I am currently letting GPT-5-Pro review my weekly post draft. Do you predict I will judge this was worth doing?

For LLMs right now, I think of it as four ‘speed tiers’: 1. Quick and easy. You use this for trivial easy questions and ‘just chatting.’ Matter of taste, GPT-5 is good here, Sonnet is good here, Gemini Flash, etc. Most of the time you are wrong to be here and should be at #2 or #3 instead. 2. Brief thought. Not instant, not minutes. Use primarily Claude Opus 4.1. 3. Moderate thought. You can wait a few minutes. Use primarily GPT-5-Thinking and back it up with Claude Opus 4.1. 4. Extensive thought. You can wait for a while. Use GPT-5-Pro and back it up with Opus in Research mode. Consider also firing up Gemini Deep Research or Deep Thinking, etc, and anything else you have handy cause why not. Compare and contrast.

RT @VitalikButerin: A good book, worth reading to understand the basic case for why many people, even those who are generally very enthusia…

I do think the underlying fact is mostly true but this statistic is because there are only 12k members total and the Girlfriend subreddit decided to move in with the Boyfriend one since they were effectively very similar.

The big confusion with GPT-5 is they released 4+ models at once - GPT-5-chat that routes, GPT-5, GPT-5-thinking, GPT-5-pro. Thinking and Pro seem better and like better upgrades, whereas GPT-5's core idea of routing seems mostly a big miss.

Had a disappointing dinner trying out a new place despite a very high rating, so I looked at the reviews, and... Come on, at least remove the em dashes, but I didn't check before going and that's on me.

It's weird writing up reactions when so much of it is 'give me back my model picker' or 'give me back 4o' when you now have a model picker and you have 4o.

All right, it's time. GPT-5 reaction thread. Hit me.

I am continuously amazed how people say things like 'it's 10x lowering my productivity' and haven't paid a monthly fee to undo that.

Truly it is just like us.