Wait, Compute, Win?

Before asking whether civilisations aestivate, you need to ask whether any exist to aestivate. The Sandberg 2018 Monte Carlo with wide log-uniform priors shows P(N < 1) ≈ 30–50%: there may be nobody to hibernate. The “aestivation viable” scenario uses Sandberg priors because the thesis was written for precisely that prior regime — if N is small, the silence is already partly explained by rarity, and aestivation is a candidate explanation for whatever small residual exists. The optimistic preset is the scenario where the aestivation thesis is doing the most heavy lifting: if N is large, aestivation must explain not a residual but the overwhelming majority of absent civilisations — every single one went dark simultaneously, by the same thermodynamic logic. That is a much stronger claim.

Landauer’s principle says every irreversible bit operation costs at least k_B T ln 2 joules in heat dissipation. Lower the temperature, lower the cost. Sandberg et al. (2017) showed that waiting 10¹²–10¹³ yr for the CMB to cool from 2.73 K to < 10⁻³ K multiplies the irreversible-ops budget per joule by a factor of order 10³⁰. The gain is genuine and enormous — if the limiting factor is Landauer cost and if the cold future reservoir is genuinely new (not accessible today). Bennett, Hanson & Riedel (2019) contested exactly this conditional: deep space is already at ~3 K or colder today; a sufficiently large radiator can already approximate the future cold sink. If you can dissipate into a ∼3 K reservoir right now, you capture most of the gain without waiting. The reversibility fraction slider in the tool separates the two cases: near-zero reversibility means you pay full Landauer cost and aestivation gives a 10³⁰× gain; near-one reversibility means you pay almost nothing now and the waiting gain vanishes.

Once the universe has cooled to < 10⁻³ K, what can you actually compute per joule? The Bekenstein-Landauer-Lloyd tool shows three limits simultaneously: the Bekenstein bound (maximum information content of a finite region at finite energy), the Landauer limit (minimum energy per irreversible operation at temperature T), and the Lloyd limit (maximum operations per second on a physical system of energy E, from the time-energy uncertainty principle). In the cold future, the Landauer limit approaches zero and the Lloyd limit is the binding constraint. In the aestivation scenario, the payoff is the Landauer ops/s per watt: at 10⁻⁹ K, you get ~10³⁰× more operations per joule than at 2.73 K today. In the MTH scenario, you instead use the IMBH BZ power (P ≈ 10³⁶–10³⁸ W) combined with the cold-universe Landauer limit — giving an operations budget that dwarfs any aestivating civilisation purely through power scale.

The aestivation argument requires civilisations to survive hibernation for 10¹²–10¹³ years — roughly 70–700 times the current age of the universe. For the MTH to be compatible with aestivation, the IMBH must still exist after that wait. Hawking evaporation gives the answer directly: the evaporation timescale scales as M³. For the ω Cen IMBH at ≥ 8,200 M_△, the evaporation time is ~10⁸³–10⁸⁶ yr — roughly 70 to 73 orders of magnitude longer than any aestivation wait. The IMBH not only survives the wait; it is essentially eternal on aestivation timescales. The tool also shows that at this mass, the Hawking temperature is ~10⁻¹⁸ K — far below the CMB — so the black hole is currently absorbing, not evaporating. It will begin to evaporate only when the CMB cools below its Hawking temperature, which happens after the aestivation window anyway. The MTH endpoint is safe.