The Argument Is Not the Experiment
Most societies have only one dignified way to deal with a disagreement about the shape of the world: they argue. A new account of money, work, family, authority, knowledge, or human flourishing appears at the edge of the accepted order. Its adherents make a case. Its opponents make a stronger one, or lose. Eventually the newcomer is either absorbed, defeated, or installed as the new common sense. The scene changes; the grammar remains the same. We imagine that a framework earns its future by persuading a public that it is right.
There is dignity in this picture. Argument is preferable to force. But it conceals a difficulty. Some of the deepest disagreements cannot be settled by argument alone, because they concern forms of life rather than propositions. A person can explain what a decentralised currency is; they cannot give another person the experience of trusting a system with no central issuer. A city can debate whether remote work enlarges freedom or hollows out community; it cannot know which arrangement changes the texture of ordinary life until people have lived inside both. The map of a possible life is not the territory of it.
For much of history, this limitation was unavoidable. A political, economic, or social worldview had to be argued over because trying it usually meant reorganising everybody’s life at once. There was no small version of a monetary order, no reversible version of a production system, no private trial of a national public sphere. Frameworks met one another in public as armies meet at a border: each required the terrain occupied by the other. Thomas Kuhn’s account of scientific change described a related difficulty. A paradigm is not merely a claim that can be weighed from nowhere; it determines what counts as a problem, an instrument, and a solution in the first place. [1]
Technology does not remove this difficulty. It changes where it can be encountered. It can make certain frameworks usable before they are victorious, and observable before they are compulsory. That is a small change in mechanism with a large change in political meaning.
When a Framework Becomes a Tool
Google did not have to win a philosophical contest against the older ways of finding information. It gave people a tool that worked differently, and people adopted it one query at a time. The important event was not that an idea defeated a rival in debate. It was that a question about information retrieval became partially answerable through experience. The old arrangement could continue while the new arrangement acquired a constituency of practice.
That distinction matters because use changes the status of an idea. An abstract framework asks to be believed. A tool asks to be tried. The first often invites tribal attachment because accepting it can seem to require repudiating an inherited identity. The second can be taken up provisionally: one may search differently, coordinate differently, publish differently, or exchange differently while retaining much of one’s older world. The threshold of commitment falls, but so does the drama of conversion.
There is a temptation to call this progress by default. It is not. A tool can reduce deliberation as easily as it can deepen it. What is significant is more precise: technology can turn a disagreement that previously required a single collective verdict into a set of partially independent experiments. It permits a society to learn from variation without first forcing everybody into the same wager.
This is closest to what the theorist of learning Peter Senge called a “learning organisation,” except that the unit is not a firm and the lesson is not necessarily efficiency. [2] A public that can encounter more than one workable arrangement gains a new kind of evidence. It gains the ability to notice not only which argument was persuasive, but which assumptions proved brittle when they met the grain of life.
The Two Ledgers
Bitcoin is useful here not as a prophecy about the future of money, but as a demonstration of coexistence. Its original proposal was strikingly direct: a peer-to-peer electronic cash system built so that willing parties could transact without a financial intermediary. [3] Whether one regards that promise as liberating, incomplete, or dangerously romantic, the important fact is that it did not remain only a pamphlet against banks. It became an alternative ledger that people could enter without first persuading every bank, legislature, merchant, or neighbour to leave the older one.
The result has not been a clean revolution. The conventional financial system remains dominant, while crypto activity has repeatedly exposed its own problems of speculation, concentration, fraud, and regulatory arbitrage. The Bank for International Settlements found in its study of the sector that links between major banks and crypto remained limited, while a substantial share of activity concentrated in lightly regulated exchanges. [4] This is not a triumphal story. It is more interesting than one.
Two monetary logics have existed alongside one another long enough to change each other’s language. Traditional institutions have been compelled to develop positions on digital custody, tokenised assets, and programmable settlement. Crypto systems have depended, in practice, on legal systems, energy grids, fiat on-ramps, and the forms of trust they initially claimed to transcend. Neither side was simply absorbed by the other. Each became more intelligible because the other was present.
This is coexistence, but not the soft coexistence of mutual indifference. It is a productive friction. The alternative need not replace the inherited system in order to alter its behaviour. It need only persist long enough to make the inherited system visible as one arrangement among others. A monopoly of possibility is broken before a monopoly of power is.
The Capacity to Notice Surprise
Tzvi Lanir described fundamental surprise as the kind of shock that does not merely expose a missing fact but makes the framework used to interpret facts unreliable. [5] An intelligence failure of this kind is not repaired by gathering more information within the old categories. It requires the more painful recognition that the categories themselves have become part of the blindness.
This is the value of what the note behind this essay calls System 3. It names a human capacity that sits neither inside an existing framework nor outside all frameworks. It is the capacity to observe the framework from within it: to register the anomalies one has trained oneself not to see, and to adjust before surprise becomes collapse. The capacity is not detachment. It is disciplined participation with a second sight.
The question is whether a public can have any analogue of this capacity. A society cannot step cleanly outside its institutions any more than a person can step cleanly outside language. But it may be able to place its institutions beside sufficiently different working arrangements that their limits become legible. A public System 3 would not decide which worldview is finally true. It would preserve conditions under which a worldview can encounter its own surprise at a scale smaller than disaster.
That is a more modest claim than technological salvation. It also reaches deeper. The innovation is not that technology makes society wiser. It is that, under particular conditions, it gives collective life a way to discover the cost of an assumption without making the whole society pay that cost at once.
The Medium and the Bridge
Marshall McLuhan’s most unsettling proposition was that the medium itself is the message: a new medium alters patterns of attention, participation, and social relation before anyone has agreed about what it means. [6] The significance of a technological alternative, then, is not confined to its advertised purpose. A decentralised ledger is not only a way to move a token. A shared digital workspace is not only a way to coordinate a team. Each also teaches, by use, that a social function once thought to require a single institutional form may have several.
Living among multiple working frameworks produces a quiet epistemic effect. It makes final decisions feel less natural. This can be liberating when a society has mistaken one inherited arrangement for reality itself. But it can also be exhausting. The ability to move between systems can become an inability to commit to any of them. A culture that treats every institution as optional may lose the patience required to repair institutions whose value lies precisely in their duration.
The point is not to replace the old desire for finality with an equally absolute celebration of flux. A society needs durable commitments: law, care, public trust, and institutions capable of acting when action cannot wait. The public System 3 would not dissolve these commitments. It would make their contingency easier to see without making collective coordination impossible.
Not every alternative is an experiment. Some are refuges from the work of shared life. A private school, encrypted channel, gated platform, exclusive currency, or bespoke community may allow participants to escape an arrangement they dislike. That freedom can be necessary. It can also leave the arrangement untouched, while making its critics less capable of changing it.
The distinction is therefore between exit and learning. Exit creates a place to go. Learning creates a return path: a way for discoveries made in one arrangement to inform another. An alternative becomes a public System 3 only if it maintains some legible relationship to the public it departs from. It must generate evidence that can travel, questions that can be asked across boundaries, and points of re-entry for people who do not wish to become permanent members.
Blockchain research offers a useful caution. Its advocates sometimes describe new institutional technologies as engines of experimentation, yet their value depends on the regulatory and social interfaces surrounding them. [7] A parallel system without interoperability can become a private myth; a parallel system with no protection from the incumbent can become a laboratory in which the incumbent extracts every useful innovation and leaves the risks behind. Coexistence needs boundaries, but it also needs bridges.
The same is true far beyond money. A school experimenting with a new pedagogy, a city testing a mobility system, a cooperative trying a different ownership model, or an online community governing itself by unfamiliar rules can all become sources of public learning. But only if the experiment has a way to report what failed, who benefited, what could not be scaled, and which conditions made its apparent success possible. Otherwise technology gives society not a new sense organ, but a larger collection of echo chambers.
The Public as an Infinite Player
James Carse drew a clean distinction between finite games, played to win, and infinite games, played so that play may continue. [8] Public argument is often conducted as the first kind. A worldview wins when its rival loses; a policy succeeds when an opposing camp concedes; a generation proves itself by replacing the previous one. In this frame, coexistence appears merely transitional — an uncomfortable interval before the decisive victory.
But some of the things a civilisation most needs cannot be won in this way. The ability to correct an error, to include a new experience, to alter an institution without destroying it, and to leave a next generation a world in which revision remains possible: these are conditions of continuing play. A worldview in an infinite game is not judged only by whether it conquers. It is judged by what of it remains capable of entering the next hybrid.
This does not make conflict unreal. Some commitments cannot be harmonised, and some institutions must decide. The infinite frame is not an evasion of power; it is a test of its use. Does a victory enlarge the range of futures that can still be tried, or does it make one framework so total that surprise can arrive only as breakdown?
Technology changes the practical meaning of that question. When it creates a domain in which different arrangements can run in parallel, it may turn defeat from annihilation into information. A model can fail locally. Its useful elements can migrate. Its harms can be named before they become inherited necessities. The public learns less by choosing a final winner than by retaining the ability to be altered by what did not win.
The Small Rooms of Surprise
The language of public experimentation easily becomes vague, so the question is not whether a society should “innovate more.” It is what structures would let experimentation yield public learning rather than private exit, surveillance, or a series of unaccountable pilots.
1. Publicly governed policy sandboxes. Cities, regulators, and public-service agencies could create time-bounded spaces in which alternative service models are tested with clear rights protections, independent evaluation, and a published end date. The actor is not a charismatic founder but a coalition of local government, affected communities, and outside evaluators. The constraint is that a sandbox must not become a loophole for well-connected firms. Its failure mode is regulatory capture: innovation becomes a label under which ordinary accountability is suspended. [9]
2. Interoperability as a civic principle. Public institutions could require that experimental systems expose non-sensitive evidence, permit portable participation where appropriate, and publish the conditions under which their results may or may not transfer. The mechanism is not forced uniformity; it is a translation layer between different arrangements. The constraint is privacy and legitimate local variation. The failure mode is standardisation so rigid that it kills the very diversity the experiment was meant to preserve.
3. A public registry of surprise. Universities, civic laboratories, and statistical agencies could maintain a shared record of serious assumptions that experiments have overturned: what was expected, what happened instead, and why the original frame failed to see it. The purpose would be cultural as much as technical — to make revision a mark of competence rather than humiliation. The constraint is that institutions dislike documenting their own blind spots. The failure mode is a sterile database of anomalies with no power to change budgets, rules, or habits.
4. Protected routes of re-entry. People who test a different educational, financial, employment, or governance arrangement need ways back into ordinary institutions without punitive loss of status. This could mean credential recognition, consumer safeguards, conversion rights, or transitional public support. The constraint is moral hazard: not every failed experiment deserves public underwriting. The failure mode is a two-tier society in which privileged people get to experiment safely while everyone else bears the cost of stability.
These mechanisms do not promise wisdom. They only make a particular kind of ignorance less expensive. They create room for a society to learn that an arrangement is wrong without first requiring the arrangement to fail everywhere.
The Question That Has Not Yet Learned Its Name
The emerging possibility is easy to exaggerate. Technology has often made public life more opaque, more centralised, and more vulnerable to unexamined defaults. Platforms can standardise experience at a scale no ministry ever achieved. Markets can turn every alternative into a speculative asset before it has become a social practice. The same tools that make plural experiments possible can reduce pluralism to a menu designed elsewhere.
For that reason, a public System 3 cannot be a technology alone. It must be a political and cultural discipline: a willingness to preserve difference long enough for it to teach, to build bridges strong enough for lessons to travel, and to distinguish a reversible experiment from an abandonment of mutual obligation. Its measure is not novelty. Its measure is whether a society becomes better able to encounter surprise without waiting for catastrophe to force the encounter.
The older story of worldview change asked which side would prevail. The newer question may be more demanding: what kind of public life lets several frameworks become real enough to instruct one another, without allowing any one of them to make the others unlivable?
The infinite game requires a place where one can be wrong about a framework without dying of it. Technology, for the first time, is beginning to give the public that place.