Essays on the far future of intelligence
← All essays The guide

Into deep time · part 3 of 3 · by Bill Guan

Stars will keep shining for a hundred trillion years, so almost all of the future is still ahead of us

The big question

How long will the universe have stars, and how far into that story are we?

The answer changes how the present looks. You're about to take a tour of deep time, from the Sun's last days to the final black holes, and meet the strategies a mind might use to stretch itself across eras far longer than everything that has happened so far.

Here's a thought I keep coming back to: whatever intelligence is around to watch the Sun die will probably be as different from us as we are from bacteria. Stranger still is how early in the story we are, because the universe is about 13.8 billion years old, which sounds ancient until you line it up against how long it will keep going.

Your guess first

Squeeze the whole age of stars, from the first ones lighting up to the last ones burning out, into a single 24-hour day that starts at midnight. What time is it now?

The last stars die in a hundred trillion years, and black holes outlast them

THE DEEP-TIME RULERyears, with every tick ten billion times longer1020 1040 1060 1080 10100 THE AGE OF STARSthe universe so far13.8 billion yearsthe last stars burn outabout 100 trillion yearsDEGENERATE ERAcooling stellar corpsesprotons decay, if they ever domore than 2.4×1034 yearsBLACK HOLE ERASun-mass black holes evaporatemore than 1067 yearsthe biggest black holes evaporateabout 10100 yearsDARK ERAlasts forever

Here's the long view, using the best estimates I could find. In a little over a billion years the brightening Sun makes Earth "uninhabitable for higher forms of life" 1, and in about 5 billion years it becomes a red giant 2, after which "planet Earth will not be able to escape engulfment" 3. By about 100 trillion years from now (1014), star formation has ceased and the longest-lived small stars burn out 4.

After that the numbers get silly. Protons decay after more than 2.4×1034 years, if they decay at all, which no experiment has seen yet 5. Black holes the mass of the Sun take more than 1067 years to evaporate, and the largest ones, with the mass of whole galaxies, last about 10100 years 6.

Fred Adams and Greg Laughlin, who mapped this out in detail in 1997, split the future into eras. We live in the Stelliferous Era, the age of stars, which ends around 1014 years. After that come the Degenerate Era of cooling stellar corpses and the Black Hole Era running out to about 10100 years, and then the Dark Era, which lasts forever 4.

Squeeze the age of stars into one day and we're 12 seconds in

THE AGE OF STARS AS ONE DAYMIDNIGHTNOON6 AM6 PMNOW12 secondspast midnightstill to come:23 H 59 MIN 48 Sthe last stars burn outat the next midnightZOOM IN ON THE FIRST 20 SECONDSso far: 13.8 billion yearsthe Sun becomes a red giant: 16 sNOW: 12 S0 s20 seach second = about 1.2 billion years

The number that gets me is this one. If the Stelliferous Era is a single day, each second of that day stands for about 1.2 billion years, which puts the universe's 13.8 billion years at about the first 12 seconds after midnight. About 0.014% of the age of stars has passed, and the other 23 hours, 59 minutes and 48 seconds are still to come.

Even the Sun's death barely moves the clock, since its red giant phase about 5 billion years from now lands at roughly 16 seconds past midnight. And the day itself is only the opening act, because once the last stars go out at the next midnight, the Black Hole Era alone runs on until about 10100 years, another 86 powers of ten further out.

A mind could stretch a fixed budget by thinking slower

A thinker trying to last into these eras runs into the problem that energy keeps getting scarcer, and people have proposed some clever strategies for dealing with it.

The first is to slow down. In 1979 Freeman Dyson showed that, in the universe as it was understood then, a mind could think forever on a finite energy budget by thinking more and more slowly and hibernating in between 7. The discovery that cosmic expansion is accelerating broke this plan, because space never gets cold enough for the trick to keep working, and Lawrence Krauss and Glenn Starkman concluded in 2000 that "life cannot be eternal," though it "can be astronomically long" 8.

Waiting for the cold could make thinking far cheaper

SLEEP UNTIL IT'S COLD?THE AESTIVATION IDEA, 2017zzZsleep throughthe warm erawake when theuniverse is cold×1030 more computingTHE REPLY, 2019think nowcold places to dumpheat exist todaywaste heatso there is no reason to wait

The second strategy is to wait. Computing is cheaper when it's cold, because the minimum energy needed to erase a bit of information falls with temperature. In 2017 Anders Sandberg, Stuart Armstrong and Milan Ćirković suggested that advanced civilizations might be lying dormant until the universe cools, since waiting "can produce a 1030 multiplier of achievable computation" 9. They called it the aestivation hypothesis, after the summer equivalent of hibernation, and offered it as one possible reason we don't see anyone.

Charles Bennett, Robin Hanson and Jess Riedel weren't convinced. They replied in 2019 that the universe today is full of places where waste heat can be dumped just as cheaply, which can be done "at any time, and is not improved by waiting for a low cosmic background temperature. Thus aliens need not wait to be active" 10.

Something stranger is possible too. A sufficiently advanced intelligence might even try to find and reach base reality, if the reality it lives in turns out to be one layer of something larger. I honestly have no idea what to do with that one, and it's a good reminder that we have very little sense of what a mind a billion times smarter than us would even want.

Boltzmann brains show how the far future can sink a theory

Deep time throws up some genuinely weird puzzles, and my favorite is the Boltzmann brain. Given enough time, random fluctuations in a cold, empty universe could in principle assemble a brain, complete with false memories, out of nothing, and some cosmological models predict that such accidental observers would eventually vastly outnumber real ones.

The physicist Sean Carroll argues that this is a reason to reject those models, since a theory that predicts you're probably a random fluctuation gives you no reason to trust the observations you used to arrive at it. In his words, such theories "cannot simultaneously be true and justifiably believed" 11. I find that a lovely example of how thinking about the far future can tell us something about the present.

Being this early makes the opening scene matter enormously

If almost all of the future lies ahead, then everything happening now is the opening scene of a very long story, and the stakes of how it starts are enormous.

Nick Bostrom made this case in 2003. He estimated that every second we delay settling our local supercluster of galaxies costs "over 1013 potential human lives," or around 1029 if minds can run on computers 12. You might expect that to argue for racing ahead as fast as possible, but his conclusion runs the other way: "a single percentage point of reduction of existential risks would be worth (from a utilitarian expected utility point-of-view) a delay of over 10 million years" 12. The lesson, in his words, is that "we ought to maximize its safety, i.e. the probability that colonization will eventually occur" 12.

My assessment: whatever you think of the utilitarian framing, the arithmetic of deep time is humbling. Hundreds of trillions of years of possible future sit on the far side of the next few centuries, so if minds last, we are ancestors in the most literal sense, and almost every mind that will ever wonder about any of this comes after us.

Sources

  1. NASA Science, "Stars." science.nasa.gov/exoplanets/stars/
  2. NASA Science, "Star Types." science.nasa.gov/universe/stars/types/
  3. K.-P. Schröder and Robert Connon Smith, "Distant future of the Sun and Earth revisited," MNRAS 386, 155 (2008). arxiv.org/abs/0801.4031
  4. Fred Adams and Gregory Laughlin, "A dying universe: the long-term fate and evolution of astrophysical objects," Reviews of Modern Physics 69, 337 (1997). arxiv.org/abs/astro-ph/9701131
  5. Super-Kamiokande Collaboration, "Search for proton decay via p→e+π0 and p→μ+π0 with an enlarged fiducial volume in Super-Kamiokande I-IV," Physical Review D 102, 112011 (2020). arxiv.org/abs/2010.16098
  6. "Hawking radiation," Wikipedia (citing Don Page, 1976). en.wikipedia.org/wiki/Hawking_radiation
  7. Freeman Dyson, "Time without end: Physics and biology in an open universe," Reviews of Modern Physics 51, 447 (1979). journals.aps.org/rmp/abstract/10.1103/RevModPhys.51.447
  8. Lawrence Krauss and Glenn Starkman, "Life, the Universe, and Nothing," Astrophysical Journal 531, 22 (2000). arxiv.org/abs/astro-ph/9902189
  9. Anders Sandberg, Stuart Armstrong and Milan Ćirković, "That is not dead which can eternal lie: the aestivation hypothesis for resolving Fermi's paradox" (2017). arxiv.org/abs/1705.03394
  10. Charles Bennett, Robin Hanson and C. Jess Riedel, "Comment on 'The Aestivation Hypothesis for Resolving Fermi's Paradox'," Foundations of Physics 49, 820 (2019). arxiv.org/abs/1902.06730
  11. Sean Carroll, "Why Boltzmann Brains Are Bad" (2017). arxiv.org/abs/1702.00850
  12. Nick Bostrom, "Astronomical Waste: The Opportunity Cost of Delayed Technological Development," Utilitas 15(3), 308 (2003). nickbostrom.com/papers/astronomical-waste/