STUDIO TORIUMI / OBSERVATORY日本語
When humanity ends is more or less settled. It's fixed by physics, so there is nothing to argue about. That isn't the problem. The problem is that plans to avoid it barely exist. What is certain, what is moving, and what has stalled — laid out with numbers and sources.
Neither of these is a "might happen." Both are a question of "when."
The Sun is slowly brightening. As it brightens, rock weathering accelerates and carbon dioxide drains from the atmosphere. Once CO₂ falls too far, plants can no longer photosynthesize, and the oxygen supply stops.
The Earth that follows is a world rich in methane, poor in carbon dioxide, and without an ozone layer. The only survivors will be organisms that don't breathe oxygen.
By this estimate, the stretch of Earth's life during which the planet can hold on to its oxygen amounts to only 20 to 30 percent of the whole. We are already in the later part.
Kazumi Ozaki & Christopher Reinhard, "The future lifespan of Earth's oxygenated atmosphere," Nature Geoscience (2021)
The Sun will expand to somewhere near Jupiter's orbit, then shrink to roughly Earth's size and become a white dwarf. The inner planets will be lost.
This is not just calculation. In 2021, the Keck Observatory caught a star system with a Jupiter-like planet orbiting a white dwarf at roughly Jupiter's distance. The inner planets were gone. In effect, we have seen the future solar system.
University of Tasmania et al., Nature (October 2021) / W.M. Keck Observatory
The thin line at the left edge is the roughly 300,000 years Homo sapiens has existed — 0.028% of the total. The time remaining until the deadline is about 3,600 times the whole history of our species.
Talk of "human extinction" tends to blur settled physics together with talk of probabilities. Keeping them apart lets you be properly afraid of each.
The moving ones and the stalled ones, side by side in the same list.
In September 2022, NASA's spacecraft deliberately crashed into the moon of an asteroid, and genuinely changed its orbit. The first time humanity has altered the motion of a celestial body.
ESA's Hera launched in October 2024 and arrives on site in 2026. Its job is to survey the aftermath in detail and turn this method from "a one-off experiment that worked" into "a technique you can use again."
This is the one area producing real results. Though the only threat it guards against is asteroids.
The plan was to send tiny probes to Alpha Centauri at 15–20% of the speed of light. Of all the roads out of the solar system, this was the most famous.
In September 2025, an indefinite halt was reported. Of the planned $100 million, about $4.5 million had been spent. The papers from Phase 1 remain.
No flagship program aimed "beyond the solar system" is running today.
It really is moving forward. But Mars and the Moon both orbit the same Sun. When the Sun expands, they go with it.
Settlement works against planet-wide disasters, war, and asteroids. It is not an answer to the deadline itself. Plenty of articles conflate the two, so keep them separate.
Not a single program on Earth is working on the death of the Sun.
The missing funding isn't indifference. A billion years sits too far outside the units of human decision-making. Elections run on four years, corporate plans on five; the longest national undertakings span decades. Humanity does not yet have a mechanism that can handle a number like a billion years.
The real problem is not the Sun. It's that we can't handle timescales of that length. And that one is a problem of institutions, not technology — so work on it can start today.
Written only as far as the facts allow, without exaggeration.
Last collected:
Where this comes from. The official RSS feeds of NASA, NASA Science, and ESA's space safety and space science teams. Filtered on keywords touching asteroids, planetary defense, interstellar flight, and exoplanets. It's automatic, so items only loosely related may slip in.
Individual voices come and go, so only official organizational pages are listed. Working down from here is, in the end, the most reliable way.
Last checked: August 31, 2026. Timescales vary from study to study. The figures here are exactly as the sources state them. Don't settle for "it is said that" — always trace claims back to the primary papers. This is the subject where exaggeration spreads most easily.