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

Into deep time · part 2 of 3 · by Bill Guan

Fire unlocks less than a billionth of the energy in matter, and a spinning black hole unlocks 42%

The big question

How much of the energy locked inside matter can anyone actually get out?

Every kilogram of stuff holds a staggering amount of energy. You're about to climb the ladder of ways to release it, from a candy bar to a black hole, and see how low on that ladder our whole civilization still stands.

Einstein's E = mc² is usually taught as a fact about physics, but I find it more useful to read as a fact about waste. A kilogram of anything, whether it's coal or chocolate, holds about 9×1016 joules locked up in its mass 1, and almost everything humans have ever done to get energy, from a campfire to a nuclear power plant, taps a tiny sliver of it.

Your guess first

Imagine wrapping the whole Sun in solar collectors and catching every photon it gives off until it dies. What share of the Sun's mass would you turn into energy you can use?

Each rung of the ladder runs on different physics

THE ENERGY LADDERshare of the mass that comes out as energySPECULATIVEBLACK HOLESNUCLEARCHEMISTRYBlack hole evaporation*90%A "sphalerizer"*50%?Dyson sphere around a quasar42%Spinning black hole engine29%Fusing hydrogen into helium0.7%Fission of uranium-2350.08%Dyson sphere around the Sun0.08%Burning gasoline*0.00000005%Burning coal*0.00000003%Digesting a candy bar0.00000001%×8 MILLIONmore from fissionthan from a candy bareach rung = 10× more of the mass* still to be checked against the book [TBD]

In Life 3.0, Max Tegmark lists ways to turn matter into usable energy along with how much of the mass-energy each one actually releases 2, and it's one of my favorite tables in any book. The shares run from 0.00000001% for a candy bar to 90% for an evaporating black hole, ten powers of ten apart, and every big jump up the ladder means a different kind of physics. Four rows, marked with an asterisk, still need checking against the book [TBD].

Everything from digesting lunch to burning oil sits on the bottom rung, and so does every battery. Chemical reactions shuffle electrons between atoms, and the energy involved is minuscule next to the mass of the atoms themselves. Tegmark puts the candy bar at "merely 0.00000001 percent efficient" 2, which means your lunch releases about one ten-billionth of the energy it contains.

Nuclear reactions tap the forces inside the atom

Splitting or fusing atomic nuclei taps the much stronger forces holding protons and neutrons together. Tegmark notes that today's reactors "still fail to extract more than 0.08 percent of their energy" 2, which is still about 8 million times better than a candy bar. Fusion does better again, because when four hydrogen nuclei fuse into helium, "0.028758 u, or about 0.7%, of the original mass of 4 Hydrogen atoms is converted to energy" 3 (u being the atomic mass unit), and that's what powers the Sun.

The row that surprised me is the Dyson sphere, a swarm of collectors wrapped around an entire star. Even if you surrounded the Sun with them and caught every photon until it died, Tegmark writes, "we'll still never convert more than about 0.08 percent of the sun's mass to energy we can use" 2.

Gravity does the heavy lifting around a black hole

The next jump comes from gravity, and it's enormous. Matter spiraling into a black hole heats up and radiates before it crosses the edge 4, and for "a maximally spinning black hole delivering energy at a whopping 42 percent efficiency" 2, nearly half of everything you drop in comes back out as light. Astrophysicists put the theoretical maximum at 42.3%, falling to about 30% for a more realistic spin 4. That's why Tegmark's table pictures a Dyson sphere around a quasar, the blazing core of a galaxy where a giant black hole is feeding.

A spinning black hole also stores energy in its rotation. Roger Penrose showed that you can steal some of it by sending matter in close and catching what flies back out, and the rotational energy you can extract this way tops out at 29% of the hole's mass 5.

The top rungs need physics nobody can harness yet

The last two rungs are where Tegmark gets playful. A "sphalerizer" would use an exotic process predicted by the standard model of particle physics, called the sphaleron, which can convert protons and neutrons, three at a time, into much lighter particles from the electron family 6. Tegmark describes compressing matter "to a couple of quadrillion degrees" and letting it re-expand once sphalerons "had converted most quarks into electrons and related particles" 2.

And at the very top, black holes slowly evaporate into radiation, a prediction of Stephen Hawking's. Tegmark's table credits evaporation with 90%, which assumes you could feed matter into a small black hole and catch the radiation that comes back out, something nobody knows how to do.

We use five parts in a hundred trillion of the Sun's output

OUR SHARE OF THE SUNTHE SUN'SOUTPUT3.828×1026 wattsall of it = Type IIALL OF HUMANITY1.9×1013 wattsin 2025 = Type 0.73×20 TRILLIONhow much more the Sun puts out than we use

In 1964 the Soviet astronomer Nikolai Kardashev proposed classifying civilizations by how much power they command 7. A Type I civilization was one at roughly Earth's level at the time, about 4×1012 watts, while Type II controls the output of a whole star and Type III a whole galaxy. Carl Sagan later turned this into a sliding scale with a much higher bar, putting Type 1.0 at 1016 watts 8, and judged that "our present civilization would be classed as something like Type 0.7" 8.

Updating that with today's numbers, the world used a bit over 600 exajoules of energy in 2025 9 (an exajoule is a billion billion joules), which averages out to about 1.9×1013 watts and puts us at roughly Type 0.73 on Sagan's scale. The Sun puts out 3.828×1026 watts 10, so humanity currently uses about five parts in a hundred trillion of what our own star produces, which means the Sun puts out about 20 trillion times what we use, and nearly all of our share comes from the bottom two rungs of the ladder.

Astronomers have searched the sky for Dyson spheres

HOW TO SPOT A DYSON SPHEREcollectors around a starcatch its lightwaste heat glowsin infrareda star with toomuch infrared?THE SEARCH SO FAR2024: seven candidate stars flagged2 checked by Webb,2026: galaxiesthe rest: probablydusty galaxies toopurple smudge = a dusty galaxy far behind the star

In 1960 Freeman Dyson pointed out that a civilization harvesting its star's energy would glow with waste heat, and could be spotted as a star with too much infrared light 11. That idea launched real searches, and in 2024 a project called Hephaistos combed through sky-survey data and flagged seven candidate stars 12.

Follow-up work suggested they were probably distant dusty galaxies sitting behind the stars 13, and in July 2026 a team using the James Webb Space Telescope confirmed it for two of them: "the infrared excess does not originate from Dysonian megastructures... but from background galaxies" 14. So far, nobody up there seems to have built one either.

Feeding black holes would multiply what each kilogram gives

Everything we run today, from lunch to the power grid, comes off the bottom two rungs, and even our nuclear plants extract less than a tenth of a percent of their fuel's mass-energy.

My assessment: the ladder is the best picture I know of how much room the future has. A civilization that learned to feed black holes the way quasars do would get billions of times more energy from each kilogram than fire ever gave us, and at that point Kardashev's star-powered and galaxy-powered civilizations start to look like engineering problems.

Sources

  1. Seth Lloyd, "Ultimate physical limits to computation," Nature (2000): "A one kilogram computer has average energy E = mc2 = 8.9874 × 1016 joules." arxiv.org/abs/quant-ph/9908043
  2. Max Tegmark, "Our Next Billion Years," excerpt from Life 3.0, Discover (2017-10-06). discovermagazine.com/the-sciences/our-next-billion-years
  3. Jeff Mangum, "Proton-proton cycle question," National Radio Astronomy Observatory (2023). public.nrao.edu/ask/proton-proton-cycle-question-why-is-the-…
  4. Z. Kovács, L. Á. Gergely and P. L. Biermann, "Maxima of the energy flux and of the radiation efficiency of accretion discs around black holes," MNRAS 416, 991 (2011). academic.oup.com/mnras/article/416/2/991/1056478
  5. "Penrose process," Wikipedia (citing Sean Carroll, Spacetime and Geometry, 2003). en.wikipedia.org/wiki/Penrose_process
  6. "Sphaleron," Wikipedia. en.wikipedia.org/wiki/Sphaleron
  7. Nikolai Kardashev, "Transmission of Information by Extraterrestrial Civilizations," Soviet Astronomy 8, 217 (1964), as quoted in Milan Ćirković, "Kardashev's Classification at 50+" (2015). arxiv.org/abs/1601.05112
  8. Carl Sagan, The Cosmic Connection (1973), as quoted in Ćirković (2015), same link as [7].
  9. Energy Institute, Statistical Review of World Energy 2026, as reported by DieselNet (2026-07-02). dieselnet.com/news/2026/07energyreview.php
  10. Andrej Prša et al., "Nominal values for selected solar and planetary quantities: IAU 2015 Resolution B3," Astronomical Journal 152, 41 (2016). arxiv.org/abs/1605.09788
  11. Freeman Dyson, "Search for Artificial Stellar Sources of Infrared Radiation," Science 131, 1667 (1960), as quoted in Jason Wright, "Dyson Spheres" (2020). arxiv.org/abs/2006.16734
  12. Matías Suazo et al., Project Hephaistos II, MNRAS (2024). [TBD: full title] arxiv.org/abs/2405.02927
  13. Tongtian Ren, Michael Garrett and Andrew Siemion (2024). [TBD: full title] arxiv.org/abs/2405.14921
  14. Erik Zackrisson et al., JWST follow-up of Dyson sphere candidates (2026-07-10). [TBD: full title] arxiv.org/abs/2607.09460