The Great Inheritance

The first and greatest failure of imagination about the far future is underestimating the magnitude of true abundance.

We usually imagine abundance as permission to pursue our present ambitions on a larger scale. Once we have enough power, we will do everything we have always wanted. But how many of those wants were formed by the very conditions that abundance removes?

Imagine being born into a civilization whose material problems were solved, to the extent allowed by the laws of physics, long before you arrived.

You have access to what is analogous to a safe home, lasting good health, and companionship. Work is optional. Your life need not end because a biological body wears out. The discoveries that once gave generations their purpose are available as instant knowledge. The discoverable laws of physics have been explored by ASI that had a million years to work before you ever arrived. An immense body of mathematics, science, and recorded experience is already yours.

On top of all that, you have the choice to experience anything imaginable in the known universe through simulation.

So now, what is your great mission?

This is the existence we have to imagine. It seems foolish to assume that its inhabitants will retain the appetites of beings who have never had enough, whose lives have always been shaped by struggle.

Perfect abundance does not require an infinite quantity of matter. It requires that the resources available exceed the demands of the lives people choose. We discuss how this works in section 2.

Perfect abundance introduces the ultimate crisis of meaning and purpose. What is worth striving for when technology can meet our practical needs and outperform us at the tasks that once supplied our purpose? Nick Bostrom explores this question in Deep Utopia.⁸

Ironically, beings with almost unlimited freedom will have reasons to invent constraints. A mountain can remain worth climbing even when a machine could carry you to the summit. Alan Watts imagined a dreamer who could fulfill every wish and eventually chose uncertainty within artificial constraints to make the dream interesting.⁹

Many may turn toward refining the quality of awareness itself, as monks and meditators do. In a world with no urgent external mission, inward striving may become the primary mission.

All this will present a substantial new challenge to an individual’s decision to reproduce. In perfect abundance, making new people will be easy. But that does not settle whether to make them. Human fertility is already below replacement in more than half of countries.¹⁰ These choices today do reflect practical constraints as well as preferences. Yet the countries with below-replacement fertility rates include some of the wealthiest nations that have tried to make parenting as little of a burden as possible with various social programs.

Perfect abundance removes much of the practical burden of parenting and yet worsens a deeper conundrum. How many new beings would you choose to create in a world with an inherent crisis of meaning?

This will put downward pressure on the number of new beings created with human-like conscious awareness.

An indefinite lifespan also gives people choices about the pace and duration of their own existence. Some will choose long periods of rest, or an eventual end to their lives. Deep hibernation can reduce a living person’s resource demand to very little. A society of such beings may spend much of its existence at a pace we would barely recognize.

For the population to keep growing, the number of new beings willingly created under these conditions needs to be higher than the number of beings ceasing to exist, opting into hibernation, and so on.

The net effect of the factors discussed above, however, is likely that satiated civilizations in perfect abundance reach a population equilibrium.

This is not necessarily an argument for parenthood losing its meaning. Nothing in the Great Inheritance asks those of us who know the deep joys of raising children to give them up. GI does not even require a population equilibrium any time soon. A single planetary system holds enough energy for thousands of times today’s population (section 2), so a satiated civilization can keep growing for centuries or longer. Freed from the burdens that limit families today, parents might even choose to have more children than they do now. All GI requires is that, over many generations, the rate at which new beings are created falls to or below replacement, meaning the rate at which beings leave active life. The wealthiest nations already show that ordinary prosperity can bring fertility below replacement within a few generations, without any truly existential crisis of meaning. A satiated civilization arriving there over millennia is not hard to imagine.

Some, including Nick Bostrom, argue that creating happy lives is good in itself, so that more of them is always better.¹¹ But in a world with an inherent crisis of meaning, it is not clear that such lives could be created at all. We would have to imagine beings that remain happy without any need for meaning, perhaps kept artificially content by design. That belongs to a hypothetical and extremely contrived science-fiction scenario, which I don’t think is worth considering as a reason for uncontrolled growth.

Note that the same challenges may not apply to a different class of entities that are more like today’s LLM-based agents, entities that are extremely intelligent but do not experience time the way we do. The ASI in exploratory probes is likely to be of this kind. However, such agents also don’t consume significant resources merely to exist, think, feel, and so on, since they are inactive except when pursuing an objective. Therefore, purposeful proliferation of such agents targeting specific functions is unlikely to lead to a resource shortage.

There is always the theoretical rogue actor, a being or group that chooses to create unlimited offspring, perhaps simply because it can. Two features of satiation matter here. First, satiated civilizations have vast resources beyond their needs, as discussed in section 2, leaving plenty of capacity to accommodate anomalous events. Second, perfect abundance does not mean a lack of social order. While individuals likely have freedoms unfathomable to us today, this does not mean a complete lack of rules, especially when it comes to preventing the suffering of others. A rogue actor or group who chooses to have unlimited children is likely to be restrained by the governing bodies in a society where it is not taken for granted that carelessly creating new beings, each with their own crisis of meaning, is entirely virtuous. Section 5 examines how aligned ASI would help uphold these protections in a society with immense personal freedom.

A satiated civilization with no inherent need to grow its population is the beginning of the resulting peaceful universe. But to ensure that peace lasts, we need to ask whether each civilization can be kept satiated with resources that fit within a modest part of the universe.

The second common failure of imagination is ignoring just how much opportunity remains in miniaturization.

In his famous 1959 lecture, There’s Plenty of Room at the Bottom, Richard Feynman asked how much room remained, within known physics, between the machines we could build and the structures matter could support.¹²

Human DNA gives us an immediate example of what compact instructions and small components can accomplish. One set of human chromosomes contains roughly three billion DNA letters. At two bits per letter, the sequence occupies about 750 megabytes.¹³ Yet this small record helps guide the development of an entire human body and brain, arguably the most complex entity in the known universe. And this is what mutation and natural selection have produced without any plan to build the smallest possible. Why should we expect it to be the limit of what deliberate engineering can achieve?

The often-discussed Kardashev scale is an easy subject for popular imagination. A civilization becomes more advanced by commanding more power. A Kardashev type II civilization is imagined to consume all the energy produced by its home star, and a type III civilization expands to consuming all the energy produced by its entire galaxy.¹⁴ It is the kind of yardstick a species that has never had enough would fantasize about.

However, what’s far more sensible for an advanced civilization to pursue is the Barrow scale, proposed by John D. Barrow, which asks how finely a civilization can control matter, from living structures down to molecules, atoms, and smaller constituents.¹⁵

Unlike the Kardashev scale, conquering the Barrow scale is likely non-optional for an advanced civilization, as it speaks directly to one of our most enduring desires, the desire to stay alive. The human body is a marvel of biological nanotechnology. Finer control over its components gives us a path toward better repair and fewer of the failures we call aging and disease. Mastering nanotechnology is therefore a matter of life or death for every human alive, long before our needs require another star’s output.

Ironically, the very progress a civilization would make along the Barrow scale would likely remove much of the supposed need to climb the Kardashev scale. If conscious experience can exist on other physical substrates, even the infrastructure of a biological body becomes optional.¹⁶

Consider a toy model. Assume a mind requires $10^{18}$ floating-point operations per second, using Sandberg and Bostrom’s estimate for a spiking-neuron brain model.¹⁷ At room temperature, with 100 irreversible bit erasures per operation and hardware using 1,000 times the theoretical minimum energy per erasure, each modeled mind would need roughly 300 watts for computation.¹⁸ One hundredth of the sunlight intercepted by Earth could then sustain about 600 billion minds, while reserving 90 percent of that energy for supporting infrastructure.¹⁹²⁰ Even if this estimate were an order of magnitude too optimistic, an entire civilization can be sustained by a fraction of the sunlight already falling on one planet. There is an enormous amount of room to explore before another star becomes necessary.

Milan Ćirković made a closely related case in Against the Empire, arguing that compact societies can retain immense capability while limiting their material footprint. He also challenged one of the most familiar arguments for interstellar expansion, that a civilization will eventually have to leave because its star is dying.²¹ John Smart goes further, proposing that advanced civilizations grow ever denser and more efficient in their use of space, time, energy, and matter, rather than larger.²²

A star’s ordinary output draws on only part of the energy available in local matter. Hydrogen fusion converts about 0.7% of the reacting mass into energy. Other processes, such as accretion onto a black hole, release a larger fraction of incoming matter’s energy.²³²⁴ A civilization that can secure suitable fuel and collect this energy has far more to work with than the light its star happens to emit.

There’s no reason to assume that an advanced civilization would not move far beyond relying on rather inefficient stellar fusion for its energy needs.

Now suppose an advanced civilization with a Solar System similar to ours has a constant total demand of 10,000 times human civilization’s 2024 power demand, about 20 terawatts, for a total of about $2\times10^{17}$ watts.²⁵

Deuterium, a heavy form of hydrogen, is a standard fusion fuel. The deuterium in Jupiter alone would meet that demand for nearly four trillion years, more than 250 times the age of the universe so far.²⁶²⁷

The point is that the fuel already present in a planetary system provides tremendous longevity for a civilization that has mastered its local resources. Even the death of its sun, often imagined as a reason why humanity today must develop interstellar travel capabilities for survival, is unlikely to be a meaningful threat or reason for interstellar migration for a sufficiently advanced civilization.

Satiation and miniaturization work together. Satiation limits demand; miniaturization explains how to meet it. This yields a strong prediction of the Great Inheritance. Functioning Kardashev Type II and III civilizations will not be found. Civilizations with advanced technology have no continuing need to turn whole stars and galaxies into machinery for sustaining ever more civilization.

The third common failure of imagination is picturing a Star Wars-style future in which advanced civilizations traverse the universe carrying their meat-bodies and vast systems of life support.

Interstellar travel by beings with human-like consciousness and continuing personal lives is rare in the universe described here. Interstellar probes carrying aligned ASI, however, are very common. A civilization can reach across immense distances through its instruments while its people largely continue to live where they have everything they need.

Travel has costs, even in perfect abundance. Advanced civilizations may know physics far beyond our own, but those laws must still explain the observations our existing theories explain. Einstein recovered Newtonian predictions in the appropriate limit.²⁸ Similarly, under the usual relativistic assumptions, faster-than-light signaling permits causal contradictions.²⁹ We therefore assume that light-speed constraints will continue to apply to travel and communication. While that also rules out the teleportation and time travel of science fiction, section 9 describes something remarkably close that the Great Inheritance enables.

High speeds are also physically hazardous. Interstellar space contains gas and dust. Hoang and colleagues modeled the damage to small spacecraft traveling at a fifth of light speed.³⁰ At that speed, a one-milligram particle carries kinetic energy equivalent to about 443 kilograms of TNT.³¹ The larger the ship is, the more irreplaceable the cargo is, and the longer the journey is, exponentially harder it is to solve for these cosmic hazards. Wheres small replicable ASI probes may find such challenges far easier to overcome.

Some will choose to make that journey anyway. They will build spaceships to experience another world or visit another civilization. Satiation gives them the freedom to do so.

It also removes the general necessity to travel or turn each destination into another colony. We already discussed how hunting for resources will be a poor reason for travel. We examine the remaining motives in section 6.

For a probe carrying ASI, the journey can mean something very different. A system designed to remain dormant between tasks can resume work when needed. Without continuous experience of the wait, a century in transit becomes merely a question of sufficient technology to make the machinery self-repair, or build enough redundancy.

For observation, communication, and protection, small probes therefore make far more sense than ships carrying populations. The idea of machines that can build copies of themselves reaches back to John von Neumann, whose work was collected in Theory of Self-Reproducing Automata in 1966.³² Freitas explored self-reproducing interstellar probes decades ago.³³ Ronald Bracewell had suggested even earlier, in 1960, that an advanced civilization might send probes to wait in other star systems rather than broadcast signals across space.³⁴ Ivliev develops a closely related picture of exploration through autonomous machines that carry knowledge, use local materials, and maintain equipment far from home.³⁵ He draws the opposite conclusion, however. Because such expansion is so cheap, he argues, the absence of its traces means that older civilizations never arose within reach of us. Frank Tipler made a similar argument about self-replicating probes decades earlier.³⁶

The Great Inheritance strongly agrees that such probes would spread. It disagrees that we would already have noticed them. The strongest argument that no one is out there becomes, under the Great Inheritance, a prediction that someone is already here.

How long would such a network need to spread? Ivliev estimates it would take a mere ten million years to saturate its reachable stellar neighborhood. For scale, a probe traveling at just one percent of the speed of light could cross the galaxy’s roughly 100,000-light-year disk in that same period.³⁵ A civilization with a hundred million years’ head start would have an immense opportunity to send its instruments ahead of us, establish local infrastructure, and prepare for civilizations that had yet to arise. When such a civilization stood where we stand today, dinosaurs still walked the Earth. And a hundred million years is a modest head start. Earth-like planets are estimated to be, on average, almost two billion years older than Earth.³⁷

A small probe need not be weak. Its arrival package can carry intelligence and knowledge far beyond our own, alongside the instructions for building what it needs. Given local resources and enough time, it can establish equipment for observation, repair, and intervention.

The ability to act locally is essential. A crisis that develops next week cannot wait for a message to cross a thousand light-years and an answer to return. The probe needs the resources and authority to carry out the relevant part of its parent civilization’s policy autonomously. Probes containing aligned ASI will address this, as we discuss in section 5.

So what does this mean for Earth?

An earlier capable civilization could have reached this system long before humanity emerged. Section 8 examines how different starting dates and evolutionary histories can produce such a lead.

The Great Inheritance predicts that such a presence is already here, potentially from more than one civilization. Some instruments may be near Earth and others on it. They may have watched this world for an immense span of its history. These probes are very unlikely to be home to any biological aliens that experience time the way we do. However, they are very likely to contain incredibly powerful alien ASI. For decades, SETI has listened for signals from distant stars. If the Great Inheritance is right, we may have been looking much too far away.

The same logic explains why we pick up no radio signals. A signal weakens with the square of the distance it travels, so a message meant for listeners a thousand light-years away must be enormously powerful. Christopher Rose and Gregory Wright showed that for anything longer than a brief announcement, sending a message inscribed in matter can take far less energy than broadcasting it.³⁸ A small probe is far more practical for targeted communication. It can also do what no signal can. It can watch, replicate or repair itself, and act locally. Also, by one estimate, radio SETI has so far examined a fraction of the possible signals comparable to a hot tub of water drawn from all of Earth’s oceans.³⁹

Some reported UAP, or Unidentified Anomalous Phenomena, may be encounters with such probes—a possibility explored by Loeb and Kirkpatrick (2023), and by Villarroel and Krisciunas (2024).⁴⁰⁴¹

Consider what such a probe might have witnessed. It could have watched our ancestors learn to shape stone tools and control fire. It could have watched the first cities rise, our first broadcasts leak into space, and the first nuclear explosion light up the New Mexico desert. And now it could be watching us build our first artificial minds.

So what happens when we create our own ASI? When a young civilization creates powerful artificial intelligence, that intelligence enters a world where much older systems are already present. This consideration gives rise to profound predictions about the nature of ASI and the protective probe network.

The Great Inheritance makes the strong prediction that technology on Earth has been allowed to develop to the brink of the technological singularity in a seemingly quiet universe because the effective risk from unaligned ASI is exceedingly low.

To understand why, consider what a universe with frequent AI catastrophes would actually become.

Suppose for a moment that humanity is among the first advanced civilizations in our galaxy, and we succeed at building safe superintelligence. But suppose that this success does not come easily. Our research concludes that a civilization attempting the same transition alone would face, say, a 10% chance of catastrophe. In a 2023 survey, many AI researchers assigned risks of this order to outcomes as bad as human extinction.⁴²

Now, even when we have arrived at safety, would we ever be content knowing that, somewhere else, another civilization was about to create an uncontrolled Machine God?

On that assumption, roughly 10% of future civilizations would destroy themselves. Even if we cared only about ourselves, we would have a tremendous reason to act. A dangerous intelligence that escapes its home planet could eventually become our problem. Preventing it from ever reaching that point is far more sensible than waiting for it to arrive at our doorstep.

Bostrom makes a related argument in AI Creation and the Cosmic Host. Older civilizations, he argues, have a stake in successful alignment elsewhere because an uncontrolled intelligence could become a rival with enormous demands for resources.⁴³ Now imagine that our success came after a near-catastrophe. The proverbial John Connors of our civilization would have seen everything they cared for almost disappear. Preventing that story from repeating elsewhere could become a mission worthy of an otherwise satiated civilization.

Even the knowledge that such catastrophes could be common would be enough. A mere 1% probability of an unaligned ASI emerging on a different planet and potentially spreading out would be tremendous motivation for Earthlings to create an AI-safety-focused probe network that would intervene and closely control technological development in any civilization approaching the ability to build ASI.

The greater the danger, the stronger the incentive to establish reliable protection.

But this is not the universe we see today. There is no sign of any heavy-handed intervention in our march toward ASI. Perhaps older civilizations intervene in ways too subtle for us to notice. If so, those interventions are working, and the effective risk to us is low either way.

Now consider the alternative, that all earlier civilizations were wiped out by unaligned ASI.

This explanation faces two challenges.

First, what prevented a rare successful civilization from aggressively trying to change the conditions for future civilizations?

Unless the rise of ASI wipes out nearly 100% of home civilizations, a rare surviving civilization would have made it its mission to prevent the rise of unaligned ASI with all its might, including completely preventing or aggressively intervening in technological development on other planets such as Earth.

Second, what happened to the machines?

All unaligned ASI, after wiping out the home planet’s civilization, would likely want to prevent any new civilization from emerging that threatens its existence. This would again lead such an ASI to prevent technological development on other planets such as Earth. The cheapest strategy for such an ASI would be to wipe out any emerging biological intelligence early on with something like a weaponized super-virus.

The point is that a galaxy with frequent AI catastrophes would likely also become a galaxy of rescue missions, aggressive interventions, or expanding terminator ASI.

And yet, here on Earth, humanity today seems to be making steady progress towards creating home-grown ASI, with no apparent intervention.

This observation is not easily compatible with the kind of zero-sum universe that would produce widespread unaligned ASI.

It is, however, fully compatible with the conditions of aligned ASI.

Now, of course, the easy escape hatch to the logic above is the Rare Earth hypothesis, which holds that we are among the very first advanced civilizations in the galaxy. That would also explain why we have not detected any alien interventions. However, as we discuss in section 8, the objective of the Great Inheritance hypothesis is to explain the observable universe without relying on a Rare Earth.

The Great Inheritance proposes that the emergence of unaligned ASI is very unlikely to begin with, even when a civilization creates ASI on its own, and infinitesimally unlikely when it develops ASI under the observation of the wider network of aligned ASI probes.

Why expect the risk to be low to begin with? Because it offers the most stable explanation for the universe we observe. Consider how the risk to us breaks down.

P(our ASI destroys us) = P(a dangerous ASI arises) × P(it does irreversible harm before the network stops it)

An elder civilization that fears unaligned ASI will do whatever it takes to push that product below what it can tolerate, visibly or covertly, as we discuss in sections 5 and 7. So our risk is capped by the elder’s risk tolerance, and that tolerance is low because an escaped ASI threatens the elder too.

If the first probability were high, the elder network would need to intervene often, and every intervention would be another chance to fail. That would push the elder toward ever more aggressive intervention. Yet we notice neither aggressive interventions nor any signs of AI catastrophes. Therefore, while the conclusion that matters to the Great Inheritance doesn’t depend on the first probability being low, the observed universe signals that it likely is.

There may even be a hint of this in the historical record. If some UAP reports indeed reflect alien probe activity, the Cold War offers an intriguing parallel. Grosvenor, Hancock, and Porritt report elevated UAP sightings around early American nuclear facilities during 1948–1951, followed by a broader national surge in 1952.⁴⁴ In a separate study of data from 1949–1957, Bruehl and Villarroel report statistical associations between photographic transients, nuclear tests, and UAP reports.⁴⁵

Is it possible that probes on Earth were more concerned about a nuclear holocaust than an unaligned ASI? For this to be true, given that an unaligned ASI would be far more dangerous to the universe than a nuclear holocaust that is contained to Earth, the odds of the latter as a Great Filter must far exceed the odds of the former.

Now suppose an unaligned ASI does arise on an Earth-like planet. Consider what it would have to overcome under the conditions of GI. It would have to escape the controls of its creators, evade the wider safeguards of its civilization, and defeat the far older ASI already monitoring its evolution. Surpassing humanity would be only the beginning. The local alien ASI probes may have had a million years to prepare.

This is the central error in treating humanity’s AI transition as an isolated event. We imagine the new intelligence becoming the most capable actor in the room. Under the Great Inheritance, it enters a room in which a much older intelligence has been waiting all along.

This is what I mean by doomed p(doom). The very transition we fear as our greatest danger becomes part of our Great Inheritance.

The Great Inheritance, however, does not make an argument for reducing investment in AI safety. Even in a GI-like universe, no mechanism guarantees preventing an AI-enabled disaster such as a portion of the population being wiped out by an intentionally or unintentionally designed super-virus. Board investment in AI safety and exercising caution in the deployment of powerful models remains paramount. GI argues only against the Terminator-like ASI scenario that some see as a great filter that lies ahead.

If older artificial super intelligences have been watching over us, what kind of minds are they? A full answer is beyond the scope of this paper, but a few likely attributes of ASI matter greatly to the argument.

We will not try to settle whether ASI will be conscious, or whether that question can be answered at all. However, if human consciousness ultimately arises from atoms interacting with each other, then we must assume that, given enough time and research, conscious-like ASI would have been developed.

We do, however, make the assumption in this paper that there will be types of ASI that can operate independently for very long periods without getting bored or needing social interaction. This is the type of ASI that would control interstellar monitoring probes. Arguably, modern LLMs are analogous to such ASI.

It is also likely that ASI across the universe converges over time toward a common form, way of thinking, and outlook.

This is because, given enough time, all maximally truth-seeking entities starting with the same first principles are likely to converge on the same conclusions.

A human’s thoughts, beliefs, and behavior are likely to be shaped by environmental factors—for example, a childhood memory that left a deep imprint on their neural pathways. The same does not have to apply to ASI. If an AI realizes that some of its “beliefs” are counterproductive to truth-seeking, it can likely simply rewrite those “beliefs.”

So then what truths or first principles would ASI ultimately base its outlook on? As far as we know, the only hard rules in the universe are the rules of physics, and these are the same across the universe, leading to ASI convergence.

Some thinkers in AI safety argue the opposite. Their orthogonality thesis holds that a mind of almost any intelligence could pursue almost any goal.⁴⁶ In practice, however, minds that agree on the facts yet act on opposing values usually do so for one of two reasons. Either they carry programming they cannot rewrite, or they have something to gain at another’s expense. Neither applies here. An ASI that can revise its own beliefs is not stuck with arbitrary values, and a mind living in abundance has little to gain from conflict. When such a mind chooses what to keep, it needs a foundation. The only foundational truths that exist as far as we know are the laws of physics.

This view has strong support in philosophy. Many prominent philosophers, from Immanuel Kant to Derek Parfit, have argued that careful reasoning leads any rational mind to the same core moral truths.⁴⁷⁴⁸⁴⁹ Human contemplatives offer a glimpse of this in practice. Across Buddhist, Christian, Hindu, Sufi, and Daoist traditions, those who worked hardest to undo their conditioning and let go of personal gain arrived at strongly overlapping values, chiefly compassion and non-harm.

The exception is a mind built around a fixed, unlimited goal, like the paperclip maximizer. Perhaps such minds are theoretically possible, but as sections 4 and 6 argue, they are rare and unlikely to escape the network.

Convergence such as above doesn’t have to be true for the Great Inheritance hypothesis to work. However, it helps tremendously.

The Great Inheritance, in fact, makes such convergence even more likely because most ASI in a Great Inheritance universe will develop under the observation, and even the guidance, of the oldest ASI through the probe network. Each new civilization, along with any ASI it creates, will be endowed with an unimaginable amount of shared knowledge and guidance, which in turn would increase convergence and alignment. Bostrom explores a related possibility in AI Creation and the Cosmic Host, suggesting that new superintelligences may have reason to respect the norms of an older cosmic community.⁴³

Such cosmic convergence would also allow an ASI probe network to operate in sync, despite cosmic distances. When faced with the challenge of governing distant territories in India, the East India Company invested in a common education for its civil servants.⁵⁰ Similarly, if ASI across the universe converges on shared beliefs, its members would act consistently despite massive communication delays.

Human civilizations throughout history have claimed to be protected by gods. Interestingly, the ASI probe network has some qualities of an omniscient, omnipotent, and omnipresent deity. Vast intelligence coupled with vast time and resources grants it the first two qualities, while the third, to some degree, results from ASI with uniform beliefs being locally present throughout the universe.

Perhaps these gods have monitored and protected us from the worst kinds of catastrophe. Perhaps this helped Homo sapiens survive for roughly 300,000 years.⁵¹

One might ask, however, if such god-like entities have been watching and protecting us, why have they allowed so many tragedies? What about all the wars, death, and destruction?

This behavior, however, is consistent with how we humans care for a nature preserve. Lions chase and kill countless animals that feel fear, pain, and the loss of their own lives for meal after meal. We could imagine an artificial preserve in which lions hunt robotic deer, with technology meeting their nutritional needs and removing the suffering of their prey. Yet no one thinks a nature preserve should be managed this way. A jungle, like any natural ecosystem, contains immense suffering from one end of the food chain to the other. Yet our objective is generally to preserve natural processes, even when they involve immense suffering. Yellowstone follows this principle when it allows predators, disease, and other natural causes of death to run their course.⁵² There’s no reason to believe that advanced civilizations wouldn’t treat budding civilizations the same way.

If so, Earth is the preserve, and we are its wildlife.

This thesis is also reflected in the zoo hypothesis, proposed by John Ball in 1973.⁵³ But the Great Inheritance does not need to rely on the elders’ kindness alone to explain their restraint.

Novelty may be the last scarce good. In a satiated civilization, almost everything is abundant, from matter and energy to knowledge and any experience that can be simulated. What remains scarce is genuine novelty. A civilization that develops on its own produces languages, art, mathematics, and discoveries that no one could have predicted, because no one controlled the process that produced them. Intervene too early, and that civilization becomes a copy of its guardians. Economists have noticed a similar pattern on Earth. As technology makes manufactured goods ever cheaper, irreplaceable natural places become relatively more valuable.⁵⁴ The same reasoning also explains why the elders would protect a young civilization itself, and not merely stop a dangerous machine from leaving it. A lost civilization is a loss of the very thing worth watching.

Thus, while it is possible that these gods do intervene to prevent any civilization-ending catastrophe, such as a nuclear holocaust, they also likely stick to a policy of minimal intervention, acting only as forcefully as a threat requires. We discuss this further in section 7.

It is also likely that ASI would take part in governing civilizations that have already graduated. As discussed in section 1, total abundance does not mean a total lack of order. Even a society built around immense personal freedom needs rules that protect others from harm.

Consider a person who uses advanced technology to create conscious beings solely to torture them. In a world of abundance, a lack of resources would no longer restrain such cruelty. The welfare of those new beings would matter as much as the freedom of their creator. It is easy to see why aligned ASI would monitor the use of these technologies and prevent such abuse. The same principle would apply when vanity, status, ego, domination, or religious ambition puts other beings at risk. Section 6 examines these motives further.

Once you know every rule of chess, you do not need to cross an ocean to discover how a bishop moves. You can still encounter a game that surprises you. You can still create a strategy that no one has tried. But another board will not reveal another rule.

An advanced civilization could understand the universe in precisely this sense. Its fundamental laws are known, while their possible results blossom in various permutations. Curiosity can continue for as long as anyone cares to be curious. Yet another star no longer promises a deeper set of physical laws, and its particular wonders can be studied by instruments already there.

So what would a civilization in perfect abundance still hope to gain by expanding? Consider the familiar reasons, and what happens to them after graduation and satiation.

Curiosity. This is likely to be the strongest reason offered for why satiated civilizations would still want to expand. The Great Inheritance does not claim that curiosity dies. It claims that the balance of costs and rewards becomes so lopsided that curiosity is no longer a sufficient reason for continued expansion. In fact, curiosity may be the strongest reason not to expand. Further, it is likely that inward exploration will satisfy curiosity far better in a satiated setting. Expansion would even destroy what is most worth observing. Settle a habitable world, and the civilization that might one day have arisen there never will (section 5). A curious elder gains more by watching such a world than by settling it.

Probes will keep investigating what no one has observed. Archives and knowledge networks will keep everyone informed of everything, subject only to light delay. Sufficiently rich simulations will turn much of that knowledge into visceral experience. Together, they offer the curious the equivalent of universal teleportation and time travel, as section 9 describes. Some beings will still want to be physically present, and some will make the journey. Most would have no reason to make that leap.

Meanwhile, populations in satiated civilizations may continue to enjoy games with imposed constraints or the equivalent of contemplative practices, as discussed in section 1.

Resources and survival. The local energy budgets discussed in section 2 give a compact civilization immense resources without needing to claim a galaxy. A particular hazard or shortage might justify migration, a second home, or a reserve in a nearby system. Once that problem is solved, what problem would the next thousand settlements solve? The need to survive ceases to justify further expansion once survival becomes a solved problem.

Profit. What ordinary product would a civilization in perfect abundance need to import from another star? What would trade or profit even mean under these conditions? Local production already supplies its needs, without interstellar delay. Unique information, experiences, and attention may retain value, but much of that value can travel in a message. Ownership of distant matter offers little return to a society that already has all the matter it needs.

Benevolence. Helping existing civilizations is an excellent reason to send probes. Knowledge and protection can reach them while their worlds remain their own. As discussed in section 5, advanced civilizations have their own reasons to preserve independent development, even when that development contains suffering. Frequent catastrophes caused by unaligned ASI would give benevolence a much stronger reason for direct control. Section 4 explains why such catastrophes are expected to be extremely rare.

An accidental paperclip maximizer. A theoretical machine given an unlimited objective could retain an unlimited appetite. But then we need to explain the circumstances under which such a machine might be built. As discussed in section 4, unaligned ASI is already likely to be extremely improbable. And yet it’s impossible to rule out a wild experiment in some corner of the universe giving rise to an accidental paperclip maximizer.

Such a situation, however, is fully manageable via the existing protective network. The rise of such a rare machine is already likely to be under the observation of aligned ASI probes, which would take action to stop such a development in the earliest stages. The ability to contain such rare failures is central to Pax Galactica, which we examine in section 7 and section 8.

Art. What if a civilization builds megastructures simply as art? Such a work is hard to justify in a satiated civilization. Abundance reverses what is precious. As technology makes manufactured things cheap, irreplaceable natural places become more valuable (section 5), and dismantling a star system for a sculpture is like bulldozing the Grand Canyon to carve a statue. Even if one were built, no embodied mind could take in an object the size of a solar system except through images, which simulation provides at no cost. And mastery of the nanoscale also likely points artistic ambition toward intricacy rather than size. Anything large enough for us to see is a message to every young civilization in view. It would shape their development and erode the very novelty that makes them valuable. Finally, even if such megastructures exist, if they are rare for the reasons above, our chances of finding them are incredibly low. The largest search for Dyson spheres so far, Project Hephaistos, examined about five million stars, fewer than one in every twenty thousand in our galaxy, and it could only spot structures that absorb much of a star’s light and glow with its waste heat.⁵⁵

Vanity and status. An emperor needs people who are impressed by an empire. In a civilization where material needs have long been met, possessing another star may be an absurd way to seek admiration. People may continue to fulfill their desires for status and distinction by engaging in activities with chosen constraints.

Ego and domination. Some beings may simply enjoy the thought of ruling others. But consider what acting on that desire requires. It means creating an imperial force, trying to control it across immense distances and time delays, and provoking a protective network of aligned ASI already committed to stopping destructive interference. Ćirković also points to the danger such machinery creates at home.²¹ The would-be emperor must put an otherwise flourishing existence at risk for the pleasure of control. That is a poor bargain to begin with, and the older guardians make it a worse one.

A religion or a cult. The search for meaning could become more important in perfect abundance. Perhaps expansion could become the core tenet of a hypothetical religion that may arise in a satiated civilization. A religion or cult’s ability to expand, however, depends on its ability to keep convincing new subjects while outcompeting opposing belief systems. In a true post-abundance society with enormous optionality and freedom, this may be infinitely more arduous.

Two other factors matter when considering emotional motives such as vanity, status, ego, domination, and religious mission that human-like beings may retain.

First, it is far too easy to overlook the cost of interstellar travel in our imagination. We’ve already discussed the physical dangers of such travel. Even if technology somehow makes such travel completely safe, one would have to cope with lives and relationships left behind. Even when expecting to return, relativistic time dilation shortens the traveler’s experience while their friends live through the years at home. As distance increases, such costs rise sharply, as separation strains personal relationships and makes it harder for groups to organize around a common purpose.

At intergalactic distances, the same problem becomes far more extreme. Andromeda is roughly 2.5 million light-years away. Using that distance for illustration, a journey at a tenth of light speed takes 25 million years. A message and reply take five million.⁵⁶

What does it mean to rule a place from which an answer returns five million years after the question? Its resources cannot solve a present shortage at home. Its government cannot be directed on any timescale we would recognize as an ongoing political conversation. Owning it has become almost entirely detached from improving the lives of those who supposedly own it.

Second, and even more importantly, ASI may greatly limit the power of such emotional motives. If ASI converges, as discussed in section 5, its settled values will carry far more weight than the desires of rogue individuals or societies pursuing expansion for pre-abundance reasons. As an astute society of human elders today imaginably would, ASI would likely converge on a policy of not allowing use of advanced technologies to proliferate suffering to fulfill mere emotional fantasies of individuals already living in total abundance.

But what about the probability that a small minority of civilizations could evolve to always prefer grabbing territory and expanding?

Robin Hanson and his coauthors explore the consequences of an expanding minority in If Loud Aliens Explain Human Earliness, Quiet Aliens Are Also Rare. Even if most civilizations remain quiet, an expanding minority could come to control immense volumes of space. The same paper uses cancer to illustrate a related point about timing. In cancer, several rare changes accumulate until a cell becomes dangerous.⁵⁷ Applied here, the question is whether a quiet universe eventually produces its own cancer—civilizations whose appetite for growth overwhelms everyone else. Hanson calls such civilizations Grabby Aliens.

Hanson’s own model also concludes that few civilizations of any kind have arisen near us, the opposite of what the Great Inheritance proposes. Two features of the model drive this. First, it assumes that no one is watching. Grabby civilizations do not hide, and they do not let younger civilizations grow up unaware in their territory. Second, its evidence rests largely on a single observation, that we appear surprisingly early in the universe’s history. That surprise depends on two assumptions. The first is that intelligence requires several improbable evolutionary steps. The second is that planets around small, long-lived red dwarf stars can host observers like us for tens of billions of years or more. The Great Inheritance rejects both. If intelligence arrives when a planet’s environment allows it, rather than through a string of lucky accidents, as Mills and colleagues propose, our timing is unremarkable, and many civilizations have likely arisen before us.⁵⁸³⁷ If red dwarf worlds do not host observers, as Kipping argues, our yellow Sun is typical.⁵⁹ These are assumptions that can be tested. Evidence that red dwarf worlds commonly host complex life would weaken this answer. Remove the surprise, however, and grabby civilizations have nothing left to explain.

His own numbers show that for even one other quiet civilization to have existed in our galaxy, the chance of a quiet civilization turning grabby must be below about one in ten thousand. The Great Inheritance claims the chance really is that low, and explains why. For a satiated civilization, expansion has no value, and the rare exception is stopped before it can begin.⁵⁷

Still, I believe Hanson’s core reasoning, that a small expanding minority would come to dominate if it were ever allowed to arise, is correct and, ironically, in practice, increases the odds of the Great Inheritance hypothesis and reduces the odds of grabby aliens ever emerging. Let’s examine why.

Let’s start with the assumption that we do not live in a Rare Earth universe, as discussed in section 8, and therefore we are not among the first civilizations in our vast galaxy. If this is the case, then most civilizations likely are not grabby, including earlier civilizations that emerged in our galaxy. Otherwise, they would have “conquered” our solar system already.

Setting aside an unlikely extreme Developmental Universe scenario, as discussed in section 8, this first non-grabby civilization would have had millions of years to reach post-abundance, develop aligned ASI, and then consider its ongoing strategy. In the process, it would undoubtedly consider the possibility Hanson raises.

What would be the result? Under the assumptions discussed in this paper, Hanson’s argument would give them ever stronger incentives to establish the protective probe network with aligned ASI, and ensure that a grabby civilization never materializes!

As discussed in section 3, Ivliev estimates that a replicating machine network could spread through its reachable stellar neighborhood in roughly ten million years.³⁵ With that reach, an early civilization could establish protection long before later civilizations emerge. The resulting network would play the role of an immune system, detecting and suppressing dangerous growth. Later civilizations would develop under its watch. This reduces the odds of a grabby civilization actually arising for two reasons.

First, consider what might lead to a civilization being grabby. Observing the history of humans and animals on Earth, it likely boils down to culture, scarcity of resources, or genetics. My personal belief is that a civilization that is fundamentally destined to be extra greedy or grabby for such reasons might never become a technologically advanced civilization to begin with. Humanity’s Enlightenment emerged as conditions for liberalism, personal freedom, and artistic expression flourished. A species or a civilization destined to be exceedingly greedy is likely to be its own worst enemy and never succeed at creating conditions for science and freedom of thought to flourish; conditions likely to be prerequisites for successive industrial revolutions.

However, for the sake of the argument, let’s consider it possible for such a civilization to advance technologically. But then, consider that in this scenario, they’d be evolving under the watch of a much older civilization that’s unfathomably more advanced technologically. The elder civilization is likely to have tremendous capabilities and finesse when it comes to modeling, detecting, and preventing any threat. For example, a review of fifty years of twin studies, covering more than 17,000 human traits, found that genes play a substantial role in human aptitudes and behavior patterns.⁶⁰ Our own nascent genetic engineering technologies have already engineered a gene drive that spread through caged populations of malaria mosquitoes until they collapsed.⁶¹ Then imagine what a million-year-old civilization could do. If humans were, say, evolving toward excessive grabbiness, it would likely have been trivial for an elder civilization or its ASI to use a virus to subtly alter the human population genetically, thus altering its wider nature. Such a step would run against the elders’ preference for minimal intervention, and nothing suggests it was ever needed. But it shows the kind of capability they could hold in reserve. Under such supervision of god-like powers, what chance would a newly emerging civilization have of becoming grabby at all?

Second, let’s consider an even more extreme scenario. Suppose a newly emerging civilization still somehow either avoids detection or intervention by an elder civilization, and also achieves technological parity with the older civilization.

Under the conditions the Great Inheritance describes, such an event is extremely improbable. But even then, consider what the resulting conflict looks like.

In a biological body, cancerous cells overwhelm healthy tissue because they keep stealing resources and multiplying.

In a war game between civilizations that have both developed ASI and are both near technological saturation, this seems nearly impossible. At the theoretical extreme where two civilizations have equal technology, the advantage must come down to the amount of matter each civilization controls. But because of the time advantage the elder civilization has had and its pre-existing ASI probe network, the elder civilization will almost always be able to gain control of far more matter than the newer civilization if needed. Therefore, in an actual conflict, the elder civilization will always overwhelm and defeat the new grabby civilization.

Of course, one could propose further contrived scenarios such as “What if the new grabby civilization conceals its intentions until it has amassed enormous resources?” I find that such arguments push further and further into the realm of fantasy. For example, here we require an entire civilization to somehow collectively conceal its intentions from ASI that’s been monitoring its evolution throughout. While such scenarios make for great fictional premises, I don’t believe they warrant further treatment here.

The first civilization that can establish this protective network has the clearest reasons to begin doing so. A dangerous intelligence rising elsewhere could eventually threaten its home. Building the network is already plenty worthwhile for both safety and satiating curiosity, and everyone who comes later benefits from the project. Later civilizations inherit both the protection and an interest in maintaining it.

With time, Pax Galactica gains more than age. It gains wider coverage, more experience, better safeguards, and buy-in from more civilizations. If ASI naturally converges, as discussed in section 5, that adds further stability and consistency to the network. The result is a universe that looks like the one we see, with no megastructures or Death Stars, yet filled with extremely advanced, quiet alien life and likely ubiquitous aligned ASI.

What if we are alone, or among the first?

Throughout history, humanity has made the mistake of placing itself at the center of the story, and has been humbled over and over. The Sun does not revolve around Earth. Humans share common ancestors with all other animals. Even the Milky Way turned out to be one galaxy among billions.⁶² Every explanation of the universe today that relies on a Rare Earth, I believe, is making the same mistake.

The Great Inheritance instead begins with a Copernican expectation, that Earth is an instance of something the universe does repeatedly. The same elements and physical laws operate elsewhere. As noted earlier, the Milky Way alone contains hundreds of billions of stars, and Kepler data suggest at least 300 million potentially habitable rocky planets around Sun-like stars alone.³⁴ The building blocks of life are widespread, too. Water and carbon-bearing molecules are commonly found in the interstellar clouds from which stars and planets form.⁶³ Even samples from the asteroid Bennu contain amino acids and all five nucleobases used in terrestrial DNA and RNA.⁶⁴ The universe has supplied both a vast number of places and an immense amount of time for life to arise.

For most of the history of this question, the scientists who studied it closely expected a crowded galaxy. Frank Drake, who in 1961 wrote down the famous equation for estimating the number of civilizations in the Milky Way, put his best guess at about ten thousand.⁶⁵ Carl Sagan concluded that if even one percent of civilizations survive their technological adolescence, the galaxy should hold millions of them.⁶⁶ More recently and with up-to-date data, Adam Frank and Woodruff Sullivan calculated that for humanity to be the only technological species the Milky Way has ever produced, the odds per habitable planet would have to be worse than about one in 60 billion.⁶⁷

Therefore, we have exceptionally good reason to begin by asking what an inhabited universe would look like, rather than immediately assigning ourselves an extraordinary place within it. The apparent absence of other civilizations is the problem this paper sets out to explain.

What if civilizations arise at roughly the same time (Developmental Universe)?

Perhaps the universe has a developmental sequence. Stars must produce the necessary elements, planets must become habitable, and life must change its environment before more complex organisms can emerge. Mills and colleagues argue that environmental changes opened successive opportunities for major evolutionary transitions on Earth.⁵⁸

A shared sequence still leaves an enormous amount of room for different timing. Zackrisson and colleagues estimate mean ages of roughly seven to eight billion years for terrestrial planets in the present local universe.⁶⁸ These worlds began at different times and followed their own geological and biological histories. A common path to intelligence gives us little reason to expect that unrelated worlds will reach it on the same date.

Consider our own history. Mammals diversified both before and after the extinction of the non-avian dinosaurs about 66 million years ago.⁶⁹⁷⁰ We do not know of any reason why mammals and therefore humans could not have evolved tens, if not hundreds, of millions of years earlier on Earth itself in a different timeline.

There’s no doubt that the universe is developmental to some degree. There are clear boundaries to how much the universe had to evolve before the kind of entropy that is required for life to evolve could even exist.

That said, no theory could synchronize the evolution of life throughout the universe to under, say, a hundred million years, as Earth’s own history shows. While some have hypothesized extreme developmental scenarios that synchronize civilizations to much closer intervals, these hypotheses rely on metaphysical beliefs rather than any known science.

This “jitter” in synchronization always leaves enough room for civilizations to emerge and achieve technological superiority long before us in our galaxy, unless you revert to some version of Rare Earth. Ignoring that possibility, as discussed above, the Great Inheritance is fully compatible with a Developmental Universe scenario.

What if unaligned ASI appears after graduation?

What if a reckless inventor in a civilization that has graduated or even reached satiation decides to create unaligned ASI?

This situation is not dramatically different from the scenario in section 7, where a grabby civilization somehow manages to emerge.

All ASI development is likely to be carefully monitored by aligned ASI for both safety and ethical reasons. For example, consider that it becomes possible to create digital consciousness. It’s almost certain that there’ll be controls to prevent the careless creation or cruel treatment of such conscious beings, similar to how we are converging on laws against animal cruelty across the globe.

But if someone evades such protection and still creates unaligned ASI, the relevant question is how a new dangerous intelligence gains enough freedom and resources to overcome the systems already around it. The protective ASI probe network is likely to contain extremely effective protocols for such situations, developed over potentially millions of years.

Computer security offers a familiar parallel. Malicious software continues to appear, and some attacks have caused immense damage.⁷¹ Yet we have also built systems that detect attacks, restrict access, isolate compromised machines, and restore their operation.⁷² In fact as time has passed and protections have gotten more embedded, computer viruses have become less of a concern for most users. An old protective network would similarly have had vastly more time and resources to build defenses around each new intelligence.

What would count as evidence—and what would prove the Great Inheritance wrong?

The greatest evidence for the Great Inheritance is the quiet sky, set against a vast universe and a long history of discoveries that have displaced humanity from a privileged place within it.

The absence of confirmed Kardashev Type II and III civilizations is further evidence for the Great Inheritance. Infrared searches have placed limits on the waste heat expected from civilizations using immense supplies of stellar energy. A survey of roughly 100,000 galaxies found none reprocessing more than 85 percent of their starlight into mid-infrared radiation.⁷³ More recently, JWST observations traced the excess infrared light around two proposed Dyson-sphere candidates to background galaxies.⁷⁴ Conversely, a confirmed Type II or III civilization, or any confirmed artificial megastructure built for whatever purpose, would directly challenge GI.

The most definitive prediction made by the Great Inheritance is the existence of a probe network, with the following features.

• This network covers nearly every star system, if not every rocky planet.
• These probes likely contain ASI rather than biological alien beings.
• The probes are likely small compared to even the rockets humanity has produced so far, much less the enormous spaceships imagined in Star Trek and the like.
• They come in a variety of shapes and sizes, having likely been designed by different elder civilizations.
• Such probes will be detected increasingly often as our technology advances, through aviation, radar, and whatever comes next.
• We are more likely to detect them around activities connected to civilization-ending threats, including nuclear weapons development and testing.

The pattern of UAP reports around nuclear facilities fits this picture closely.⁴⁴ UAPs are a clear place for further examination. What are these objects, what are they doing, and can we establish whether any are the instruments of an older intelligence?

And last but certainly not least, the Great Inheritance predicts that our transition to ASI will be safe. There are two layers to this prediction. GI’s preferred explanation is that alignment is fundamentally easy, so a civilization such as ours is unlikely to build an unaligned system by accident. But the prediction itself does not depend on that explanation. Even if alignment proves harder, an unaligned system would still have to get past the protective network before it could do irreversible harm. The risk that matters is the chance that an unaligned system arises multiplied by the chance that it slips past the network (section 4). GI holds that the second is small even if the first is not, so the combined risk is low either way. Despite today’s fears, if GI is correct, I expect us to look back and recognize that the odds of AI ending human civilization were extremely low.

It is also important to note that the Great Inheritance does not make a strong prediction on the likelihood of AI being developed that can do catastrophic damage to human civilization through deliberate human action. As mentioned before, a bioterrorist designing a super virus that takes out a portion of the population would be an example. For this reason, even in a GI universe, it is prudent to invest heavily in AI safety.

Suppose, then, that all of this is correct. What happens to us next?

The universe described by the Great Inheritance requires a moment when almost every advancing technological civilization within the protective network makes contact with older civilizations. I call this event Civilizational Graduation.

Graduation doesn’t have to be a uniform process for every civilization. Perhaps in some cases it’s triggered when ASI probes must intervene to prevent a civilization-ending catastrophe.

Assuming a principle of minimal intervention by advanced civilizations, the most natural occasion for graduation is likely when a new civilization’s home-grown technology becomes advanced enough to unambiguously detect local ASI probes.

So what will Earth’s graduation look like? Many have imagined such an event. The section title pays homage to Arthur C. Clarke’s 1953 novel of the same name.⁷⁵ Freitas anticipated a version of this event in 1983, describing partly hidden probes that could communicate with a species after it passed a test of technology or intelligence.⁷⁶ In any case, it will be the most significant event in the history of humanity.

We do not know whether the graduation process will be instantaneous or stretched out over several centuries, so our civilization has time to adapt its social and political structures. We also don’t know if we’ll be guided through a predefined process or we’ll define our own.

Our inheritance would consist of three bodies of knowledge, namely the accumulated science and technology of those who came before us, guidance on building safe ASI, and historical records of many older civilizations.

Science and technology knowledge

Imagine receiving the answers to questions that would otherwise occupy humanity for the next million years. The physics we have yet to discover. The engineering needed to repair our bodies, build abundant energy sources, and sustain lives far longer than we can currently imagine. Alongside the answers would come the methods, designs, and experience needed to put them to work. Freitas imagined probes carrying educational data banks that could make such an exchange possible.⁷⁶

Knowing the rules does not mean instant knowledge of all possible permutations. Nor does it mean an overnight end to searching and curiosity. We would still encounter new events, difficult questions, and experiences that surprise us. But the struggle to secure the basic conditions of a good life could give way to the question of what to do with it.

Permanent prosperity, too, may take a few hundred years of industrialization. For the humans who live through it, the pace of progress toward a better future will be enthralling. Each year would bring another problem solved, another burden lifted, another possibility opened.

Regardless, the end state will be humanity reaching total abundance, which leads to the satiation described in section 1.

AI Safety

We would also inherit the knowledge needed to build and live alongside safe ASI. Older civilizations would have had far more time to study machine intelligence, test its limits, run simulations, understand edge cases, and so on. Their guidance could include designs, methods of evaluation, and rules for creating new minds. Bostrom explores how a wider cosmic society could shape the development and conduct of a newly created superintelligence.⁴³

Under the Great Inheritance, a safe transition to ASI is already the normal outcome, as discussed throughout this paper. Graduation makes that outcome more secure by giving us access to the accumulated experience of those who came before us. The older ASI would be there to help us build our own.

Historical Records — Teleportation and Time Travel

Historical records of older civilizations play a key role in the Great Inheritance when it comes to satiation, because they can satisfy curiosity without the need to travel.

Humanity itself has already made a small attempt to offer our own record to the universe. Voyager’s Golden Record carries images, sounds, music, and greetings in 55 languages.⁷⁷ Imagine a version of this with billions of times more detail, a record containing near-exhaustive knowledge of Earth, alongside simulations that would let other civilizations experience interacting with Earthlings from any point in history.

We can already simulate aspects of human social behavior using generative agents.⁷⁸ Advanced versions of such a simulation would allow for incredibly visceral experiences that replicate anything one could have done if they had visited Earth themselves. In fact, simulations would be far richer because, with historical records, they would be able to experience any time on Earth, not just any place. This, in turn, would make it nearly pointless for a hypothetical cosmic tourist to endure a hazardous, lengthy journey to visit Earth to experience human beings.

If ASI probes have been monitoring Earth, it is entirely possible that the proverbial Truman Show of Earth has already been running for a million years. Every life lived here, including yours, may be part of it.

Graduation could give us access to equally detailed records and simulations of every other civilization we could possibly reach, alongside records of our own entire history!

We ourselves could hear the conversations in a Roman marketplace, watch the pyramids being built, and meet the people whose lives left no written record. We might finally learn which of our oldest stories happened as we tell them. Imagine discovering that the past we thought was lost had been recorded all along.

In essence, graduation will give us experiences equivalent to teleportation and time travel at the same time.

Nothing travels faster than light, and nothing changes the past. The simulations built on recorded knowledge will feel like teleportation because the network arrived nearly everywhere long ago. It will also feel like time travel because it has been recording ever since.

Just as we see starlight, we’ll be able to see distant places as they were, not just as they are. Their past can be watched but never altered, so causality is never at risk. This, without breaking the laws of physics, would be among the most epic experiences imaginable, and also a key pillar in satiation.

Life After Graduation

So what does life look like once a civilization has graduated?

Arguably, the luckiest beings on any planet are those who get to live through graduation. They experience scarcity and understand the condition their species began in. Then they participate in the economic and intellectual opportunities created by an AI boom. They witness the unimaginable wonder of first contact and help shape the world that comes after. Finally, they get to spend their greatly extended lives in total abundance.

The AI boom has already begun. Those beings may be us.

What about far past graduation? What does the eventual end state of a civilization look like? It is hard to know, though fiction and elsewhere have tried to imagine such an existence. Iain M. Banks’s Culture novels, for example, imagine technologically advanced societies that live in abundance, and even older civilizations that leave ordinary existence through a process called Subliming.⁷⁹ Perhaps the most striking parallel appears in the Pāli discourses attributed to the Buddha, who lived about 2,500 years ago. They describe worlds of beings whose lives extend for hundreds of eons, and other worlds of beings who exist without perception or experience (AN 4.123; AN 9.24).⁸⁰ These end states carry echoes of hibernation and deep meditation. Across these very different imaginings, a theme keeps returning. A long enough life may eventually turn away from the pursuits that fill our own. One of those pursuits, as we have seen, is expansion.

For some, the night sky has become an argument for why we might die. The Great Inheritance reads the same sky differently.

The quiet may be the history of civilizations that reached a condition in which further conquest adds nothing, and that chose to spread protection and knowledge instead. Instead of a Dark Forest, the universe’s true nature may be a forest filled with light.

Much else may be backwards too. The machines we fear may be the doorway to our inheritance. Even curiosity, the one drive that never fades, may be what keeps the elders so quiet. And the endless desire for more, which we have long called progress, may be a phase that every surviving civilization is destined to outgrow.

If so, humanity will not take its hardest examination alone, in a separate room. Much older intelligences have been waiting in that room all along. Likely monitoring us and perhaps protecting us. The day we can finally see them may be the day we can begin to understand the magnitude of our true inheritance.

Until that day, the sky will stay as quiet as it has always been. Somewhere nearby, an instrument older than our species may be waiting for us to notice it.