Everyone’s been raving about o1, but honestly, I’ve found it lacking for creative writing. It felt like asking a brilliant math professor to help with crafting a fun story.
But I just tried 4o again after a month, and wow—what a difference! It’s exponentially better at writing captivating and engaging paragraphs and stories. There’s a lot more emotion in its writing now, and it seems way more confident.
Has anyone else noticed this?
Here’s a fun creative experiment: ask it to generate a choose-your-own-adventure scenario based on anything you like—movies, games, real life, whatever.
When it presents you with the choices, don’t pick any of them. Instead, choose some random, absurd action.
It’s amazing how it seamlessly incorporates your outlandish choice into the story without missing a beat.
I’ve been using it daily for a few months, and I’ve noticed the message “you are using a new version of GPT 4o” at least twice in the past week. I keep track of which version I’m using because they have different strengths.
We know that Strawberry is being used to train Orion, which is likely GPT 5 or something very similar.
While I can only speculate, it seems likely that they’re incorporating improvements from the next model into GPT 4o for a soft launch, testing, and to reduce resource requirements, among other things.
Yep. I asked it to write a short femme fatale/detective noir piece, and I was quite amused that the character’s perfume was “Cigarettes and Roses.” It looks like GPT has a bit of cheeky flair!
I noticed that too, but I’m not sure what changed.
Yeah, 4o’s writing definitely feels more vibrant and creative now—I’ve noticed the same!