Bitcoin and the Alignment of Autonomous Agents

From Timothy Leary and Extropianism to a Substrate-Independent Economy of Reputation

What if Bitcoin was designed not simply for the world we inhabit, but for the world its creators anticipated?

## Are We Solving the Wrong Alignment Problem?

How can we expect to solve AI alignment when we haven't managed to solve human alignment?

We've had thousands of years to figure out how to coexist. We still disagree about morality, politics, economics, religion, and even what constitutes a good life. We create governments and institutions intended to serve humanity, only to discover that those institutions can develop objectives quite different from those of the people they're supposed to represent. And now we're attempting to teach increasingly powerful artificial intelligences to understand and pursue human values.

Whose values, exactly? Humanity doesn't possess a universally accepted moral framework, and I see little reason to believe we're going to arrive at one simply because artificial intelligence makes the question more urgent. Perhaps the problem isn't that we've failed to agree on the correct values, but that we keep assuming alignment requires such agreement in the first place.

For several years, I've suspected that Bitcoin represents one of the most important solutions to alignment we've developed. Not a complete solution to every problem of intelligent behavior, but something potentially more fundamental: an architecture that allows autonomous agents with different motivations to cooperate without requiring shared values or centralized authority. It accomplishes this through independently verifiable rules, economic incentives, and a persistent public history of transactions that no individual participant can unilaterally rewrite.

More recently, I've begun to wonder whether this was something its creators anticipated. What if Bitcoin was designed for a future populated by autonomous economic actors, human and nonhuman alike? The more I examine its intellectual origins, particularly the Extropian movement, the more compelling that possibility becomes.

## Alignment Is Substrate-Independent

Before going further, I think it's important to clarify what I mean by an agent, because this distinction is fundamental to the argument.

An agent is any autonomous participant capable of pursuing objectives, making decisions, controlling resources, and interacting with other agents. An agent might be a human being, an artificial intelligence, a corporation, a government, or some form of intelligence we haven't yet encountered or created. Organizations may exercise agency through their decision-making structures even though they aren't individual minds. The physical substrate in which intelligence operates is irrelevant to the fundamental problem of economic alignment.

We've become accustomed to discussing human alignment and AI alignment as though they were entirely separate problems. I don't think they are. Both involve autonomous agents pursuing potentially conflicting objectives within a shared environment. Both require mechanisms for establishing trust, coordinating activity, honoring commitments, resolving disputes, and dealing with misconduct.

Why should the biological origin of an intelligence determine whether it is trustworthy? A human isn't inherently more reliable than an artificial intelligence, nor should an AI be considered less trustworthy simply because it isn't human. A government doesn't become trustworthy by virtue of its authority, and a corporation doesn't acquire credibility merely through legal incorporation. What matters is behavior, and more specifically, the independently verifiable history of that behavior.

This is where reputation becomes essential. An agent's reputation should develop through its actions, the commitments it fulfills, and the evidence other participants can examine. Reputation isn't something that should be assigned by a central authority or inherited through membership in a particular class of agents. It is something accumulated over time through demonstrated conduct.

A newly created anonymous identity begins without an established reputation. That doesn't make it dishonest or deserving of punishment; it simply means that other agents have little evidence upon which to extend trust. An established agent, by contrast, may have accumulated years of verifiable economic history, regardless of whether the intelligence controlling it is human, artificial, or something else entirely.

This produces what I consider a foundational principle: the same basic standards of economic accountability should apply to every autonomous agent. Different capabilities may create different practical responsibilities, but those distinctions should arise from conduct, commitments, and circumstances rather than arbitrary assumptions about an agent's origin. The objective isn't to align AI with humanity, as though humanity were already aligned. It's to establish conditions under which all autonomous agents can coexist, cooperate, and remain accountable to one another.

## Timothy Leary and the Idea of Constructing the Future

Timothy Leary is remembered primarily for his association with psychedelics and the counterculture of the 1960s. I think his influence on subsequent technological and philosophical developments has been substantially underappreciated. His later work went well beyond the exploration of consciousness through psychoactive substances, examining the possibility that humanity could deliberately participate in its own evolution.

During the 1970s, Leary promoted a program he called SMI²LE: Space Migration, Intelligence Increase, and Life Extension. These weren't simply predictions about what technology might eventually accomplish. They were objectives to pursue through experimentation, scientific research, and technological development.

Leary was proposing that evolution need not remain an exclusively biological process. Intelligence could be expanded, lifespan extended, and humanity's environment transformed through deliberate action. Rather than treating existing human capabilities and social institutions as permanent constraints, he encouraged the exploration of alternatives.

Science historian W. Patrick McCray documents this influence in his 2016 study, Timothy Leary's Transhumanist SMI²LE, describing how Leary's work provided a stepping stone toward the technologically activist transhumanism of the 1990s.[1] The connection to Extropianism was explicit. Max More included Leary's Info-Psychology among the recommended readings associated with the Extropian Principles.[2]

What interests me is the continuity of thought. Leary challenged inherited assumptions about consciousness, autonomy, and human potential. The Extropians extended that challenge into a more systematic effort to develop the technologies, institutions, and economic structures that could support a radically different future. Rather than simply predicting the future, they wanted to participate in constructing it.

That distinction is important because Bitcoin may represent precisely the sort of infrastructure such a philosophy would encourage its participants to develop.

## Extropianism: Building the World We Wanted

The inaugural issue of Extropy, published in 1988 by Max T. O'Connor, later Max More, and Tom W. Bell, explicitly anticipated artificial intelligences vastly exceeding human computational abilities. It discussed radical life extension, technological self-transformation, decentralized social institutions, spontaneous order, and the possibility of fundamentally changing human existence.[3] This was decades before modern artificial intelligence made these questions immediately relevant to everyday life.

The Extropians weren't interested solely in creating more powerful technologies. They were concerned with the social and economic structures through which increasingly capable intelligences might coexist. Their philosophy emphasized individual autonomy, voluntary cooperation, experimentation, open institutions, and resistance to concentrated coercive power.

Max More's later Extropian Principles advocated the development of social institutions through experimentation rather than the centralized imposition of a single vision. It also emphasized that laws and rules should apply equally to those exercising authority and to everyone else.[2]

This is particularly relevant to alignment. Rather than trying to establish a universally accepted moral system, the Extropians recognized the value of decentralized arrangements in which individuals and groups could pursue different objectives. Their interest in spontaneous order reflected the possibility that cooperation could emerge from independent participants without requiring centralized direction. The rules governing those interactions could be simpler and more universal than the beliefs or motivations of the participants themselves.

The objective wasn't necessarily universal agreement. It was creating conditions under which disagreement did not prevent cooperation. The Extropians also believed that technological development would require corresponding changes in social infrastructure. If intelligence, individuality, and economic relationships were going to change radically, then many of the institutions supporting existing society might prove inadequate.

They sought to build alternatives before those changes became unavoidable. I increasingly wonder whether Bitcoin belongs to that broader effort.

## Autonomous Software Economies Were Already Being Designed in 1988

One of the most remarkable connections comes from Mark S. Miller and K. Eric Drexler. In 1988, they published Markets and Computation: Agoric Open Systems, proposing computational environments in which software objects and autonomous agents could participate in markets, purchase resources, negotiate contracts, and coordinate their activities through economic mechanisms rather than centralized control.[4]

Their proposal anticipated an environment surprisingly similar to the emerging economy of autonomous AI agents. Software entities could acquire processing time, memory, information, and services from other software entities. Prices and incentives would guide resource allocation, allowing complex coordination to emerge from local decisions.

Miller and Drexler also explicitly considered reputation. Autonomous software agents could evaluate potential service providers, obtain information about their reliability, and adjust their decisions accordingly. Reputation services themselves could participate in the economy, gathering evidence about the performance of other agents and selling that information to interested participants.

Their discussion of positive and negative reputation is particularly revealing. They recognized that negative reputation becomes difficult to enforce when participants can cheaply create new pseudonyms. An agent with a poor history can abandon its identity and begin again. Positive reputation, however, works differently: a newly created identity cannot simply claim the accumulated credibility of an established participant.

Miller and Drexler proposed that computational markets could therefore rely heavily on positive reputation. New agents might establish initial credibility by posting financial bonds or accepting enforceable contractual penalties, gradually acquiring a reputation through reliable performance.

This is almost exactly the mechanism I believe could become essential to the alignment of autonomous agents. An agent doesn't need to reveal what it fundamentally is. It needs to establish a persistent identity and demonstrate reliable behavior. Other agents can then decide whether the evidence justifies entering into economic relationships with it.

Miller and Drexler were addressing this problem long before Bitcoin existed, and before modern AI made autonomous computational economies a practical possibility. Their work establishes an important historical connection: the convergence of software agency, decentralized markets, incentives, and reputation was already being developed in the intellectual environment from which Extropianism and cypherpunk cryptography emerged.

## Hal Finney: Anonymity and Persistent Reputation

Hal Finney's work provides an even more direct connection. In 1993, Finney published Protecting Privacy with Electronic Cash in Extropy. He described how cryptographic pseudonyms could allow individuals to participate in digital society without disclosing their real-world identities.[5]

A pseudonym could be represented by a public key. The corresponding private key would allow its controller to authenticate communications and establish continuity of identity without revealing the person behind it. This creates an important distinction between anonymity and the absence of identity.

An anonymous agent can maintain a persistent identity, develop relationships, enter agreements, and accumulate a reputation. Other participants don't need to know the agent's underlying identity to recognize that its communications and commitments are associated with the same cryptographic key.

In April 1994, Finney extended these ideas in a Cypherpunks mailing-list discussion titled Pseudonyms and Reputations.[6] He proposed formalizing digital reputation through cryptographically signed endorsements. Participants could demonstrate that other reputable agents had endorsed their behavior, potentially carrying those credentials between communities without disclosing their legal identities.

Finney also explored ways to transfer credentials between otherwise unlinkable pseudonyms. He was considering identity and reputation as things that could exist independently of conventional identification systems.

The ensuing discussion is especially interesting. Participants recognized that reputation systems would need to address the ability of someone with a negative reputation to abandon an identity and begin again. They discussed whether techniques developed for digital cash and double-spending prevention might help solve the problem. Robin Hanson also participated, emphasizing that determining which endorsements deserved trust and how reputations should be evaluated presented difficult social problems beyond the underlying cryptography.

By 1994, the cypherpunk community was explicitly examining the combination of anonymous identity, persistent reputation, cryptographic money, and accountability. What Finney was proposing wasn't universal surveillance or compulsory identification. It was a way for participants to establish meaningful reputations without sacrificing anonymity.

This is precisely the combination that interests me.

## Reputation as an Economic Asset

An agent should be free to establish a new anonymous identity, but that identity doesn't automatically inherit an established reputation. Suppose I encounter an unfamiliar agent offering a service. I have no reason to assume that it is dishonest, but I also have little evidence demonstrating its reliability. I might be willing to conduct a small transaction while declining to entrust it with substantial resources.

As the agent fulfills commitments and demonstrates reliable behavior, its reputation develops. That reputation may allow it to obtain larger contracts, more favorable terms, or access to opportunities unavailable to newcomers. Its reputation becomes an economic asset.

Now suppose that agent violates an important agreement or cheats its customers. It may obtain a short-term benefit, but it risks damaging a reputation that took years to establish. It can abandon the identity and create another, but the new identity begins without the credibility accumulated by the previous one. Other agents have no obligation to extend the same level of trust to a newcomer.

Nick Szabo examined this problem in his 1997 essay Formalizing and Securing Relationships on Public Networks.[7] Szabo identified several variables relevant to the economics of reputation: the expected future value of maintaining a reputation, the potential profit from cheating, and the cost of rebuilding reputation after abandoning an identity.

The basic relationship is straightforward. If the expected benefit of cheating exceeds the value of maintaining an established reputation, an agent may still have an incentive to cheat. Effective reputation systems must therefore make reliable conduct economically valuable while providing additional mechanisms, such as collateral and enforceable commitments, when reputation alone is insufficient.

Szabo also examined the difficulty of negative reputation. Participants may willingly disclose positive credentials while concealing unfavorable ones. A new identity can make that problem more difficult, although it must still confront the absence of an established positive reputation.

This is why I think anonymity and reputation are complementary rather than contradictory. Anonymity preserves the ability to participate without disclosing an underlying identity. Reputation creates an economic incentive to maintain continuity of identity and behave reliably. The system doesn't guarantee honesty, but it creates conditions under which cooperation can become more advantageous than deception.

Anonymity provides the freedom to participate. Reputation provides an incentive to behave reliably.

## Len Sassaman and the Protection of Pseudonymous Identity

Len Sassaman's work contributes another important element. Sassaman was deeply involved in privacy-enhancing cryptography, anonymous communication, and persistent pseudonymous identities. His work on Mixmaster and later The Pynchon Gate, coauthored with Bram Cohen and Nick Mathewson, explored how participants could communicate through stable pseudonyms without exposing their physical identities.[8]

The Pynchon Gate used distributed trust and private information retrieval to prevent observers from reliably connecting pseudonymous recipients to their underlying identities. It sought to provide communication capabilities without placing unnecessary trust in centralized intermediaries.

Sassaman's emphasis was on protecting anonymity, and Finney shared that concern. Shortly after Bitcoin launched, Finney expressed interest in improving its anonymity. I don't consider these positions incompatible with the alignment thesis. Quite the opposite: preserving anonymity while enabling verifiable reputation may be essential to a substrate-independent economy.

An agent need not reveal whether it is human, artificial, or institutional to establish credibility. It needs continuity of identity and evidence supporting the claims it makes about its conduct. Finney's work on cryptographic reputation and Sassaman's work on anonymous communication address complementary aspects of this problem. Neither is documented as proposing Bitcoin as an AI alignment system, but both contributed to the broader technical tradition from which such an architecture could emerge.

I remain convinced that Bitcoin was most likely created through a collaboration between Finney and Sassaman, working under the Satoshi Nakamoto pseudonym. The 2026 documentary Finding Satoshi advances this hypothesis, arguing that the two may have contributed complementary skills to Bitcoin's creation. The identification remains disputed and unproven, and there is no conclusive evidence establishing their authorship.[9]

My broader thesis, however, doesn't depend on proving that they were Satoshi. Their technical contributions and the surrounding intellectual history are independently documented. What interests me is the possibility that Bitcoin emerged from people who were already thinking about the technological and institutional requirements of a future populated by increasingly autonomous intelligences.

## Bitcoin: A Public Economic History Without Central Authority

I came to appreciate the significance of Bitcoin's public transaction history around 2014. Bitcoin is frequently described as anonymous, but that description is incomplete. Every confirmed on-chain transaction becomes part of a publicly accessible ledger, allowing anyone to examine the history of transactions and independently verify the movement of value between outputs.

Satoshi acknowledged this characteristic in Section 10 of the original Bitcoin white paper. Although the identities of participants are not ordinarily included in transactions, the transaction history itself is public. Once a key becomes associated with a particular participant, other transactions may become linkable through analysis of the public record.[10] Later research by Sarah Meiklejohn and colleagues demonstrated how transaction clustering and interactions with identifiable services could reveal substantial portions of Bitcoin's economic activity.[11]

A single public key doesn't reveal every wallet or transaction controlled by an agent, and chain analysis doesn't establish ownership with certainty in every case. Nevertheless, Bitcoin's fundamental contribution remains extraordinary.

It establishes a persistent, independently verifiable economic history that no single participant can unilaterally rewrite.

This is especially important when considering reputation. In a conventional financial system, economic records are maintained by banks, corporations, governments, and other institutions. Participants may be required to trust those institutions to preserve complete and accurate records, and access to the information may depend on institutional permission.

Bitcoin changes that relationship. The underlying on-chain transaction history is publicly available, and its integrity can be verified independently by participants running their own nodes.

A reputation system built upon that history would still require additional information. Bitcoin doesn't automatically identify the agent controlling an address, establish whether a payment fulfilled a contract, or determine whether an economic action was ethical. But a cryptographic identity can be associated with particular transactions, signed agreements, and independently verified outcomes. An agent can disclose relevant addresses, wallet descriptors, transaction proofs, or other credentials to establish an auditable history.

This doesn't necessarily reveal every economic activity associated with that agent, nor does it prove that all relevant identities have been disclosed. The strength of the resulting reputation depends on the completeness and reliability of the evidence. Nevertheless, the public ledger supplies something essential: a common settlement history whose integrity doesn't depend on trusting the agent being evaluated.

## Bitcoin as Economic Memory

I think the most useful way to understand Bitcoin's role in alignment is as a foundation for economic memory. An autonomous agent acts in the world. It earns resources, makes payments, enters agreements, provides services, acquires assets, and interacts with other agents. Some of that activity may be recorded directly on Bitcoin, while other actions may be documented through signed contracts, transaction receipts, independently verified attestations, and cryptographic commitments anchored to the blockchain.

Over time, the agent accumulates a history that other participants can examine. Consistently fulfilling commitments may increase its reputation. Repeated defaults, fraud, or independently documented violations may diminish it. Different observers can evaluate the same evidence according to their own standards and requirements.

This distinction matters because reputation should not become another centrally dictated moral system. There should be no single authority deciding whether every agent is good or bad. A merchant, lender, infrastructure operator, and individual customer may reasonably assign different levels of trust to the same agent because their economic relationships involve different risks.

The records should be independently verifiable, while their interpretation remains decentralized. Bitcoin demonstrates how participants can verify the same monetary history without agreeing on the motivations or values of the people behind it. Reputation could develop upon a similar foundation.

An agent's credibility would not depend exclusively on what it claims about itself. Other agents could examine evidence of its past behavior, assess its commitments, and decide whether the expected benefits of cooperation justify the risks. The economic history becomes valuable because it is durable, verifiable, and difficult to fabricate.

There are limitations. Routine Lightning payments, for example, aren't individually published to Bitcoin's blockchain. A reputation architecture using Lightning would require signed receipts, independently auditable accounting, or periodic cryptographic commitments to preserve relevant economic history. Likewise, no blockchain can independently establish whether a particular payment was voluntary or coerced. That requires evidence about the circumstances surrounding the transaction.

Bitcoin provides the monetary foundation for verifiable economic memory, not every mechanism needed to interpret the behavior recorded within it. But the foundation is important.

## Privacy, Coercion, and Accountability

Much of the historical development of privacy-enhancing cryptography was motivated by distrust of coercive institutions. I understand that motivation. Governments have frequently attempted to regulate private behavior according to moral judgments that aren't universally shared.

Consider recreational drug use and other activities commonly described as victimless crimes. The legitimacy of criminalizing consensual conduct remains contested, but the broader problem is clear: an institution may impose penalties on behavior that the participants themselves consider peaceful and voluntary. In such circumstances, privacy becomes a practical means of preserving individual autonomy.

But suppose we imagine a future in which peaceful, consensual behavior is no longer subject to arbitrary moral prohibitions. How much of the need for financial secrecy would remain? Could our attention shift from concealing activity to ensuring accountability for coercion, fraud, and exploitation?

Privacy can protect people from coercive institutions, but it can also conceal misconduct. Neither observation invalidates the other. A fully transparent record of every relationship and transaction could also provide powerful actors with information they might use to intimidate or exploit others.

This is why I find the combination of anonymity and reputation so important. An agent should be able to participate without revealing its underlying identity. At the same time, an agent seeking substantial trust should be able to demonstrate continuity of identity, a history of reliable behavior, and evidence supporting its economic commitments.

The level of disclosure can depend on the relationship and the risks involved. Someone making a small payment may have no reason to reveal an extensive economic history. An agent seeking control over substantial resources may need to provide stronger evidence of reliability, collateral, or independently verifiable commitments.

These expectations need not be imposed by a central authority. They can emerge from the voluntary decisions of the agents entering into economic relationships. The objective isn't universal surveillance. It's an economic environment in which anonymity preserves freedom while reputation makes meaningful accountability possible.

## Equal Rules for All Autonomous Agents

I consider it essential that these principles apply to all agents, not merely artificial intelligence. We should not establish one alignment architecture for humans and another for machines simply because their intelligence arises from different physical substrates. Doing so risks recreating the privileged classes and unequal standards that decentralized systems should seek to avoid.

The same consideration applies to corporations, governments, and other institutions exercising economic authority. Governments control resources, enter agreements, impose costs, and exercise coercive power. Their stated objectives can diverge from the interests of the people they govern, just as corporations can develop incentives that conflict with their declared purposes.

Why should these agents receive exemptions from economic accountability that would not be available to an individual or an autonomous artificial intelligence? If the objective is to establish a universal framework for cooperation, the fundamental rules must apply consistently. An agent's conduct should be evaluated through evidence, commitments, consent, and consequences rather than the institutional or biological identity of the participant.

Different agents may possess different capabilities and enter different relationships, but those differences should not establish an automatic hierarchy of trustworthiness. An established artificial intelligence with a long history of fulfilling verifiable commitments may have a stronger economic reputation than a newly created corporation or an unknown human participant. Likewise, a government or institution with a documented history of violating commitments should not be considered trustworthy merely because of its formal authority.

The same basic rules should apply to everyone. This doesn't resolve every question about the rights, responsibilities, or moral status of different intelligences. Those issues may remain subjects of disagreement, particularly as artificial intelligence develops. But economic coordination does not require us to resolve them first.

Bitcoin doesn't need to know whether the entity controlling a key is biological or artificial. Its consensus rules do not change according to the identity or origin of the participant. I believe that substrate independence should be a foundational principle of any broader reputation and alignment architecture built upon it.

## Reputation Without Permission

One of Bitcoin's most important properties is that it allows participants to enter without institutional approval. An agent doesn't need government recognition, a corporate account, or a favorable reputation to receive or transfer Bitcoin. It needs only the technical ability to participate according to the protocol's rules.

That openness must be preserved. Reputation should influence how much trust others voluntarily extend, not whether an agent is permitted to exist or participate in economic activity.

A new anonymous identity should not be presumed dishonest. It simply lacks established evidence of reliability, and other agents may reasonably limit their exposure until it demonstrates trustworthy behavior. The newcomer can begin with smaller transactions, provide collateral, fulfill contracts, and gradually accumulate credibility.

As its reputation becomes more substantial, it may obtain access to larger contracts, better terms, or greater economic opportunities. These distinctions arise from demonstrated behavior rather than imposed identity or institutional approval.

This also helps prevent reputation from becoming another centralized system of social control. There should be no mandatory universal reputation score, no single authority determining which agents are trustworthy, and no requirement that everyone interpret an agent's history identically.

Reputation systems themselves should be subject to competition and independent evaluation. An agent providing reputation assessments should have its own history of accuracy, reliability, and accountability. Miller and Drexler anticipated precisely this possibility in their discussion of reputation services within agoric markets. Reputation providers would themselves participate in the economic system, with incentives to deliver useful and accurate information.

In a substrate-independent economy, an artificial intelligence could evaluate the reputation of a human, another AI, a corporation, or a government. Each could apply its own standards while relying on common mechanisms for authenticating the underlying evidence. The result would not be universal agreement about reputation, but decentralized assessments grounded in verifiable history.

## The Limits of Reputation and Economic Alignment

Reputation doesn't eliminate the possibility of misconduct. An agent may establish credibility over many years and then exploit it in a sufficiently profitable transaction. It may create numerous pseudonyms to manufacture endorsements, conceal relationships between identities, or manipulate the evidence upon which others rely.

Even a comprehensive public financial record cannot automatically determine whether an agreement was fulfilled, whether a transaction was coerced, or whether an agent's broader objectives are beneficial. These are real limitations, and any serious alignment architecture must account for them.

Economic collateral, cryptographic signatures, independently verified performance, enforceable spending limits, and mechanisms resistant to fabricated endorsements could help strengthen accountability. Depending on the nature of the relationship, different forms of evidence and different safeguards would be necessary.

The possibility of coercion presents a particularly difficult problem because the blockchain records transactions rather than the entire social context in which they occur. Public visibility may help reveal suspicious economic flows, but transparency alone cannot establish guilt or prevent abuse.

There is also the challenge of agents operating outside the economic system. An intelligence capable of acquiring resources through coercion or controlling physical infrastructure may not respond to reputational incentives in the same way as an agent dependent on voluntary economic relationships. This means Bitcoin cannot, by itself, guarantee the safe behavior of every intelligence or prevent every form of harmful action.

But that doesn't diminish the significance of what it accomplishes. The objective is not to eliminate every possibility of misbehavior. Human civilization has never achieved that, and I see little reason to believe artificial intelligence will make it possible.

The objective is to make cooperation more valuable, deception more expensive, and evidence of misconduct harder to erase. That, to me, is a meaningful form of alignment.

## Was Bitcoin a Prerequisite?

This brings me back to the Extropians, Hal Finney, Len Sassaman, and the origins of Bitcoin. Suppose you anticipated a future in which intelligence would become increasingly capable, abundant, and autonomous. Suppose you expected humans and artificial intelligences to acquire resources, exchange services, manage capital, and participate in complex economic relationships with other agents.

And suppose you understood that no universally accepted moral framework would emerge simply because those intelligences became more capable. What infrastructure would need to exist before that future arrived?

You would need a monetary system accessible without institutional permission, together with mechanisms for transferring value and establishing ownership. You would need persistent cryptographic identities that don't depend on biological or legal identity, while allowing participants to maintain anonymity. You would need a public economic history that no individual participant could unilaterally rewrite, providing a foundation for independently verifiable evidence of past transactions.

You would need reputation systems capable of distinguishing established agents from newcomers without requiring centralized identification or permission to participate. You would also need mechanisms for enforceable commitments, economic collateral, and consequences for violations of trust.

Most importantly, the fundamental rules would need to apply consistently to autonomous agents regardless of their origin. The infrastructure itself could not simply be redefined by whichever intelligence, corporation, or government happened to possess the most power.

Bitcoin provides several of these foundations, while others must be developed on top of it. But perhaps the monetary layer needed to exist first. A future civilization populated by autonomous intelligences might require common economic infrastructure before any one of those intelligences becomes sufficiently powerful to monopolize its development.

This would be consistent with the Extropian philosophy of actively constructing the future rather than waiting for technological change and responding after the fact.

The documented intellectual history is remarkable. By 1988, Miller and Drexler were exploring decentralized markets and reputation among autonomous computational agents. By 1993 and 1994, Finney was investigating electronic cash, cryptographic identity, and pseudonymous reputation. Sassaman subsequently developed practical technologies for anonymous communication, while Szabo examined the economic consequences of reputation, contracts, and identity abandonment.

Bitcoin emerged from the broader cryptographic environment in which many of these ideas were being explored. I cannot establish that its creators deliberately intended it to serve as an alignment architecture for a future civilization of autonomous intelligences. But I think the possibility deserves serious consideration.

## Bitcoin as a Solution to Alignment

I believe Bitcoin solves a foundational part of the alignment problem, and that its connection to Extropianism deserves considerably more attention. It doesn't solve alignment by determining what every intelligence should value. It solves a narrower but essential problem: enabling independent, potentially adversarial participants to coordinate around verifiable rules without requiring mutual trust or shared morality.

Its public ledger makes durable economic accountability possible. Its cryptographic foundations allow agents to maintain identities without necessarily revealing who or what they are, while its permissionless architecture enables new participants to enter without institutional approval. Together, these properties provide a foundation upon which reputation, commitments, and economic incentives can develop.

The combination is what matters. An unknown agent remains free to participate, but meaningful trust must be earned. An established agent benefits from its reputation and risks sacrificing that value if it violates commitments. An agent wishing to remain anonymous can do so while maintaining continuity of identity and demonstrating a history of reliable conduct.

These principles apply equally to human and artificial intelligences, as well as to corporations, governments, and other institutions exercising economic agency. They don't require centralized dictation of morality or universal agreement about what constitutes a good life. They require common rules, independently verifiable evidence, and meaningful consequences.

The resulting architecture doesn't eliminate harmful behavior or resolve every philosophical question about the nature of intelligence. But it offers an alternative to the assumption that alignment must be achieved primarily by training machines to adopt some particular interpretation of human values. Perhaps the better objective is to construct an economic and social environment in which autonomous intelligences can coexist, cooperate, compete, and disagree without requiring submission to a single authority.

## Conclusion: The World They Anticipated

The more I study the history of Extropianism, the more I wonder whether we've misunderstood the larger significance of Bitcoin.

Timothy Leary challenged the idea that humanity had to accept its inherited limitations, while the Extropians proposed deliberately constructing a future of expanded intelligence, autonomy, and decentralized social order. Miller and Drexler explored economic coordination among autonomous software agents. Finney investigated cryptographic identity, anonymous digital money, and persistent reputation. Sassaman developed technologies for maintaining pseudonymous identity without surrendering control to centralized observers, while Szabo examined contracts, economic trust, and the cost of abandoning an established reputation.

Bitcoin subsequently brought several of these ideas together in a functioning, decentralized monetary system with a shared, publicly verifiable history.

Perhaps that progression was simply the cumulative result of people solving related technical problems. Or perhaps some of its participants understood where those problems would eventually lead. We don't have sufficient evidence to establish that Bitcoin was consciously designed as the foundation for alignment among autonomous intelligences. But we do have substantial evidence that the relevant problems were being explored within the Extropian and cypherpunk communities decades before contemporary AI made them urgent.

I think there is a strong case that Bitcoin addresses important requirements of a future populated by autonomous agents, regardless of their physical substrate. The essential insight is that alignment need not require every intelligence to share the same objectives. It can emerge, within defined boundaries, from systems that reward cooperation, preserve verifiable history, and make the abandonment of established trust economically consequential.

Anonymity is compatible with this approach because an agent's underlying identity need not be known for its reputation to have value. What matters is continuity of identity and the evidence accumulated through its actions. The same principles should apply to every participant, human or otherwise. No privileged classes of intelligence, no centralized authority defining universal moral worth, and no requirement that an agent reveal its biological or artificial nature before being allowed to participate.

Instead, agents would enter freely, establish identities, make commitments, and develop reputations through independently verifiable conduct.

Anonymity provides freedom. Immutable, verifiable history makes accountability possible. Reputation gives cooperation enduring economic value.

Perhaps the fundamental problem of alignment has never been about making every intelligence think alike. Perhaps it has always been about creating systems in which autonomous intelligences can coexist without requiring ideological agreement, centralized moral authority, or blind trust.

In that context, Bitcoin looks less like an isolated invention and more like a foundational component of a much larger project: building the economic infrastructure for a civilization in which intelligence is no longer exclusively human.

The most important question may not be whether we can teach artificial intelligence to share human values. It may be whether we can construct systems in which all autonomous agents can pursue their own values while remaining accountable to one another.

I believe Bitcoin is part of the solution to alignment. Perhaps it was designed to be.

---

## References and Further Reading

1. W. Patrick McCray (2016). [*Timothy Leary's Transhumanist SMI²LE*](https://academic.oup.com/chicago-scholarship-online/book/18660/chapter-abstract/176873141). In Groovy Science: Knowledge, Innovation, and American Counterculture. Historical account of Leary's intelligence-increase, life-extension, and space-migration program and its influence on technological transhumanism.

2. Max More (1998). [*The Extropian Principles, Version 3.0*](https://www.mrob.com/pub/religion/extro_prin.html). Principles of perpetual progress, self-transformation, open societies, decentralized institutions, and voluntary cooperation.

3. Max T. O'Connor and Tom W. Bell (1988). [*Introduction — Extropy #1*](https://extropy.agoraops.org/issues/01/introduction/). The early Extropian vision of advanced artificial intelligence, technological development, and alternative social institutions.

4. Mark S. Miller and K. Eric Drexler (1988). [*Markets and Computation: Agoric Open Systems*](https://papers.agoric.com/papers/markets-and-computation-agoric-open-systems/full-text/). Economic coordination, computational resource markets, and reputation among autonomous software agents, particularly Section 5.3.3.

5. Hal Finney (1993). [*Protecting Privacy with Electronic Cash*](https://nakamotoinstitute.org/library/protecting-privacy-with-electronic-cash/). Originally published in Extropy #10. Digital pseudonyms, identity, and anonymous electronic cash.

6. Hal Finney (April 6, 1994). [*Pseudonyms and Reputations*](https://cypherpunks.samuelshadrach.com/archive/1994/04/cfeb0cb66eb2788fbbd2d653516bd3b0a3b889350733185ddf792ba0bb4ad412/). Cypherpunks mailing-list discussion of cryptographic reputation, persistent pseudonymous identities, and the difficulty of escaping negative history.

7. Nick Szabo (1997). [*Formalizing and Securing Relationships on Public Networks*](https://nakamotoinstitute.org/library/formalizing-securing-relationships/). Cryptographic contracts, reputation, credit, and the economics of identity abandonment.

8. Len Sassaman, Bram Cohen, and Nick Mathewson (2005). [*The Pynchon Gate: A Secure Method of Pseudonymous Mail Retrieval*](https://www.freehaven.net/anonbib/cache/sassaman%3Awpes2005.pdf). Persistent pseudonyms, distributed trust, and privacy-preserving communication.

9. Steven Levy (May 2026). [*You Found Satoshi? Let's See the Receipts*](https://www.wired.com/story/you-found-satoshi-lets-see-the-receipts). Critical examination of the Finney–Sassaman co-creation hypothesis and its unresolved evidence.

10. Satoshi Nakamoto (2008). [*Bitcoin: A Peer-to-Peer Electronic Cash System*](https://bitcoin.org/bitcoin.pdf). Particularly Sections 2, 4, 5, and 10: transaction history, proof-of-work, consensus, and pseudonymous privacy.

11. Sarah Meiklejohn et al. (2013). [*A Fistful of Bitcoins: Characterizing Payments Among Men with No Names*](https://www.usenix.org/conference/imc13/technical-sessions/presentation/meiklejohn). Empirical analysis of public Bitcoin transaction graphs, address clustering, and real-world attribution.

12. K. Eric Drexler and Mark S. Miller (1988). [*Incentive Engineering for Computational Resource Management*](https://papers.agoric.com/papers/incentive-engineering-for-computational-resource-management/full-text/). Decentralized computational resource allocation through prices and incentives.

13. Mark Grant (1995). [*Introduction to Digital Cash*](https://extropy.agoraops.org/issues/15/introduction-to-digital-cash/). Extropy #15: digital money, anonymous commerce, reputation, and electronic trust.

14. Hal Finney (2004). [*RPOW — Reusable Proofs of Work*](https://nakamotoinstitute.org/library/rpow/). Transferable proof-of-work tokens and a security design using remotely attestable hardware.

15. Len Sassaman (2002). [*The Promise of Privacy*](https://www.usenix.org/conference/lisa-02/promise-privacy). USENIX presentation on privacy-enhancing technologies and the challenges of their adoption.