just for some random guy to ask if she's single
We are cooked 😭
Sponsored partnerships like this should be illegal
Like some company is literally paying money for people to blantantly LIE bout their model
Wtf
I think I've gotten so used to Opus 5.5 that I forgot how dumb other models can be.
We use Cursor at work, and Grok 4.7 sometimes does the dumbest shit, even on simple tasks.
It often completely misses the intent behind what you're asking for
Google gatekeeping Argon for two weeks, only to drop it exclusively for Ultra subscribers lol
Must be one hell of a slap in the face for regular Google users.
wait.. up to WHAT? 1000 agents? Usage limits gonna get drained like crazy
Replying to @ClaudeDevs
To use it, configure your agent with multiagent type: multiagent_20261001 and then ask Claude to run a workflow. Claude will write the plan and handle orchestration of up to 1,000 agents per run.
Mathematicians are handling the current situation of AI solving math problems really terribly imo. Even worse than software devs.
Feels like they care more about protecting their profession than actually solving math problems.
Why wouldn’t mathematicians want solutions to these huge open maths problems?
shoutout @mathandcobb for being a consistent voice in this conversation
Thank u musk for changing that and now using opus !!!!!!!
After testing Grok Bot non-stop for several days, you can already tell where the future is heading.
Sadly, the models behind Grok Bot are still too weak to handle meaningful, larger tasks without getting frustrating results. Opus 5.5 really highlighted the weaknesses of other models. With Grok 4.8/5, Grok Bot could be on top (it already is kinda, great product) 🙌
Hopefully, the limits will be increased 🤪

















