Commentary on X
@gdb I don't think I'm alone in saying the better price/perf is Anthropic as apparently no one else has a capable model. If your model is incapable of completing the task, you're a waste at any price point.
11:01 AM · Jul 15, 2026@gdb I think that's bad PR talking about price/performance, it makes you look like underdog, giving AMD vibes. You're the one started it all, you should aim for the absolute best models, my 2 cents.
10:32 AM · Jul 15, 2026@gdb This kind of open communication is insane (in a good way).
10:34 AM · Jul 15, 2026@gdb This is awesome!! I love this attitude.
10:31 AM · Jul 15, 2026View all
@davidkogan_ @gdb Thanks man,, gonna do it now.
10:05 AM · Jul 15, 2026Note the current expectations are still around test time compute/more tokens for a given task This is not the case Tokens per task will now drop even as quality improves Cost per intelligence equivalent token will drop 100x Jevon's paradox does not apply here
8:07 AM · Jul 15, 2026