@WenShijin@rodydavis Got ultra 200 today. Usage feels very good and I enjoy the models a lot more than gpt (Claude is still better though I understand you can’t). I’d switch
@tunahorse21@Techniacus Carbon is just sooo slow. If they sped it up and let people use ultra in any harness it would be up there with Anthropic models.
@alvysdungeon@Joby_Fi@thsottiaux@AdamHoltererer Dude you sound like you are fishing for a benchmark task or something 😭. Openai just doesn't have taste like opus 5.5 has rn. The models are ok and still impressive just a step behind both on benchmarks and communication with user and making fewer mistakes.
@SStricklandMMA@waltermasterson Absolutely dude. An American company that has the privilege of selling to our advantageous market should never be allowed to hire cheap foreigners.
@oikpoop@mytypefyi Nobody’s “intimidated” by women. The one on the left looks like a nice, fun, smart person and the one on the right looks like a stupid, evil bitch.
@kimmonismus Without a doubt that’s what fable 5.5 would have been. That’s not their best internal model they almost certainly have models leaps and bounds ahead of this.
@Mattffrench@kimmonismus You are either a bot or stupid. “Best model” on betting markets measure https://t.co/ZGNerreMJE text arena scores. No correlation to how good the model is at real tasks.
@homecleaningguy@getProvenAI@coreyganim Fuck that. It’s not “just capitalism” and we make the rules here. You could argue for child labor under the same guise.
We don’t want to get used to it. We want change and punishment for people and companies harming American citizens. We have smart people willing to work hard.