Technological Uniformity Saves the Simulation

Nick Bostrom’s formulation of the Simulation Argument is a rigorous reworking of what is, at its heart, an ancient question. Are we living in a created world? He distills the answer into three stark possibilities, a trilemma:

1. Doom: Almost all civilizations destroy themselves or otherwise fail before developing the capacity to create detailed simulations of their ancestors.

2. Abstinence: Some civilizations develop this technological capacity, but almost all choose not to simulate conscious agents, for ethical or other reasons.

3. Simulation: If our civilization survives and runs ancestor simulations, then simulated agents would vastly outnumber non-simulated agents, and, all else equal, our credence that we’re simulated should be very high.

In a critical analysis, Brian Eggleston highlights an unstated assumption underlying Bostrom’s formulation: we are not alone as technological pioneers. Specifically, Bostrom’s trilemma only entails the third possibility, that we’re simulated, if we assume that some other civilization – not only our future descendants – already ran ancestor simulations before our present. Without this assumption, we could imagine humanity as the first or only simulator, collapsing the trilemma into merely weak possibilities without force.

Principle of Technological Uniformity

However, I believe there’s an unstated intuition, yet another unrecognized assumption, behind Bostrom’s formulation. And that intuition, when identified and formally expressed as an assumption, fully addresses Eggleston’s criticism and maintains the force of the trilemma. I call that assumption the Principle of Technological Uniformity (PTU):

If a given technology is feasible, beneficial, and once achieved by a civilization, then, all else equal (barring radically unique physics or values), that technology probably has been or will be achieved by other civilizations operating within similar conditions.

PTU has at least three important characteristics. First, it provides rational grounds for supposing that becoming a simulator increases the probability that other simulators exist. Second, it reflects and extends many empirical precedents, including those broadly categorized as convergent evolution. Third, it maintains the strength of Bostrom’s trilemma, solving the problem that Eggleston identifies without appealing to exceptionalism.

Philosophical Support

PTU is grounded in the principle of mediocrity, or what people sometimes call the “Copernican” principle. As Copernicus removed from Earth the privilege of being the center of our conceptualization of the cosmos, so PTU would have us resist the temptation to privilege our human civilization uniquely in time or space or significance. Instead, given uncertainty and no evidence for our uniqueness, we should regard ourselves as typical. Human civilization on Earth is just one among many technological civilizations, subject to similar physics and incentives.

PTU is related to the anthropic principle. When considering the possibility of others developing technology like us, we shouldn’t default to flattering assumptions that would make us exceptional. To the contrary, as we observe ourselves developing increasingly detailed simulations, that should raise our credence that others have developed similar technologies before us. PTU is a meta-induction about technology diffusion, similar to but distinct from the self-sampling assumption.

Eggleston’s criticism hinges on the possibility that another civilization has already become a simulator. PTU invites us to assume this possibility has a significant probability, both for the philosophical reasons mentioned above, as well as empirical reasons shared below. We should assume that, where potential and incentive align, scientific discoveries and technological developments propagate, not deterministically, but frequently enough to warrant inductive reasoning from specific observations to generalizations.

Empirical Support

PTU is not merely a philosophical abstraction. Biological, cultural, and technological evolution are replete with convergence. Life and intelligence within similar constraints repeatedly arrive at similar solutions to similar problems.

Biological examples abound. The camera eye, constructed from radically different biological materials, appears to have evolved independently in both cephalopods (ancestors of the octopus) and vertebrates (ancestors of humans). Powered flight apparently evolved independently in insects, birds, and bats, reflecting similar constraints in aerodynamics. Where an environment rewards a particular function, nature finds a way to achieve that function, often more than once.

Cultural examples also abound. Agriculture seems to have begun independently on multiple continents, perhaps millennia apart. Writing systems appeared at disparate times in Sumer, Egypt, China, and Mesoamerica with little to no evidence of direct transmission. Even intricate toolmaking, such as fluted point technology used for projectiles like arrows, appears in archeological records on separate continents, despite a paucity of evidence for contact or common inheritance among them.

Today, we can observe technological convergence in the rapid global proliferation of computation. Moore’s Law (and its generalization into Kurzweil’s Law) is not just a local phenomenon. Countless independent groups have, motivated by shared incentives, moved computation and simulation technologies forward at accelerating rates. Now we construct virtual realities, whether for entertainment or science or otherwise, as if echoing an ancient call to create new worlds.

Saving the Simulation

Eggleston is right to caution against hidden assumptions. If, in actual fact, humanity were to become the only civilization to reach the simulation threshold, Bostrom’s trilemma would collapse. And the force of his articulation of the Simulation Argument would evaporate. PTU solves this problem, recognizing another hidden assumption, by insisting that when technology is both feasible and valuable, history predicts independent rediscovery.

It would be extraordinary to claim that human civilization will become the first and only simulator, uniquely lucky (or uniquely destined) in our technological potential. The probability that we are or will be the only civilization with our particular potential in any given technological domain shrinks against the weight of empirical precedent in the biological, cultural, and technological evolution that we have already observed or discovered. Assuming we eventually become simulators, the probability that we ourselves were simulated rises accordingly.

Backed by PTU, the Simulation Argument trilemma stands strong:

1. Doom: Almost all civilizations perish before becoming simulators.

2. Abstinence: Almost all civilizations choose not to become simulators.

3. Simulation: We almost certainly live in a simulation – one among unimaginably many.

To assert that humanity is a unique exception to PTU would be radical anthropocentrism, inconsistent with logic and history.

Conclusion

The Principle of Technological Uniformity is deeply rooted in both philosophical reasoning and empirical evidence. From that dependable ground, it resolves the problem that Eggleston identifies in Bostrom’s formulation of the Simulation Argument. If we recognize ourselves as neither central nor exceptional, but rather as part of the grand convergent patterns of the cosmos, then we should also recognize that we could not have become simulators unless other simulators already exist. Thus, Bostrom’s trilemma persists with full force.