ARPA-H and XPRIZE Push Longevity Science Toward Measurable Function

Latif Nasser framed the central problem as a double question: how people can live long, healthy lives, and how institutions can change the world so that more people can do so. Scientific research is one answer, he said, but not a clean one. In medicine and public health, ambitious ideas are surrounded by uncertainty; hope and promise often become “false hopes and red herrings.” Since time, money, and expertise are limited, the hard work is not simply funding research. It is deciding which claims deserve a serious bet, then structuring that bet so that success can become usable by real people.

The two models approached that problem from opposite ends of the funding process. Alicia Jackson, director of ARPA-H, described a government agency built to push money, milestones, and technical management into high-risk health problems early. Jamie Justice, who leads the health domain at XPRIZE, described a prize model that pulls teams toward a defined outcome by placing a large reward at the end. Nasser summarized their complementarity with an analogy from Justice: in a marathon, ARPA-H may be running beside teams and operating the water stations, while XPRIZE is waiting at the finish line with the trophy.

The target in longevity science, as Justice described it, is not immortality or a single disease. It is the gap between lifespan and healthspan. She said people are now living “about 30 years longer than we ever have,” but with a roughly 12-year period of living unwell inside that longer lifespan. In the United States, she put the healthspan gap at roughly 10 to 12 years; globally, she described it as roughly 10 to 15 years. XPRIZE Healthspan is therefore not asking teams to prove that a favored drug or protocol works. It is asking them to compress that gap.

Justice said the prize defines the challenge as a restoration of health and function: up to 10 years, with a goal of 20 years, across muscle, cognitive, and immune function. Those domains were chosen because they matter to people and because they map onto the regulatory language of clinical benefit: function, feeling, and survival. The prize does not limit teams to a mechanism. Drugs, biologics, gene therapies, lifestyle interventions, exercise, diet, and other approaches can enter, if they meet the same evidentiary frame.

That frame is meant to be solution-agnostic. Justice contrasted it with a grant model in which funders might support a specific metformin trial with a predefined primary outcome. XPRIZE begins instead with the gap to close, defines the start and finish line, and lets teams decide how to get there. In the Healthspan competition, she said, 790 applicants registered globally. The field was narrowed to a top 100 the previous year, when XPRIZE “gave out the first 10 million of 101.” This year, she said, XPRIZE is awarding the next $10 million for 10 teams to begin one-year trials. More teams can still enter by self-funding.

Justice emphasized that this openness changes who can participate. In the top cohorts, she said, XPRIZE saw “everyone from Harvard to a high school team in Malaysia,” publicly traded pharmaceutical companies, and early startups. Her point was not that all of them are equally likely to win. It was that she does not know in advance where the good ideas will come from, and the competition model is designed not to gatekeep the source of an answer.

Alicia Jackson traced ARPA-H’s operating model to DARPA, where she worked earlier in her career. At DARPA, she said, she saw ideas move from a laboratory investment of roughly $250,000 to a fielded national-security capability within five years. The appeal of that model was not only the money. It was the combination of a mission, autonomous program managers, technical scrutiny, and the ability to increase or stop funding as evidence emerges.

ARPA-H, she said, applies that model to health. The agency is only three years old, and its mandate is to pursue health challenges that are too large, complex, or risky for traditional funding or industry. Jackson described the program managers as the center of the model: technically deep people from academia or industry who enter government for limited terms, usually three years. They are asked to define the major problem they want to solve, explain how they will solve it, state how much money they need, and move fast enough to matter before their term ends.

The funding relationship is not a conventional grant. Jackson said ARPA-H does not “throw anything over the fence” and wait for a paper. In her words, “If you come to publish a paper, you’re not for us. All we care about is impact.” If an idea fails, ARPA-H stops the funding. If it works, ARPA-H can double down.

The pace is part of the proof system. Jackson said a traditional NIH program might take a year to launch. In ARPA-H’s model, a program manager can develop an idea, survive internal scrutiny, announce it within roughly three months, receive proposals two to three months later, and launch in less than six to nine months. The agency then works closely with teams and holds them to milestones. Jackson acknowledged that “not everyone likes to move at this pace” and that it can be difficult to hit milestones while doing science. But she argued that the model works because it refuses to let hard problems drift.

Her case study was an ARPA-H sleep program led by a program manager named Nate. The problem begins with something common: many people sleep poorly, and poor sleep is tied to deteriorating mental health, cardiovascular health, metabolic health, and chronic-disease risk. Jackson said current drugs may technically help people sleep but can be bad for the brain, and many lack durable benefit. The program manager then questioned the metrics themselves: consumer devices such as an Oura ring, Whoop, or Apple Watch rely on sleep measures rooted in a roughly 60-year-old framework, while the science of sleep has advanced toward finer structures that matter for health and brain function.

The proposed program is not a new sleeping pill. It aims to measure the quality of sleep at home in much finer detail and intervene in real time while the person is asleep. Jackson mentioned possible neurostimulation, vibrational stimulation, and peripheral technologies. The goal is that when sleep begins to degrade, a device could send a signal that improves sleep quality without a drug intervention.

The ARPA-H call then defines constraints before teams pitch. Jackson said the program manager tells the research community what he wants and what he does not want: no new drug, a technology that works at home, a defined price point, and clear requirements around data, partnerships, technology building, and verification. Teams can form around the problem, but they must accept the pace and the milestone structure.

Jamie Justice described XPRIZE’s version of the same discipline from the opposite end. XPRIZE spends less time gatekeeping who gets to try and more time defining the parameters for success. The competition must be “prizable”: someone must be able to sign up for it, the endpoints must be measurable, the judging framework must be rigorous, and the result must matter to the global challenge being addressed. If the start line and finish line cannot be defined, or if finishing would not change the problem, the idea is not appropriate for the prize model.

That does not mean weak ideas are filtered out by credentials in advance. Justice said the model allows people to struggle, fail, and demonstrate whether they succeed. The failures are still useful if they generate data. She called that accumulated evidence “compost for future innovation and design and development.” Centralized data collection is part of that logic: even if an intervention does not win, its data can be aligned with outcomes and endpoints from other programs and used to inform the next stage.

The institutional through-line is proof before preference. ARPA-H tries to force evidence early through active program management, milestone funding, and rapid decisions about whether to stop or double down. XPRIZE tries to force evidence at the finish line through predefined judging, centralized data, and a prize that cannot be won by a story alone.

When Nasser asked what looked most promising in longevity, Alicia Jackson began with the less glamorous answer: existing drugs and molecules. She said there are “a ton of molecules and drugs already out there,” with some already FDA approved. She named GLP-1s, SGLT2 inhibitors, metformin, rapamycin, and “a whole range of drugs” as examples in the broader landscape, and said candidates in this area have shown activity across a variety of model organisms. The problem, in her account, is not necessarily that science must invent entirely new interventions. It is that the field still needs human trials capable of testing whether these kinds of candidates can matter for longevity and healthspan.

The bottleneck is trial design. If the endpoint is death or disease decades later, no one will fund a 20- or 30-year trial. Jackson said ARPA-H and XPRIZE teams are both focused on biomarkers that can project forward: measures that would allow a three-year trial to say something credible about what might happen 20 years later. She did not present a single biomarker as settled. Instead, she described the biomarker problem as central infrastructure for the field.

She also rejected the idea of a single longevity cure. “I don’t think there’s going to be a silver bullet,” she said. Her expectation is a cocktail: multiple beneficial interventions combined in ways that preserve or restore function.

Then she moved to the more radical bet: replacement. Jackson described “replacement theory” in longevity as the idea that some tissues and systems degrade with age and do not adequately regenerate themselves. She gave bone marrow transplant as a familiar example of replacement already in use: if the immune system degrades, replacement is at least conceptually part of existing medicine. The same lens can be applied, more speculatively, to the heart, brain, and other tissues.

ARPA-H is funding what Jackson called a “really crazy moonshot idea” around replacing brain tissue. The agency is spending $180 million to derive induced pluripotent stem cells from a person, grow neocortical tissue that is structurally and functionally correct, and transplant it into damaged areas of that person’s brain. Jackson called the work “wild” and “very, very hard,” but said there is “no technical reason this can’t work.”

Jamie Justice said that kind of replacement logic overlaps with ideas emerging from XPRIZE’s brain-trust process. She described convening groups and asking for the ideas people are almost afraid to say aloud. Some are in development; others are killed early. One group is looking at biostasis, including possibilities such as cryopreservation for organs or even whole mammals. Another is focused on replacement organs, but from a different starting point: drug development and toxicity testing.

Justice’s point was that current models are limited. Organoids and single-system models can test certain things, but they do not capture multi-system signaling. Animal models are imperfect because, as she put it, “we’re not big mice.” A multi-organ system could help researchers understand toxicity or drug effects in a more integrated way, perhaps eventually contributing to organ replacement. That is not yet presented as the immediate clinical application. It is an enabling platform that could unlock better testing and, possibly, later replacement uses.

The tension in these examples is part of the field itself. Some of the more practical opportunities may involve drugs and molecules already known to medicine or aging research. Some of the more radical bets may sound like science fiction. Both Jackson and Justice argued that the work has to be organized so those ideas can be tested rather than accepted or dismissed on atmosphere alone.

Longevity attracts hucksters, Nasser noted, and some credible work can sound nearly indistinguishable from fantasy when heard casually. His question was how the two institutions prioritize ideas and distinguish serious science from bogus claims.

Alicia Jackson said ARPA-H hires people who are not only technically deep, but “technical athletes” able to move across fields. Their job is not to personally have the best idea; it is to find the best idea. That means talking broadly across a field, looking for ideas that go against dogma, and asking what would unlock an entire domain so that later capital can flow from public funders or the private sector.

ARPA-H also asks what is already well funded. Jackson named New Limit as an example of a major private bet in the longevity space; if others are already funding a path, ARPA-H may look elsewhere. Sometimes, she said, the agency may fund a hard test simply to determine whether a path works or not, so the field does not spend the next 20 years pursuing the wrong direction.

Program managers must pass a rigorous internal process. They are asked: What is the big idea? Why does it matter? Why has it not been done? Where is the field already going? If the field will get there in five years without ARPA-H, why should ARPA-H fund it? What are the hardest problems, and why is there any reason to think they can be solved?

The hardest problems are not hidden. Jackson said they are often placed at the front of the program as early milestones, so the agency can down-select quickly. Her preference was to spend $10 million and fail fast rather than spend “$100” and not know until the very end.

Jamie Justice described a similar filter in the prize model. XPRIZE focuses on how to ask a question rigorously, how to be thorough, and how to make scientific practice more transparent. The same burden applies whether a team is testing a gene therapy, a biologic, a repurposed drug, exercise, diet, or another lifestyle intervention.

That framework is also how Justice separates a prize-worthy problem from a product pitch. Some proposals begin with a company’s technology and then invent a problem around it. XPRIZE turns those down. Some companies ask the foundation to run competitions in areas where their own R&D has stalled; Justice said XPRIZE will not become outsourced corporate R&D, especially since teams retain their own intellectual property and commercialization paths.

Other problems are not primarily technological. Some health burdens are equity or access problems where the solutions already exist but do not reach people. A prize may not help if teams cannot get to the start line, if the finish cannot be measured, or if finishing would not make a meaningful dent in the global challenge. Justice put the constraint plainly: the foundation is not in the business of wasting philanthropic money or time.

In both cases, the answer to hype is not simply skepticism. It is a structure that decides in advance what would count as progress, then forces the claim to meet that standard.

The word “moonshot” was useful but contested. Jamie Justice works at what she called “kind of the moonshot company,” but she treated the term as less about literally going to the moon than about defining a challenge hard enough to change behavior. She cited the first XPRIZE more than 30 years ago: a competition to create a manned spacecraft that could fly 100 kilometers with three people and repeat the flight within two weeks. Since then, the model has been less about the moon itself and more about the vigor of asking people to sign up for something very difficult.

For Justice, a good moonshot is not an incremental improvement or a shelf-ready solution in search of a problem. It is set far enough out of reach that teams stop asking what they can do with the tools already in their hands and start asking what must be built, combined, or reconsidered. She described it as being in water deep enough that “your toes are not touching the ground anymore.”

That distance is the point. A dramatic endpoint can draw radical solutions, new technologies, new collaborations, and even regulatory headaches. Justice argued that moonshots can bring public attention into science, pull researchers across silos, and help unstick fields that are under-resourced or too fragmented for one company, one grant, or one country to solve.

Jackson agreed with the underlying ambition but was wary of the term. Alicia Jackson said ARPA-H avoids “moonshot” because in government it can become a label for bureaucracy rather than breakthrough. A supposed moonshot can turn into “12 different agencies with working groups and committees.” ARPA-H prefers the DARPA phrase “make a dent in the universe,” and Jackson said the agency wants to stay nimble, autonomous, “a little rogue, a little cowboy.”

Still, she said the effect on scientists can be profound. Many researchers have been trained to think only about the next incremental step. When asked what they would do with $5 million, they can answer. When asked what they would do with $10 million, they may stretch. When asked what they would do with $50 million to $100 million, many cannot answer immediately because no one has ever asked them that question in a serious way. The process can free scientists and builders to think beyond themselves, collaborate across disciplines, and imagine work that conventional academic funding does not support.

The useful version of a moonshot, in both accounts, is not grandeur for its own sake. It is a device for changing the feasible set: the collaborators people seek, the evidence they agree to generate, the regulatory problems they confront, and the scale of result they are willing to define as success.

Lisa Treviño, from the public health company DLH, pressed both speakers on whether patients are involved from the beginning. Scientists can become excited by ideas, she said, but the ultimate stakeholder is the person who needs the intervention. If patients are left behind, she warned, time, money, and lives are jeopardized.

Jamie Justice said patient and consumer voice has to be built into the definition of success, not added after the technical work is done. In the healthspan space, she said XPRIZE works with partners, including AARP, to understand what consumers and patients value. The foundation is also in discussions with groups such as FNIH around multistakeholder work groups that include patient voice.

Justice linked that input directly to the choice of endpoints. Before XPRIZE, when designing and leading trials, she said she used stakeholder and focus-group work with patients in the community to ask what mattered if researchers were going to “go after aging.” That work pointed toward function. She said a survey conducted about two months earlier in preparation for an FDA meeting found that more than 90 percent of scientists also said patient voice is hugely important. Across patients and scientists, the priority was function first: people care more about health and function than about a specific disease model of aging, mortality, or lifespan.

Alicia Jackson gave a concrete ARPA-H example from the agency’s lymphatics portfolio. In programs aimed at illuminating the lymphatic system and curing lymphatic diseases, she said ARPA-H is deeply engaged with patient advocacy groups, including the Lymphedema Society. Researchers have been paired with patients and are required to meet regularly. Jackson described that pairing as a major unlock because it helps researchers understand the lived impact of the disease.

Justice added that patient voice also has to be global. For interventions meant to address worldwide health challenges, the question is not only whether a solution works in one setting, but whether it scales and whether voices from different contexts shape its development.

A clinical psychologist in the audience asked for bioethical examples and mitigation strategies. Jamie Justice answered that some risks can be mitigated, while others have to be accepted as inherent to ambitious work. Her emphasis was on not hiding them. XPRIZE, she said, creates venues for critique: public comment, road shows, advisory input, and open discussion of second-order effects. The question is not whether advancement creates challenges. It does. The question is how programs structure constraints and invite scrutiny around the risks they create.

Justice said she has two bioethicists on her scientific advisory board: one who is future-positive and excited, and another who “hates the field.” She tells both to follow the process and publish without constraints. The adversarial function is intentional.

Alicia Jackson said ARPA-H has a dedicated function inside the agency for these questions because it is both a government body and an organization pursuing breakthroughs. The agency asks directly what headline it does not want to see in the Washington Post, then works backward: how to be transparent about what a program is and is not, how to build ethical guidelines, when to bring in bioethicists, and where to put guardrails.

Jackson drew on her DARPA experience in biology, where the combination of military and biological research carried obvious risks. Some ARPA-H programs require especially careful boundaries. She mentioned work on disordered or misfolded proteins, noting that prion disease immediately raises a different set of concerns. The agency must be clear about what it is developing, what it is not developing, who has been involved in review, and how the technology will be used. Her broader point was that technology can be used in many ways, so transparency and guardrails are not optional.

A later audience question brought the ethics problem into geopolitics and commercialization. The questioner asked how global collaboration can survive a fracturing geopolitical landscape, IP constraints, strategic concerns, and the tendency of VC-backed companies not to share.

Justice answered from the structure of the competition. The Healthspan prize is global, and she said the forthcoming top 10 teams are split evenly: 50 percent US and 50 percent non-US. At least one team is from a country where data and human samples cannot be moved in or out. There is no realistic prospect, she said, of solving that by putting people at a table and expecting regulatory harmonization.

So XPRIZE holds to its standards and pushes the burden back to teams. To compete, teams must agree to data-use and sharing terms and material-transfer agreements. If their home context does not allow samples or data to move, they must find a way to test their solution outside the country so it can be evaluated rigorously and comparably.

The prize also has to deal with regulatory arbitrage. Justice said XPRIZE is tracking US-based teams that go abroad to test gene therapies, cell therapies, or other interventions because they cannot get domestic regulatory approval. She mentioned places such as Honduras and the Bahamas as examples of where teams may choose to test. The foundation then has to decide who verifies the work and how standards are upheld while tolerating some risk.

The goal, she said, is equal viewing: the huckster and the publicly traded pharmaceutical company should face the same burden of proof. A central lab and harmonized data system are meant to make that possible. The data should be useful not only for awarding a prize, but for regulatory change, clinical practice, and a public evidentiary standard that applies even to “the bros you see on TikTok.”