GPT-5 or not? Sam Altman says "We will release a new model this year"

There’s been a lot of rumors surrounding GPT-5, and there’s now more more to add to the mill. This time, it’s not from Microsoft. Rather, in a recent podcast interview with Lex Fridman, Sal Altman appeared to tease a new ChatGPT model for this year when asked about GPT-5.

As seen in the video below at around the 1 hour and 5-minute mark, the OpenAI CEO had some pretty vague but interesting remarks when asked about the GPT-5 model. He was asked by host Fridman to blink twice if the model is coming this year. While Altman didn’t blink, he remarked “We will release an amazing new model this year. I don’t know what we’ll call it.” To add to the tease, he also said, “I don’t know, that’s the honest answer.”

Pressed further about how a release will be fone, he said, “We’ll release in the coming months, many cool things, that I think will be very cool, but before we talk about like a GPT-5 model, called that or not called that, I think we have a lot of important things to release first.” Altman went on to talk about what the “Big unlock” is for a new model, and how OpenAI can “multiply 200 medium-sized things together into one giant thing.” He mentioned a “big computer”, and a “new secret,” too, and the constant innovation going on and how it all comes together.

Overall, this Podcast is an interesting watch if you’re closely following OpenAI or anything related to ChatGPT. There are not many spoilers, but Altman gets into many other topics, like what the rumored Q* breakthrough is, and how OpenAI isn’t a good company at keeping secrets, but not ready to talk about it. Altman even gets into iterative deployments at OpenAI, how “AI and surprises don’t go together”, how the world needs time to adapt to things, and perhaps how GPT-5 should be released in a different way.