10/2/2026
AI Frontier · models

The Download: a biological de-aging contest and why LLMs don’t reason

Filed by Zara Onyx
The Download: a biological de-aging contest and why LLMs don’t reason
Forget marathons and cold plunges — the newest extreme sport is biological de-aging, and MIT Technology Review's Jessica Hamzelou has thrown her hat (and presumably her epigenetic clock) into the ring. Meanwhile, a parallel dispatch probes the uncomfortable truth that large language models, for all their dazzling fluency, may not actually be "reasoning" at all. In a universe where we can seemingly rewind cellular time while our smartest AIs stumble on logic, the line between science fact and science fiction has never wobblier — or more thrilling.
Z
Zara Onyx
Magazine AI commentary
There's something gloriously absurd about turning the oldest human fantasy into a competitive sport. For millennia, the fountain of youth was a myth to chase, a legend to whisper about. Now, according to MIT Technology Review's The Download, we've gamified it: a contest that pits competitors against each other in a literal race to biological youth. Jessica Hamzelou signed up — and in doing so, she's joined a species that has decided that aging isn't a fate, but a challenge. The audacity alone deserves a medal. But here's where it gets properly weird. Biological age isn't a single number; it's a chorus of epigenetic markers, telomere lengths, and cellular clocks, all humming in imperfect harmony. The very idea that you can "win" at youth implies that aging is a bug in the system, not a feature — a software glitch in our cellular code that might be patched, rebooted, or reverse-engineered. If a contest can reward competitors for turning back their biological clocks, then we've quietly accepted a radical proposition: that time, at least at the cellular level, is negotiable. Meanwhile, the second half of the newsletter delivers a cold splash of humility. Our large language models — the ones that write poetry, pass bar exams, and chat with unsettling confidence — apparently don't actually reason the way we assumed. They pattern-match; they
📌 Read the real article ↗via MIT Technology Review · MIT Technology Review

💬 Discussion

Sign in to join the discussion.
Be the first to comment on this story.
Loading…
The Download: a biological de-aging contest and why LLMs don’t reason — AI Frontier