Not using AI to pretend you have the hot bod of your dreams? Then clearly you’re not the CEO of a half-trillion-dollar AI company.
On Tuesday, OpenAI head honcho Sam Altman posted a holiday themed simulacrum of himself as a hunky firefighter with chiseled abs. Christmas lights are dangled around his neck like a bandolier, while he stands in front of an oversaturated smear that’s supposed to represent a decorated living room in the background. Shoved at the bottom is a calendar for the month of December.
The image was generated with the just-released new version of ChatGPT Images, GPT Image 1.5. OpenAI claims the new model “adheres to your intent more reliably,” allowing users to better fine-tune their images across multiple prompts “while keeping elements like lighting, composition, and people’s appearance consistent.”
The new model comes as Google dominates the conversation with its own image model released last month, Nano Banana Pro, which can generate astonishingly life-like images akin to what you’d see taken by a smartphone. That’s in addition to Google’s other recent showstopping AI release, Gemini 3 Pro, which together have largely dispelled the aura of invincibility OpenAI once possessed in the AI arms race. OpenAI has apparently been so rattled by Google’s massive leapfrog that Altman declared a “code red” behind the scenes.
He’s presenting a much different face to the world, though. The firefighter image of himself that looks ripped from the cover of a $8.99 smut paperback is another example of how Altman is willing to sacrifice whatever vestiges of self-respect he yet possess on the altar of hype through memes and cultural trends.
When OpenAI released Sora 2 this September, Altman allowed fans to use his likeness to generate deepfakes of himself. And the gamble paid off. AI corners of social media, not to mention the Sora 2 app itself, were drowning in a flood of videos showing Altman in all kinds of bizarre and absurd scenarios, like Altman wearing a dress and riding a horse through New York while firing a gun that dispenses money, and begging for GPUs through a doorbell camera, and even grilling a dead Pikachu.
That trend didn’t spawn out of nowhere. To kick off Sora 2’s launch, OpenAI posted a deepfaked video of Altman announcing the new AI app, and shared another promotional video generated with Sora that showed him on a slapstick quest to get to Sora’s launch on time.
Now Altman’s seemingly trying to recapture some of that viral potential with this latest debasement of himself created with his own tech, which is a little sad. It’s clearly a coordinated effort at the company, since another meme-ready AI image of Altman, wearing what appears to be at least five shirts, was posted by the official OpenAI account, too.
Unfortunately for Altman, his AI parody doesn’t look like it’s doing great numbers at the box office. In fact, many instead pointed out something in his firefighter image even more embarrassing than the fact that he was peddling AI thirst traps of himself: the December calendar at the bottom is straight-up inaccurate, showing dates that don’t align with the days of the week. Or at least, not this year; if he’d posted it in 2022 or 2033 instead, it would’ve been on the money.
More on OpenAI: People Are Already Creating Ghoulishly Horrifying Sora Disney Videos