Embryos Survive 30 Years Frozen. Until Wants Organs to Stop Racing Private Jets

Sarah Guo's childhood wish list included telepathy, mind upload, and the ability to freeze time for humans. She did not think that last one was a job. Laura Deming, cofounder and CEO of Until, is building it anyway. The company's long-term goal is reversible whole-body cryopreservation for medical hibernation: take someone on their deathbed and pause them until a critical cure exists. Near term, the work is more concrete—reversibly cryopreserve single human organs so transplant patients get them without a race against the clock.

The cold open already contains the scientific frame Deming will unpack. Water expands when it becomes ice, and tissue does not take that well. There is a temperature below which ice formation stops; get under it without crystals and you are safe for a long time. Kids have already been cryopreserved as tiny embryos for 30 years, then used to start pregnancies. The question is not whether the trick works at all. It is whether it scales.

Deming is obsessed with problems that feel as if they should be worked on and are not. Growing up, longevity seemed the obvious case. Aging is not fully recognized as a disease, so the work does not look valid, for classification reasons more than technical ones. You can already extend lifespan in many organisms; how far that goes in humans is unknown. Guo, as a school-aged child, had not even registered aging as a concept to consider.

Deming grew up homeschooled in New Zealand, without a normal biology class. As a kid she hypothesized that everyone lived to ten and then died on schedule, which made her father about eight and her mother about seven. The striking discovery was that we do *not* all hit a hard age and stop. Nobody knows what determines how long we live. If death at ten were certain, longevity would feel like an immovable limit. The uncertainty is what made the mechanisms worth chasing—through lab work, MIT, and a Thiel Fellowship.

The Longevity Fund came from taking researchers literally. Ask people in the field for the bottleneck and most said they could not get enough funding. As a teenager she decided to get money to the projects that should have it; venture capital was just the later name. After that first stretch of work, cryopreservation hit her as one of the best problems she had ever seen: counterfactual impact, raw technical delight, and a socially weird reputation. Seeing it clearly was zero-to-one. She could not imagine pouring the next decade into anything else.

Cofounder Hunter Davis is the person who comes back the next day with an answer built from first principles. Most people writing a cryopreservation primer would cite the literature. He re-derived fundamental laws of statistical mechanics as the foundation of what you should know. On their first call Deming pitched the idea and he said little; later he admitted he did not buy it. He spent a couple of days on ice-formation facts and back-of-the-envelope math, then returned: it seemed in the realm of possibility in a way that violated his first intuition.

Until's goal, as Deming frames it, is a new form of critical care. Metastatic melanoma is the anchor case. In a single year, combination immunotherapies moved expected survival from six to nine months to roughly half of patients living past a decade. Single years decide whether a patient dies of the illness they have or lives long enough for the critical cure. There is currently no way to press pause on biological time. She wants an ambulance of the future: hibernate someone on their deathbed until the therapy exists. Early use is not decades; it is covering a clinical-trial window. Davis's father-in-law got advanced cancer treatable by a therapy that arrived a couple of months after he was no longer eligible.

Whole-body reversible cryopreservation is the long-term technical implication. Guo asks about other uses—skipping ahead to see the future, going to Mars. Deming's limit is social, not technical. People define themselves by the people around them. You cannot take everybody with you. Recreational time-shifting asks you to put your current social reality on the line for an uncertain elsewhere. Mars may happen with or without this technology. In her mind the method only makes sense at first for extreme cases, where you would literally die without it.

Ice does not slam on at a temperature like a switch. Formation is stochastic: random nucleation, then extension. That is technically friendly, because you can modulate those rates and therefore the probability of ice. Spend as little time as possible in the nucleating window and you have a shot at almost none. Until focuses on avoiding ice rather than preserving *with* some crystals. Ice is bad because water expands and tissue cannot take that without substantial membrane damage.

There is a temperature below which ice formation stops—Deming's traverse from 0°C down through that line, which cryobiology treats as around −130°C. Get below it without ice and you are good for quite a long time. Human embryos have been cryopreserved for a record of over 30 years, then rewarmed and used in viable pregnancies. We already reversibly cryopreserve human tissue, including whole tiny bodies of a few hundred cells, routinely. From first principles, pausing all molecular motion and randomly restarting it used to seem crazy even at the scale of one cell. We know it works. Guo's reaction is the civilian one: wait, scientists tried it and it succeeded, so the problem is scaling.

Scaling means vasculature, mixed material properties, perfusion of chemicals in and out, and getting heat in and out fast enough. The cryobiology community has already moved the organ-scale question out of the wilderness. Greg Fahy did foundational vitrification work. John Bischof's group, and Fahy before him, showed you can reversibly cryopreserve a kidney, rewarm it, put it into a rat that has no other kidney, and watch that rat return to normal function after about a month. At mammalian-organ scale, Deming says, the problem is academically tractable *now*.

Guo visited Until's lab and was startled by how much it looked like engineers asking how to warm something quickly and safely, not a therapeutic program. Deming has tried for a long time to explain why that is the point. You can trade engineering difficulty and biological difficulty to a non-zero degree. Not 100 percent: biological questions remain, and some answers may come out negative. But better tooling makes the biology easier, and that leverage is rare in biology.

Temperature is the conceptual lever she is obsessed with. Physicists spent centuries inventing it: one measurable that compresses molecular motion, almost the relative passage of time at the nanoscale. Tune temperature and you can apply a physics toolkit that is actually useful. Almost no other biological problem lets you do that. Concretely, cooling and rewarming rate changes how much cryoprotective agent you must add to suppress ice. You want that concentration low because CPA is toxic. Faster rates, less time in the danger zone, less chemical. Instantaneous thermal control would need none; that is not the default for a large organ, so tissue toxicity never fully disappears. Guo's lab example is the everyday version: if an organ will not rewarm evenly, change the heater's surface-to-volume ratio or redistribute heat. New devices, same biology.

The company runs two tracks in parallel: scale preservation and rewarming to human-organ size, and whole-rat reversible hibernation, moving kidney lessons into the animal and doing rat-specific work. Deming would not put a near-term date on full-body reversible cryopreservation. The roadmap is clearer than when they started, and the first steps have been faster than she imagined. The giant unknown is the brain: what injury a whole-body protocol would inflict, and what fidelity is possible. That is still open neuroscience.

Recruiting against an unclear timeline on a huge scientific goal is easier when the first product is obvious. Transplant patients regularly miss organs that are already en route because of timing. Organs expire quickly after a donor becomes available—one of the most precious resources we know, Deming says, burning in hours. Hospitals charter private jets to put a surgeon on a plane, pick up the organ, and race it back. The recipient waits within a two-hour radius of the transplant center with a pager, always. Until's first product is pausing time for the organ so you can take as long as you need. That is near-term: into studies and the clinic as fast as possible.

The organ program is also a falsification test. If you talk about whole-body reversible cryopreservation and cannot dent organ logistics, there is no version of the long-term path that skips this gate. If you cannot do it, you are not the company for the long-term thing. Deming likes having a clear benchmark of whether the technology is tractable at that scale. Claiming whole-body is 100 percent certain would make them bullshitters, and they would not recruit. What they can show is the model: why they think it is in scope, and what they are testing. Strong people start skeptical, because intuition says it should not work, then they see data that violates that intuition and start reversing from first principles.

Asked how transplantation looks if Until works, Deming goes to the patient's life. You wait for a life-changing surgery with no scheduled date, under what Guo calls surgery-center house arrest. Matching happens at the last minute: someone dies, an organ appears, staff call around for whoever can get to the hospital. A transplant surgeon they work with put it in one sentence: making time not a variable changes the whole paradigm. Surgeons would not have to operate the same night they flew. Everyone in the ecosystem would have time for the best decision. Think carefully about how the system works now, she says, and it is amazing that everyone is operating this way.

If organ-scale work is already academically tractable, why was it starved? Deming's read is resources. Reversible organ cryopreservation could not attract the capital required to scale. She gives a lot of credit to pioneers, especially Greg Fahy, who fought to make vitrification something people took seriously through a long stretch when it was not a thing to focus on. Whole-body cryopreservation is still hard to discuss rationally. People either love it so much they will not question it, or they fear the science-fiction taint. She dances around what she calls an anti-meme. Her old venture instinct was to pick exactly those spots: something that feels weird, hard to think about, with no matching technical barrier. Longevity used to sit there. Cryopreservation sits there now.

Longevity became more mainstream over the last five years, academically and among consumers, and Guo notes those are linked. Deming hopes cryopreservation gets the same gift: a next-generation professor who chooses neural tissue, or whole-body-adjacent work, without shame or fear of losing a grant. If people can work on it in the open and someone will pay, funding follows. Asked for other neglected problems, she answers aesthetically. She wants to represent the molecular world so people can actually engage with it—colorful, kindergarten-like, not the flat triangles and arrows in a textbook that make no conceptual sense. If that is educational, fine. The reason is that it is beautiful, like seeing a tree for the first time: complex, fractal, light falling through the branches. That video of the world is not accessible to most people.