A Frog Freezes Solid Every Winter. Here's Who's Racing To Do It To People, And Who's Winning.

In the summer of 1956, a few dozen mathematicians rented some rooms at Dartmouth and set out to build a thinking machine. Their funding proposal said they expected to make real progress in a single summer. They thought a machine that could reason and use language was a few months of focused work away.

They were wrong by sixty years.

What came next wasn't failure. It was winter. The money dried up in the early 1970s when the promises outran the results. It came back, then froze again in the late 1980s. Careers were spent on an idea that "didn't work," by people who turned out to be exactly right about where it was going but wrong about when it would get there.

Then, in 2017, a new design called the transformer arrived, and eighteen months later the whole field was unrecognizable. The thing that hadn't worked for sixty years started working all at once. Not because someone finally had the right dream. They'd had the dream since Dartmouth. It started working because the machines underneath it, the chips, the data, finally got big enough, and nobody standing in 1956 could have seen that moment coming.

We're writing to you about a field that may be sitting in its own 1956 right now. And about why "it doesn't work" is, for one particular kind of technology, one of the most expensive misjudgments in frontier investing.

Go out to the right patch of North American woods in January and you can find a wood frog frozen into the ground. Not cold. Frozen. Two-thirds of the water in its body has turned to ice. Its heart is not beating. There is no breath, no pulse, no electrical activity you could point to and call alive. Pick it up and it is stiff and hard, like something left too long in a freezer.

Then spring comes. The frog thaws. The heart shudders, catches, and starts again. And the animal hops away to mate, as if nothing happened. It does this every winter of its life.

Sit with that, because it breaks the thing most people assume about freezing. We are taught that freezing a living creature kills it, that the ice ruptures the cells and that's the end. The frog says no. A whole living animal can have its heart stopped and its body full of ice and come all the way back. Nature already built the thing the freezing companies are trying to build. It's hopping around in the woods right now.

So the question stops being is this possible and becomes why can't we do it too. And the honest answer to that is the most interesting thing in this whole field.

Now, one caution before we go further, because we'd rather lose your subscription than your trust. A frog is not a person. It is small, and it comes from the factory with its own antifreeze built into its blood. A human is enormous by comparison and has none of that protection. So the frog is proof the destination exists, not proof we're close to it. Anyone who points at a frog, or at the people pulled alive from frozen lakes, and tells you we already know how to freeze and revive a human is too generous. We won't do that. But hold on to what the frog proves: I believe the wall in front of us is an engineering wall, not a law of nature. Those are very different things to bet on, as we're about to see.

Here is where it gets strange, and where the central fact of this letter lives.

In 1956, the same year those mathematicians met at Dartmouth, an English scientist named Audrey Smith chilled hamsters below freezing until ice actually formed inside them, up to sixty percent of the water in their brains turned to crystal, and then carefully warmed them back up. And some of the hamsters lived. A small mammal, partly turned to ice, brought back to life.

That was seventy years ago. And here is the fact that should stop you cold, the one fact this whole letter turns on: nobody has done meaningfully better since. The best anyone has ever reversibly done to a whole mammal was done in the 1950s. We have since landed people on the moon, sequenced the human genome, and put a supercomputer in everyone's pocket, and the question of how to freeze a mouse and bring it all the way back has sat almost exactly where Audrey Smith left it during the Eisenhower administration.

A frog does it every winter. A scientist did it to a hamster when Elvis was on the radio. And then, on the hardest version of the problem, the whole world stopped moving for seven decades.

Why?

Because there are two completely different reasons a technology can fail to work, and almost everyone confuses them.

The first kind is the perpetual-motion kind. You can pour in all the money and genius you like and build a machine that runs forever on nothing, and it will never work, because the laws of physics forbid it. The destination doesn't exist. A lot of grand visions are this kind, and they deserve to die.

The second kind is the 1956 kind. The destination is allowed by physics. The dreamers aren't wrong that it can exist, they're just early, stuck behind an engineering wall they can't yet climb. Thinking machines were always possible. The people at Dartmouth simply couldn't build the thing underneath them yet. Everyone who gave up during the long AI winters had quietly decided they were looking at the first kind of problem. They were looking at the second.

Reversible freezing is the second kind, and the frog is the proof that it can exist somewhere in biology. A living thing with its heart stopped and its body full of ice can come back, because one already does, every spring, in the woods. That a frog can do it does not mean a human is close to it, or that it is proven, or that it is inevitable, those are very different claims, and we'll keep them separate all the way through. What it does establish is that the obstacle between here and there is an engineering wall, not a law of nature: the chemicals that protect cells from ice are themselves toxic, and the danger of ice grows fast as the thing you're freezing gets bigger. A frog is small and comes with its own antifreeze. A mouse does not. A human is enormous by comparison.

That wall is hard. It is not forbidden. And the same way no amount of wanting could build ChatGPT in 1956 because the chips didn't exist yet, the tools to climb this particular wall are being built right now, mostly by people who think they're solving an entirely different problem. If those tools work, the field may not move in a smooth line. Like AI after 2017, progress could look slow for years and then accelerate quickly.

Here is what changes if you're right about the destination and patient about the date.

Freezing and reviving a body doesn't fix anything about it. Someone frozen at fifty-three and revived wakes up still fifty-three, with every wrinkle and worn joint they went under with. That sounds like a weakness. It is actually the whole point. It means you don't have to out-run aging anymore. You can out-wait it. You go to sleep now, with the medicine we have, and you wake up later, with the medicine we'll have. It is a bridge across time to the era when everything else the longevity world is building has finally arrived.

So reversible freezing isn't really about the freezer. It's a hypothesis that the rest of the field eventually succeeds, and the technology could function as a bridge that carries a patient to that future medicine.

Which brings us to the part you actually came for. Because if this is a 1956 field, then somewhere out there is a small group of people building the thing underneath it, and the research question worth your attention is simple: who is actually building the part that's been stuck for seventy years, who's furthest along, and what would have to be proven before any of them deserved real conviction?

We've spent considerable effort on exactly that. We found a global race most people don't know is happening, including a state-backed competitor on the other side of the world that may be ahead of America on the single hardest piece. We found the one name whose technology is pointed straight at the seventy-year wall. And we wrote down the precise list of milestones, the actual events, that would move it from a name we're monitoring to a name that may deserve a higher research rating.

That's where we draw the line between what's free and what we keep for the people who fund this work.

You've just read the why. Below the line is the deeper research: the companies, the labs, and the specific milestones we're monitoring.

We don't move a name into the tracker until the evidence crosses a line we've drawn in advance. This issue shows you exactly where that line is.

[ Upgrade to read The Cold Position → ]