Sam Altman built GPT-5 and now he’s scared of it, compares it to sci-fi thriller

In a recent podcast with Theo Von, Altman said parts of testing GPT-5 left him nervous as he compared its power and unpredictability to the Manhattan Project.

OpenAI CEO Sam Altman has set off alarm bells, not with the new capabilities of AI, but with his own reaction to it. In a recent episode of ‘This Past Weekend podcast’ with Theo Von, Altman described moments while testing GPT-5 that left him deeply unsettled. “It feels very fast,” he said, admitting that he felt “very nervous” during some of the sessions. What stood out even more was his comparison of the project to a turning point in world history: “I sort of felt like it was the Manhattan Project.”

OpenAI CEO Sam Altman's latest comments about GPT-5 have sparked more unease than excitement. (Bloomberg)

By Aishwarya Faraswal

Aishwarya is your go-to guide for turning your home, especially your kitchen, into the smartest and most efficient space on the block. If you ever find yourself wondering which air fryer actually delivers that perfect crisp, whether a new-age water purifier is worth the hype, or how to make your refrigerator work smarter (not harder), she’s the one to turn to. For over five years, Aishwarya has dived deep into the world of home and kitchen tech, reviewing everything from air fryers and microwaves to chimneys and water purifiers. Her approach is simple: cut the jargon, highlight genuine innovation, and spotlight budget-friendly solutions without compromising on quality. Her column is where fresh launches meet honest opinions. There is no sugarcoating and no fancy marketing spin. Every review comes from real, hands-on experience, giving readers the unfiltered truth, good or bad (sometimes brutal). A big believer in viral “how-to” hacks, Aishwarya loves sharing appliance hacks like microwave-cleaning tricks that involve a lemon and absolutely zero scrubbing, fridge-care tips (when sometimes all you need to do is defrost) that save you trouble, and everything in between. She tries her best for home tech to feel less overwhelming and a whole lot more fun.

Read moreRead less

That’s not the kind of analogy you expect during a product preview. GPT-5 is expected to be much faster and smarter than GPT-4. But instead of focusing on what it can do, Altman talked more about what it could mean and how little oversight there is to keep its power in check.

“There are no adults in the room,” he said, pointing to how poorly AI is being regulated right now. Coming from the CEO of one of the world’s leading AI companies, that’s a serious reality check.

Altman’s concern wasn’t limited to the tech itself. He also warned, “Something about collectively deciding we're going to live our lives the way AI tells us feels bad and dangerous.” It’s a jarring thought coming from someone responsible for building tools that could soon shape everything from education to employment to relationships.

This isn’t the first time Altman has raised flags about AI's risks. He’s previously said AI could “go quite wrong” and stressed the need for responsible development. But this time, his tone felt more personal and less calculated, more like someone seeing an unpredictable force take shape in front of them.

What’s the broader message here? GPT-5 could be incredibly powerful but that power may be arriving faster than our systems, rules and ethics are ready to handle. And when the man leading its development says he’s scared, maybe we all should be paying closer attention.