The personal forecast
It is a race between two clocks: how long the science takes to arrive, and how long you have. But nobody knows when it arrives, so we do not pretend to. We infer a whole probability distribution over the arrival year from real data, then race it against your own survival. Every number below is derived and moves as you change your age, your assumptions, or as the evidence itself changes.
The headline: pooling a model of the science with the forecasting community, escape velocity is more likely than not to arrive this century (67%), with a middle estimate around 2075. Whether you see it is a race against your own clock, so age matters most. It all waits on a first therapy proven to slow human aging, the milestone this site tracks. A transparent model, not a promise.
Will it arrive this century?
Pooled from a model of the science (mouse rejuvenation, then human trials, then compounding gains) and the forecasting community. The other 33% is the chance it comes later or the science fails outright. It hinges on a first therapy proven to slow human aging.
Your odds, if you are in the race
A scenario-weighted estimate, not a measurement: the chance you are alive whenever it arrives, across the whole distribution. Survival band 44%–49% (frozen to fast mortality improvement).
The forecast, inferred not assumed
We do not pick a year. We infer a distribution over every year, then let you read it. It pools a mechanistic base-rate model of the science, the Metaculus community distribution with named experts, and a small fails-outright share. If it happens at all, the middle estimate is around ~2075.
If it arrives at all, the scenario-weighted middle is ~2075, with the middle-80% band spanning ~2035 to after 2100. A distribution under deep uncertainty, not a date.
Are you young enough?
This isolates the part that depends on you. Whether escape velocity arrives at all is the chart above; whether you live to see it is a pure survival race to that year. At this setting, anyone 37 or younger is more likely than not to be there.
Your three possible outcomes
Every future splits three ways, and they sum to one. You reach it, or it arrives but after your time, or it comes later than this century.
Move the assumptions
Default : the share of the forecasting community’s probability on “never”, shrinking once a hallmark has a lever proven to slow human aging. Click to reset.
How fast age-specific death rates keep falling, as a Social Security Trustees low/central/high band on the SOA MP-2021 curve. Currently: Trustees' best estimate.
How the forecast is built
Two computations. First, a probability distribution over the arrival year, pooled from three referenced sources. Second, your chance of surviving to each of those years. Multiply and sum.
A small share for the chance no therapy ever works well enough: set to the 12% of the forecasting community’s probability on “never”, and shrinking once any hallmark has a lever proven to slow human aging. You can set it yourself above. The slower cases of the other two arms (long trials, slow compounding, the skeptics) make up the rest of “not this century”.
A structural chain: a first robust mouse rejuvenation (~2028–2040), then a translation lag to a proven human therapy (a lognormal fit to real drug-development base rates), then compounding. Integrated over a weighted grid.
The Metaculus community distribution (median ~2050, a quarter on not-this-century) plus named experts across the spectrum, from Kurzweil and de Grey to the skeptics, each counted once.
D(Y) = linear pool of the three arms · P(reach) = ΣY D(Y) × P(survive to year Y)
A linear opinion pool is used on purpose: it keeps the disagreement between sources instead of averaging it away into false confidence. Survival is the SSA period life table with an age-specific improvement band. The three outcomes (reach, arrives too late, not this century) sum to one by construction.
It is a scenario-weighted estimate, not a probability of record. The arrival distribution is a structured judgment under deep uncertainty, not an empirical sampling distribution. Read the numbers as “model-based degree of belief under stated assumptions,” and note that long-horizon forecasts are systematically overconfident, so if anything the true spread is wider.
It is conditional on an unproven event. The concept, actuarial escape velocity, is Aubrey de Grey’s (PLoS Biology, 2004). Every positive number assumes a first robust human rejuvenation is proven and then compounds. Neither has happened; the triggering milestone is not yet met even in mice.
A known caveat we do not hide. The starting rate (0.24 yr/yr) is record life expectancy at birth, whose early rise came largely from children surviving. At older ages the rate is lower today but has been speeding up since 1950, because gains have moved to the ages where aging acts. We keep one starting rate for consistency with the rest of the site and show the by-age rates on the index page.
The survival side is a US, period-table average. It understates a real person’s longevity (the improvement band and the ~7-year period-to-cohort gap show why) and ignores your individual risk factors. It reports a population figure, not your prognosis, and reaching escape velocity also assumes you keep receiving each future therapy.
It is educational, not medical, financial, or life-planning advice. Nothing here is a recommendation.