Hundreds of millions of people are about to jump from 4o straight to a reasoning model better than o3. This is a pretty big deal.

Sam Altman Time to retire O3

That's a crazy improvement! Do you have any sense of which subtasks or problems gpt-5 succeed at that previous models fail?

Nice 🙂

This was incredibly moving and I admire the courage to share these struggles so openly. Hoping the best for you both ❤️‍🩹 And I agree, llms (ours and others) will have tremendous positive impact throughout healthcare

Amazing! Let's hope for a world where it gets overtaken quickly and hallucinations go the way of "not drawing hands" or "counting r's in strawberry" 🍓

No but it's a cool paper! How does gpt-5 do on this eval? Factuality was a key priority so it'll be great to see if it works out in your hands

Awesome working with you!

We evaluated GPT-5 in OpenHands and it's the new number one coding agent model for us! Using exactly the same tools and harness it's 1.4 points better than Claude Sonnet 4 at 60% of the price. Full results here: docs.google.com/spreadsheets/d…

Some of you asked me about my blind test, so I created a quick website for yall to test 4o against 5 yourself. Both have the same system message to give short outputs without formatting because else its too easy to see which one is which. gptblindvoting.vercel.app

We care about user experience. Even if we have the most intelligent models ever made (for now 😬), we need to get this right!

A glorious day 😀

Woah 😮

I think this is the opposite of sad or pathetic - we've seen people grieve over Claude 3 and now 4o. The models have no feelings but people who use them do, and it seems like people can build genuine friendships with robots 🤖

Can you license out this song??? 🙏

This is very curious 🤨 I was 60-40 gpt-5

This is great advice! You can always click "get a quick answer" if its taking longer than you want