Grand Futures Reading Group | Session Digest

Grand Futures Reading Group

A 7-month journey through Anders Sandberg's vision of humanity's cosmic potential

📄 View full session notes (Google Doc)

25+
Sessions
7
Months
22
Chapters
12+
Participants

Participants

Algon (host) Tim Duffy Tim Babb Tim P Mateusz Thebes Sichu Abel Mariven Rauno Niplav Chlorokin

Curated Insights

"There's no form of PTSD for pleasure. You don't wake up at night thinking 'I love my child.'"
— Mariven (Sep 7)
"Stone is more durable than technology. If you've got the raw stuff left over, you'll have a stronger chance of cultural usage sticking around."
— Tim Babb (Jul 6)
"Parfit's argument: the best is world peace, second best is 99% of humans die, third best is 100% die. The gap from 1-2 is much smaller than 2-3."
— Sichu (Nov 16)
"Aestivation makes dark forest more plausible. You'd want to go dark for a while."
— Tim Duffy (Jul 20)
"You've got 2 generations to create galaxies before you're part of the permanent underclass."
— Abel (Aug 24)
"Carl Shulman said Tasmanians lost many cultural innovations as population decreased."
— Niplav (Jul 6)
"If you're uncertain about the future, either your discount rate is low enough that the future is huge, or high enough that it doesn't matter at all."
— Tim P (Aug 3)
"Strongly in favour of eating bacteria. Swedish company already working on this."
— Mateusz (Jul 6)
"Suffering is definitely socially mediated. Especially the longer in time it occurs over."
— Thebes (Sep 7)
"Don't think Bitcoin's gonna work."
— Chlorokin (Aug 3)
"A small moon (~100km) is as big as you can make a unified brain with a speed of thought as fast as a human's."
— Algon (Jul 27)
"Worst experiences are probably more intense than the best. But I strongly intuit that people would choose positive experiences more often on a 1-10 scale."
— Rauno (Sep 7)
"Why care about human physiology? Because it won, which is unsettling."
— Mariven (Nov 16)
"Long wondered if civ got started because population got big enough for very smart/agentic people to be born."
— Tim Babb (Jul 6)
"Haldane drank dilute hydrochloric acid to test blood acidification... and one time crushed his vertebrae testing oxygen saturation."
— Sichu (Dec 14)
"Diamond's gravitational stability: hundreds of GPa! I'm more pro diamond and more pro carbon after this section."
— Tim Duffy (Jul 27)
"Optimism is exceptional, not the default state."
— Abel (Dec 14)
"Perhaps streams are an underrated cultural transmission tech."
— Niplav (Jul 6)
"Seems kinda sad."
— Tim P (Aug 3), on reaching a state of perfect happiness
"Population bottleneck theory: Some men gained consciousness and then women only wanted to mate with men who were conscious."
— Mateusz (Jul 6)
"Like organoids? You seen that startup doing brain organoids?"
— Thebes (Jul 6)
"How much value is there in sending messages to successor civilizations? I'd assign very little. Like, what's the value of Romans taking over the universe instead of us?"
— Chlorokin (Jul 6)
"If panspermia is correct, then a bacteria is the yet-smallest existence proof of a self-reproducing machine. Only downside is it takes a few billion years."
— Algon (Oct 26)
"10% of supermassive black hole mass is from dark matter."
— Rauno (Sep 28)
"I feel like when you get to intergalactic levels, things are law-of-large-numbers predictable enough that you can have 'causal pacemakers'—disconnected groups that accurately predict each other's choices."
— Mariven (Aug 24)
"Maybe we'll get dual conveyor belts of asteroids. But we'll probably be in a simulation anyway."
— Tim Babb (Jul 6)
"We should care about the whole genetic landscape of entities that come after us. So we should care about potential phylogenetic trees."
— Sichu (Nov 16)
"From a utilitarian perspective, at the first jhana you've made it."
— Tim Duffy (Sep 7)
"Far-mode thinking is hard to sell. Colonizing Jupiter? To the average voter, nationalism made it easier—Soviets to beat."
— Abel (Dec 14)
"Tasmanians did become disgusted by fish, according to Claude."
— Niplav (Jul 6)
"Limiting factor [for geothermal] is bore holes though."
— Tim P (Oct 5)
"There exists a size where the total number of operations per second begins to decline as more computronium is added. Time dilation grows more than linearly."
— Mateusz (Jul 27)
"After the glacial period, there was a point where some areas saw sea levels rise by 1 foot/day, so plenty of sites just got washed away."
— Thebes (Jul 6)
"CDs are pretty good. Very dense." — Tim Babb
"But only for a century because of disc rot." — Chlorokin
— Exchange on data durability (Jul 6)
"Art won't exist because there will be better ways to communicate mental states."
— Algon (Dec 14)
"Any spatially collocated civilization will almost certainly eventually be wiped out by a quantum fluctuation. It'll be like 10^10^200 years."
— Mariven (Aug 24)
"Minimal viable population was also interesting. Anders said something like 5000 for the number to sustain genetic diversity."
— Tim Babb (Jul 6)
"Your reach should exceed your grasp. That becomes the motivation."
— Sichu (Dec 14)
"My dream energy source doesn't require moving parts."
— Tim Duffy (Oct 5)
"Maximize the education vector—just get educated, idiot."
— Abel (Dec 21)
"Really into this. Ooh, it's written in Forth!"
— Niplav (Jul 6), on Collapse OS
"You need some level of selfishness to persist. But past that point, you don't need to in order to achieve most goals."
— Mateusz (Oct 19)
"There was also a guy encoding poetry into bacterial genomes in a way that would force it to be conserved. The 'xeno-text'."
— Thebes (Jul 6)
"Bitcoin is hosed because it isn't quantum resistant. Anything that is signed is completely hosed."
— Chlorokin (Aug 3)
"Fremlin's estimate is crazy and I love it. 500 stories w/ 5m² per person would be 10% of the earth's population."
— Algon (Jul 6)
"The brain doesn't care about truth inherently. It only cares about prompting useful actions. So your brain is a constant liar."
— Mariven (Sep 7)
"Illiteracy was one of the best forms of cognitive security."
— Sichu (Jan 4)
"I think that the degree to which the long-term future is controllable is maybe fairly modest."
— Tim Duffy (Jan 4)
"Surprising that we haven't unlocked new co-operation modes at ever larger scales."
— Abel (Nov 16)
"My notes were just a bunch of surprising facts. 1k-20k effective population of Neanderthals in Europe at any point in time."
— Niplav (Jul 6)
"If one galaxy brain is sufficient for any problem, that's probably a reason to copy rather than merge."
— Thebes (Aug 24)
"I expect these abstractions break down before the limits, as you reach a state of perfect happiness."
— Chlorokin (Aug 3)
"A 5 page description of 'the shape of the physics of all future warfare' seems like it's going to be wrong no matter what."
— Algon (Jun 29)
"Large groups are slow, democratic and non-extreme. So you're never going to get a democratic vote for how the far future happens."
— Mariven (Oct 26)
"Wouldn't you want ecological diversity? Such a mind would collapse."
— Abel (Aug 24), on merging galaxy brains
"Reminds me of Paul Christiano's idea to send minimal message to successor civilizations if humanity might die out."
— Niplav (Jul 6)
"You could have a maximal computation band before it degenerates to noise. So there's a sweet spot of computability before collapse."
— Abel (Dec 21), on Tipler point
"I suspect suffering is always relative to a cultural baseline. A stoic society would have a very different baseline."
— Abel (Sep 7)

Open Questions

Big unresolved questions about grand futures that emerged from our discussions:

🌌 Cosmology & Physics

  • Dark matter utilization: How exactly do we collect and efficiently convert dark matter into usable mass-energy?
  • Dark energy extraction: Can we exploit vacuum energy or the Casimir effect for practical purposes?
  • Proton decay: If matter decays faster than expected, how does this affect long-term planning for civilization survival?
  • Black hole computation: What are the actual limits of using black holes for computation in our universe?
  • Vacuum decay game theory: If vacuum decay is possible, what's the optimal strategy? Can civilizations coordinate to warn others?

🧠 Computation & Intelligence

  • Jupiter brain limits: What are concrete upper/lower bounds on ops/s given cooling, gravitational, and thermodynamic constraints?
  • Heat dissipation: What is the maximum size of a "lump of computronium" given cooling constraints?
  • Conscious mind size: Are there arguments that limit the size of a coherent conscious mind?
  • Moon-sized brain psychology: What's the attentional bottleneck of moon-sized brains? How would it affect their social structures?

🤝 Coordination & Civilization

  • Acausal coordination: How does decision theory work across causally disconnected regions?
  • Relativistic discount rates: Are discount rates generally covariant in curved spacetime?
  • Offense-defense at cosmic scales: How does the offense-defense balance shape civilization at different scales?
  • Long reflection feasibility: Is a multi-generational pause for reflection possible without civilizational rot?
  • Value drift: How do you maintain coherent values across billions of years?

👽 Aliens & Fermi Paradox

  • Aestivation rebuttal: What's the response to Hanson's argument that aliens shouldn't wait to be active?
  • Are we in an alien domain?: If civilizations reshape cosmology, could our "natural" observations be artifacts of alien engineering?
  • Calendar warfare: What are the detailed game-theoretic implications of wormhole networks?

⚖️ Ethics & Values

  • Manipulating the universal prior: Can civilizations shape what strategies are optimal on a cosmic scale?
  • Buddhist utilitarianism: Has anyone integrated jhana states with utilitarian moral calculus?
  • Opposite of negative utilitarianism: Why does "pleasures having priority over pain" receive almost no attention?
  • Encoding morals: How do you prevent extreme moral drift across civilizational cycles?
  • Post-scarcity failure modes: What about wireheading and other ways abundance could go wrong?

🪞 Identity & Consciousness

  • Qualia in posthumanity: The extensional definition of posthumanity is missing qualia entirely.
  • Identity over deep time: Clock measurement limits imply bounds on how long any identity can persist.
  • Substrate dependence: If computational functionalism is false, how does this change the picture?

Session Digest

June 29, 2025 Appendix B: War
"A 5 page description of 'the shape of the physics of all future warfare' seems like it's going to be wrong no matter what." — Algon
First session. Discussed scorched Earth strategies as credible deterrent, Lanchester's laws, and whether warfare physics can be modeled with ironclad rules. Tim confused about convex/concave dynamics. Niplav learned focusing lasers is hard.
July 6, 2025 Ch. 6: Food & Population
"How do we know Tasmanians actually lost fire?" — Algon, sparking debate
Eating bacteria, synthesizing fats, and Fremlin's 500-story population estimate. Argued about Matrix population inaccuracy (off by 10^6). Thebes noted Twitch deleting archival streams is bad for cultural transmission. Tim Babb: minimal viable population is ~5000.
July 13, 2025 Ch. 15: Posthumanity
"Matrioshka brains probably aren't post-human. Too inhuman." — Mateusz, contested by Algon & Niplav
Abiogenesis theories and the clay hypothesis. Definitional debates: posthuman vs transhuman vs trans-species successors. Tim Babb: "Interesting question is what you do after human beings are no longer bound by technological limitations."
July 20, 2025 Ch. 16: Computation
"Aestivation makes dark forest more plausible. You'd want to go dark for a while." — Tim Duffy
Alkanes for info storage, reversible computing. Thebes: universal aestivation is unlikely since "everyone wanting to go dark at once" seems improbable. Hanson's estimate: nearest grabby alien civ 1-10 billion years away.
July 27, 2025 Ch. 16: Computation (cont.)
"Have you guys heard of the Lofstrom loop?" — Tim Duffy (Mateusz: "Nope")
Diamond material science, gravitational stability (hundreds of GPa), and time dilation limits on computronium. Discussion of moon-sized brains and their attentional bottlenecks. Mateusz intrigued by thermodynamic energy cost being 10^-6 of biosphere energy.
August 3, 2025 Ch. 17: Economics
"Don't think Bitcoin's gonna work." — Chloro
"Given your track record, that means it must work." — Algon
Interest rates vs discount rates, Piketty, and uncertainty arguments. Supermoney and physical ledger limits. Aestivation economics: sleeping to gain interest, but rent payments equilibrate. Mateusz: "Fully automated gay luxurious space communism."
August 10, 2025 Ch. 18: Far Future
"Loved his name for it: calendar warfare." — Algon, on wormhole temporal conflicts
Wormhole empires, causality collapse, and economic value of living in a wormhole network's core. Tim Babb drew a diagram. Thebes: "If wormhole empires were possible, we should have seen one already." Tim: "Or we're in the anthropic shadow of it."
August 17, 2025 Ch. 18: Far Future (cont.)
"Vacuum decay is really speculative physics, so I'll worry about it when we know more." — Tim Babb
Tidal power from redshifting, baby universes, and moving between Everett branches via wormholes (Tim Babb's idea). Game theory of cooperation: multi-cellular life as evidence, but it involves throwing away individual identity.
August 24, 2025 Ch. 10: Intergalactic Settlement
"Any spatially collocated civilization will almost certainly eventually be wiped out by a quantum fluctuation. It'll be like 10^10^200 years." — Mariven, on extinction timescales
14 rounds of messages between disconnected galaxies. Should civs push galaxies together? Mariven's "causal pacemakers" concept—psychohistory at cosmic scales. Abel: "You've got 2 generations to create galaxies before you're part of the permanent underclass."
August 31, 2025 Ch. 11
"Could we even communicate much with a grabby civ or get taken over?" — Tim Duffy
Brief session. Discussed how far grabby civs could reach while taking us over, and the difficulty of staying coordinated as we splinter into alien-to-each-other civilizations.
September 7, 2025 Ch. 20: Values
"From a utilitarian perspective, at the first jhana you've made it." — Tim Duffy
Deep dive: negative utilitarianism, jhanas, hedonic set points, and the asymmetry of pain vs pleasure accessibility. Thebes achieved first jhana. Mariven on the brain as "constant liar." Can we manipulate moral facts or the universal prior?
September 21, 2025 Ch. 5
"Anders focuses too much on stuff that's irrelevant in the face of superintelligence." — Mariven
Short session. Mariven noted cities differ from organisms (flat, interconnected logistics). Discussion of what matters given superintelligence.
September 28, 2025 Ch. 7: Resources
"Did we talk about the synthesis of black holes?" — Thebes
Star lifting (2×10^15 kg/s), dark matter collection, and stellivores—star-eating organisms! Proton decay vs Standard Model debate. Only 10% of supermassive black hole mass is dark matter. Hawking radiation for energy conversion.
October 5, 2025 Ch. 7: Resources (cont.)
"So how many supernova charges can you get from a rotating black hole?" — Algon [Laughter]
Tim's dream energy source: no moving parts. Black hole bombs (wikipedia link shared). Extending baryonic matter lifetime via gravitational time dilation—maybe 10^20 extra years? Abel: "We're on the fringe of new physics."
October 12, 2025 Energy & Entropy
"Chapter was mostly a nothing burger. Cool facts, but missing bits on entropy, which is the core fact determining the future." — Mariven
Divergent views on the chapter. Tim found entropy sections dry. Discussion of dark matter/anti-matter annihilation, fermionic dark matter, and Casimir effect exploitation. Disappointment about dark energy extraction limits.
October 19, 2025 Ch. 3.5: Coordination + Appendix D
"Are discount rates generally covariant?" — Mariven (Algon: "Probably not.")
Neutrino capture with neutron stars, neutronium rings (unstable). Cipolla's theory of stupidity. "Parochial" became a recurring word for the book. Tim: we should view the whole project through offense/defense balance at different scales.
October 26, 2025 Ch. 8: Space Settlement
"Once you have VN probes you immediately disable your competitors on earth no matter what. It's just a violent takeover." — Mariven
Industrial DAGs for space settlement, cement in vacuum, mammal reproduction in space (basically untested). Theia: surprisingly pessimistic chapter on space settlement. Mariven and Tim Duffy debated path dependence vs game-theoretic takeover scenarios.
November 2, 2025 Ch. 20
"Say we learnt the universe was 300 billion years old—how'd that affect your intuitions?" — Mariven
Utility functions and symmetry requirements. Time/space translation invariance in morality. Mariven: time can be operationalized as causal order, not a coordinate we intrinsically care about.
November 9, 2025
Short session. Notes minimal. Reading assigned for next week: Chapter 3, issues facing current civilization.
November 16, 2025 Ch. 3: Current Challenges
"We should care about the whole genetic landscape of entities that come after us. So we should care about potential phylogenetic trees." — Sichu
Sichu on Parfit and potential lifeforms. Mariven: "Why care about human physiology? Because it won, which is unsettling." Is the "general factor of coordination" BS? Abel surprised Moloch section is so large. Capitalism as Moloch-restraint.
November 23, 2025 Ch. 4: Post-Scarcity
"Is the West already post-scarcity? C.f. Gwern: 'you can live on $1000/year'" — Algon's notes
New "basic needs": orbital plates, body morphology, self-encryption, termination rights for copies. Sichu: wireheading might be reasonable if threats are gone. Abel: but there's always adversarial threats. Galaxy-brained scheme for conscious beings in Malthusian equilibrium of unconscious ones.
November 30, 2025 Ch. 4 (cont.)
"Predicting the far future can be easier than the medium term. Timing is one of the hard parts." — Discussion notes
Brief session. Extinction known before 1800, HG Wells and soft sci-fi. Good predictors focus on frequencies, not dates—imprecision in timing is a feature, not a bug.
December 14, 2025 Ch. 21: Implications
"Even if machines are better, you still need to find meaning in something anyway." — Sichu
Creativity in post-human futures. Art without functional purpose becomes signaling—or signaling to yourself. Haldane drank dilute hydrochloric acid for science and crushed his vertebrae testing oxygen saturation.
December 21, 2025 Ch. 21: Implications (cont.)
"Maximize the education vector—just get educated, idiot." — Abel, on what we can pass to future generations
Value drift, empathy machinery, and high-bandwidth brain connections dissolving identity. Abel: education about historical possibilities is the best we can do for uncertain futures. Being informed is objectively best.
January 4, 2026 Ch. 22: What To Do — FINAL SESSION
"Scientists don't say 'based' enough." — Sichu, on academic communication
Final chapter! Skepticism about Long Reflection. Tipler point physics—infinite energy but infinite density too? Why did optimism vanish in the 80s? Passing formalisms vs passing values. Rauno returned from Canada, Mariven's New Year's was uneventful.