Perhaps Time Is Not What Reality Contains, But What Finite Minds Must Construct
July 28, 2026
For centuries we have treated time as one of the unquestionable ingredients of reality, a thing that flows relentlessly from past to future while carrying stars, civilizations, and every human life along with it, and we have measured it with astonishing precision, built entire scientific disciplines upon it, and organised almost every corner of our existence around its apparent passage.
Yet one of the more intriguing developments in modern science is that time itself has quietly become one of the least settled concepts in all of physics, which does not for a moment mean that time has been disproven, but rather that an increasingly interesting question has begun to surface, namely what happens if time is not a fundamental property of reality but instead part of the representational geometry that every finite observer must construct simply in order to navigate an incomparably richer universe.
Notice how sharply that question differs from the familiar claim that time is an illusion, because illusions imply deception whereas representations imply necessity, and that single distinction ends up changing almost everything about how we approach the problem.
Every Observer Lives Behind an Interface
One assumption quietly underlies nearly every discussion of reality, which is the comfortable belief that we experience the world directly, and yet modern neuroscience suggests something rather different, since the brain never actually observes the external world itself but instead receives electrochemical signals generated by specialised sensory systems, combines those signals with prior expectations, compresses enormous quantities of incoming information, and then constructs a representation coherent enough to guide action.
What we experience, in other words, is not reality in its full richness but reality after extraordinary amounts of selection, filtering, and prediction have already taken place, and Karl Friston’s Free Energy Principle along with the broader framework of predictive processing both converge on precisely this idea, treating perception not as passive observation but as active inference in which the brain continually predicts the world and revises those predictions only when reality supplies enough evidence that its current model has grown inadequate.
Evolution arrives at a strikingly similar conclusion from a different direction, because natural selection never rewarded organisms for constructing perfectly accurate models of reality but rather rewarded those that survived long enough to reproduce, and Donald Hoffman’s controversial Interface Theory of Perception pushes this line of reasoning even further by proposing that evolution favours useful interfaces over objective representations, so that whether or not one accepts Hoffman entirely, the underlying insight remains compelling, since organisms require models that enable effective action rather than exhaustive descriptions of the universe.
Artificial intelligence reaches a remarkably similar destination by yet another route entirely, because large language models never encounter reality directly but only the statistical relationships buried within language, and their internal representations are compressed structures built to organise vast quantities of information into forms useful for prediction, so that their success depends not on possessing reality but on constructing representations coherent enough to act under real informational constraints.
Three independent disciplines therefore land on three remarkably similar conclusions, all of them pointing toward the same uncomfortable truth that every finite observer operates through an interface and that none of them enjoys direct, unmediated access to reality itself.
Compression Is Not Information Loss
Compression is one of the most misunderstood ideas in this whole conversation, because most people picture it as nothing more than throwing information away, and yet that is only half the story, since compression also creates structure, and it is worth sitting with that second half for a moment.
Imagine standing before an infinitely detailed landscape in which every blade of grass, every dust particle, every photon, and every molecular interaction exists simultaneously, a scene that contains vastly more information than any biological brain could ever hope to represent, so that in order to act at all the observer is forced to construct something dramatically simpler, resolving that overwhelming richness into mountains, trees, sky, a river, and a path.
What has happened here is not merely that information has been discarded but that a coordinate system has been constructed, because the observer now inhabits an organised geometry capable of supporting prediction, movement, and decision-making, which points toward a deeper principle worth stating plainly, that compression does not merely reduce reality but actively builds the coordinate system through which reality becomes intelligible to finite observers, and that observation may in the end prove even more important than compression itself, because once we accept that coordinate systems are constructed rather than simply handed to us, a great many assumptions about perception start to look considerably less secure.
What If Time Belongs to the Interface?
General Relativity describes spacetime as a four-dimensional geometric structure in which time and space become inseparable aspects of the same manifold, and within Einstein’s theory time remains entirely real, yet physics has never reached anything like unanimous agreement over whether spacetime itself is truly fundamental.
Julian Barbour has argued that time may emerge from relationships among configurations rather than existing independently; Carlo Rovelli’s relational interpretation similarly treats many physical quantities, temporal order among them, as arising from relationships rather than existing absolutely; and several approaches to quantum gravity go further still by suggesting that spacetime may emerge from deeper informational or quantum structures we do not yet fully understand.
None of these ideas represents established scientific consensus, and that genuinely matters, because the distinction between possibility and proof should never be quietly blurred, yet the mere existence of these approaches demonstrates something important, which is that modern physics no longer forces us to treat time as unquestionably fundamental but instead leaves room for the possibility that something deeper generates our experience of temporal flow.
If every observer already experiences reality through a compressed interface, then another possibility emerges quite naturally, namely that time belongs primarily to that interface, not because time is unreal but because sequential organisation may simply be the most efficient representational strategy available to finite systems trying to navigate an overwhelmingly information-rich universe, so that time under this view begins to look less like an external river carrying us along and more like the coordinate axis through which finite observers organise change.
Flatland Revisited
Edwin Abbott’s Flatland remains one of the most illuminating thought experiments ever written, inviting us to imagine intelligent creatures living entirely within two spatial dimensions, for whom length and width exhaust the whole of reality while height possesses no meaning at all, since their representational system contains no coordinate capable of even expressing it.
Now suppose a three-dimensional sphere passes through Flatland, and the inhabitants would observe something extraordinary unfolding before them, as a point appears, expands into a circle, swells to its maximum size, and then gradually contracts until it vanishes altogether, so that to the Flatlanders something deeply mysterious has clearly happened, even though to us nothing mysterious occurred at all, because they were merely watching successive cross-sections of a higher-dimensional object their representational geometry could never fully encode.
The sphere did not deceive them, since it was their interface that limited what could be represented, and the analogy grows genuinely uncomfortable the moment we turn it back on ourselves, because if our cognitive architecture evolved under real informational constraints, then why on earth should we assume our own coordinate system captures every dimension relevant to reality, and perhaps we too experience only successive slices of something considerably richer, not because higher dimensions have been proven but simply because finite representational systems inevitably possess boundaries.
The Observer’s Geometry
This line of thinking shifts the discussion away from time travel almost entirely, because popular conversations tend to fixate on whether we might travel backward through time, whereas a far more fundamental question asks whether we have even correctly identified what time actually is in the first place.
Suppose some other form of intelligence processed information differently, so that a hypothetical four-dimensional observer might perceive what we call an entire lifetime as a single coherent structure rather than a sequence unfolding moment by moment, and what appears to us as birth, childhood, adulthood, and death might amount to nothing more than different regions of one geometric object viewed all at once, requiring nothing supernatural whatsoever but only a different representational geometry, which suggests that our own sequential experience may reveal far more about the architecture of finite cognition than about the deepest structure of reality itself.
Whether such observers could actually exist remains genuinely unknown, and yet the thought experiment still exposes an important principle, namely that observation depends upon representation, representation depends upon compression, and compression depends in turn upon finite computation.
The Larger Invariant
Notice what has quietly happened over the course of this argument, because the discussion has expanded far beyond time alone, and the very same logic now applies with equal force to space, causality, identity, objects, and even probability, since every finite observer confronts the identical fundamental problem, which is that reality contains vastly more structure than can ever be represented completely.
Selection therefore becomes unavoidable, compression becomes unavoidable, and representation becomes unavoidable, so that agency itself becomes possible only because those earlier stages have already succeeded, and this invariant shows up again and again across neuroscience, evolutionary biology, artificial intelligence, information theory, and increasingly within theoretical physics itself, so that beneath a whole range of different disciplines we keep finding one recurring architecture, all of it flowing from the simple fact that finite systems cannot contain infinite information and that every intelligent system must therefore organise before it can ever hope to understand.
A Different Way of Asking Questions
Perhaps this also helps explain why scientific revolutions so often feel less like discovering entirely new worlds and more like reorganising familiar ones, because Copernicus did not create the Solar System, Darwin did not create evolution, and Einstein did not create spacetime, and each of them instead reorganised relationships that had always existed yet had previously remained hidden inside less coherent representations, so that reality itself changed remarkably little while our representation of it changed dramatically.
Perhaps our understanding of time will eventually travel that same road, and perhaps it will not, but either outcome leaves the deeper principle entirely untouched, because the purpose of inquiry was never to defend particular coordinate systems but to discover progressively more coherent ones, and that may ultimately be the most important lesson emerging from this whole framework.
The deepest question, then, is no longer whether time is fundamental but whether time, space, causality, identity, and the very objects populating our experience are features of reality itself or features of the representational geometry every finite observer must construct before intelligence, prediction, and action even become possible, and that question remains genuinely unanswered, which is precisely what makes it one of the most interesting questions science has ever been capable of asking.
