unsplash.com/Levart

OpenAI has been in the news recently, but then again, when is it not? It launched ChatGPT 4.5 on February 27 and started rolling it out in early March. CEO Sam Altman sounded like a new father, writing on X (which he still calls Twitter) that he was proud of a “preemie baby for learning how to eat on his own!” He tweeted that 4.5 was the first model “that feels like talking to a thoughtful person” and there was a “magic to it that I haven’t felt before”. On March 11, he tweeted that it had trained a new model that is good at creative writing. “This is the first time I have been really struck by something written by AI; it got the vibe of metafiction so right.” The prompt was: “Please write a metafictional literary short story about AI and grief.” The piece, just over 1 000 words, is surprisingly good, and goes beyond the shallow mimicry of earlier models. Author Jeanette Winterson, writing in the Guardian, said it was “beautiful and moving”.

OpenAI is doing very well out of GenAI. It’s valued at somewhere over $300bn or will be after a new funding round, and is making around $300mn a month. It expects to make over $10bn in sales this year. But it’s also shelling out billions in running costs. The arrival of DeepSeek has seemingly had little effect on the company, but it’s now waded into how it sees AI playing out in geopolitics, and what role it should play.