OpenAI o1 model coding performance is mindblowing. Today, I built a famous "Chrome Dinosaur Game" in less than 5 minutes. I added a new functionality where the dinosaur bends instead of jumping to avoid birds.

that's mega dope, however regarding the dino game, i'm sorry to tell you that the dinosaur has been able to bend for many years now, with the down arrow :p

really? I haven't played the game for like many years, but I saw a video and thought bend could be cool.

Your perception is that it’s built a game it’s basically stolen code from trained knowledge and knows how to put the code in the right place, not so freaky if you know …

That is one of the most childish comment I have ever read.

i feel that is better than sonnet, i have seen some benchmarks and sonnet is still the best, but the feeling is different.

sonnet is also very good for code but the current o1 models is way better.

Bemd function already been , isn't it

Your mind blows when a memory fetches a string? I mean well yeah… flip flops are amazing.

Is this really better than what we could already do? Chatbots could be used to make simple games since like 2022, right?

Can you use the o1 model? Or are you referring to o1-mini? o1 preview? Which one are you referring to?

o1-preview is best for coding