@Cail_lion@elonmusk Where’s the evidence that it isn’t happening?!? What are you even talking about?? The proof that it’s not happening is THERE IS NO EVIDENCE. Also, I was not calling election fraud in 2016. I was thinking the world has gone crazy, but I never cried fraud.
@RyanPer07224297@elonmusk This is a joke right? Did you think he had a chance to win? In the general? In LA?!? Where republican registration is around 15%?!? Oh man, this is worst than I thought.
When Codex solves a really big problem that we have been working on for a while, my first reaction is to go tell Codex. Then I remember, the model doesn’t care and that is a little sad.😔
@argofowl Poor take. This is a natural progression. The idea is to make coding more accessible not gate keep in a separate app. Coding inside of Chat makes a lot of sense. Currently a lot of my production moves in and out of Chat right now.
@DavidOndrej1 This bench is so exposed. 4.7 is not better than 5.5 in the real world. And Gemini is further behind than this shows. The whole thing is a mess.
@yashmp2004 $100 Codex is the only serious answer, but if you have $200 for Claude then you should spend the whole $200 on Open Ai Pro. Use Img Gen 2 for front end help and you’re golden.
@Teknium Hey this makes sense. OpenClaw was the opposite approach. Make it intentionally a little harder to gate keep it from people who didn’t understand what they were doing. Haven’t run Hermes yet, but looking forward to giving it a go this week.
@theo I agree with this. 4.8 is great to use and feels like it is much more in line with what I want, however it’s still Claude and can suffer when it has to dig deep to make things work in my existing code base without some real steering. Codex doesn’t need as much help.