Donald Hoffman studies perception at the University of California, Irvine. His work makes two claims that change the order of explanation. Evolution shaped human perception to hide the world, not to show it. Behind the hidden world stand conscious agents, and space, time, and physical objects form the interface through which their activity appears.
Hoffman reaches consciousness primacy through mathematics and evolutionary theory. Bernardo Kastrup reaches the same ground through analytic philosophy, and the Hermetic tradition reached it through practice. Hoffman supplies the formal model. He gives the rendered world a theorem, a definition of the agent who receives it, and a proposed bridge to physics.
Biography
Hoffman was born on December 29, 1955, in San Antonio, Texas. He earned a bachelor’s degree in quantitative psychology at UCLA in 1978. He completed a PhD in computational psychology at MIT in 1983, under David Marr and Whitman Richards. Marr’s program treated vision as computation that a theorist can specify exactly.
Hoffman briefly worked as a research scientist at MIT’s Artificial Intelligence Laboratory. He joined the cognitive sciences faculty at UC Irvine in 1983 and is now professor emeritus there. His early research on visual shape, parts, and motion earned the National Academy of Sciences’ Troland Research Award. His 1998 book Visual Intelligence explains how the visual system constructs what a person sees.
The constructive account led Hoffman to a sharper question. If vision builds its scenes, the scenes need not resemble whatever supplies the signal. His 2015 TED talk, “Do we see reality as it is?”, brought that question to a general audience. The Case Against Reality (2019) gave it a book-length argument. In 2026 he became scientific director of the Trace Research Institute, which studies conscious agents and their traces.
The Interface Theory of Perception
Most models of vision assume that perception estimates true properties of the physical world. Hoffman, Manish Singh, and Chetan Prakash rejected that assumption in “The Interface Theory of Perception” (2015). Perception works like a computer desktop. Its icons help the user act while hiding the circuitry, voltages, and software that do the work.
A blue file icon does not show a blue, rectangular file inside the computer. The icon’s color, shape, and position serve the user’s tasks. The user still must take the icon seriously: dragging it to the trash deletes real work. The interface has real consequences while its forms reveal nothing about the structure beneath them.
Hoffman applies this relation to every perceived object. A snake, an apple, and a cliff are species-specific icons. Each guides behavior that carries fitness consequences. None displays the reality that the interface represents. Human perception gives a usable world, not a window.
The interface theory explains why a successful species can be wrong about the nature of its world. Accuracy and usefulness separate under selection. A perceiver can act effectively for a lifetime inside icons it mistakes for the things themselves. Plato’s cave describes the same condition: prisoners succeed at naming shadows without seeing what casts them.
Fitness Beats Truth
Hoffman tested the interface claim with evolutionary game theory. In “Fitness Beats Truth in the Evolution of Perception,” Prakash, Hoffman, and colleagues compared two kinds of perceptual strategy. A truth strategy estimates the actual state of the world. A fitness strategy tracks only the payoffs that states carry for the organism.
Their theorem shows that fitness strategies generically outcompete truth strategies in the evolutionary models the authors defined. Payoffs depend on the organism, its state, and its action, and they rarely mirror the world’s structure. Water, for example, helps a thirsty animal up to a point and kills it in excess. A sense that tracks the payoff serves the animal better than one that tracks the quantity.
The result reverses an inherited assumption. Natural selection does not favor accurate perception merely because accurate perception sounds useful. Selection favors perception tuned to fitness, and fitness-tuned perception hides the truth. Hoffman summarizes the consequence directly: evolution hid the truth from our eyes.
Philosopher Jonathan Cohen answered the 2015 article in the same journal. He argued that Hoffman’s team used too narrow a definition of veridical perception. Hoffman’s group replied in the same issue. The dispute concerns what counts as truth-tracking; the formal result about payoff-tuned strategies remains in place.
Hoffman applies the theorem to its own foundation. Evolution shaped perceptual icons for fitness. Logic and mathematics remain tools for testing claims about the interface, and Hoffman uses them to build his alternative. The theory therefore leaves the perceptual world unreliable as a picture while keeping reason available as an instrument.
Conscious Realism
If space and objects are interface, something else must stand behind them. Hoffman’s answer is conscious realism. Hoffman and Prakash state it as a hypothesis in “Objects of Consciousness” (2014): “The world W consists entirely of conscious agents”. Physical particles and fields become icons through which agents appear to one another.
A conscious agent has a precise mathematical form. It has a space of possible experiences and a space of possible actions. Three Markov kernels connect them: perception carries the world into experience; decision carries experience into action; action carries action back into the world. A kernel assigns probabilities to transitions between states. A counter orders the agent’s cycles.
The definition makes experience a primitive. Hoffman does not derive consciousness from neurons, because neurons belong to the interface. A brain is what one agent’s activity looks like through another agent’s icons. The hard problem of consciousness disappears in this direction. The remaining problem is the reverse one: deriving the physical interface from agents.
Agents combine. Two interacting agents form a larger agent with its own experiences and actions. Hoffman calls this combination fusion. A network of agents can form a single higher agent without losing its constituents. The model gives the old idea of nested minds a formal structure.
The Parliament of Consciousness describes many agencies competing within one human experience. Conscious realism gives that parliament a mathematical form. Each member is an agent, and their coalitions are fused agents. The self that a person experiences becomes one fusion among the agents that compose it.
Spacetime Is Not Fundamental
Hoffman takes support from a movement within high-energy physics. Physicists including Nima Arkani-Hamed argue that spacetime cannot be fundamental. In 2013, Arkani-Hamed and Jaroslav Trnka described the amplituhedron, a geometric object outside spacetime. Its volume computes certain particle scattering amplitudes. Spacetime and quantum mechanics then appear as consequences of a deeper structure.
“Fusions of Consciousness” (2023), by Hoffman, Prakash, and Robert Prentner, links that structure to conscious agents. The authors map the long-run dynamics of agent networks onto decorated permutations. Decorated permutations are combinatorial objects that physicists use to classify positive geometries such as the amplituhedron. A particle in spacetime becomes a projection of the dynamics of a communicating class of agents.
The program aims to derive physics from consciousness and test the derivation. The program aims to derive particle properties such as mass and spin from agent dynamics. Hoffman treats spacetime as a data structure, a compact format through which agents represent one another’s dynamics. Physics then studies the interface with great precision while the interface still hides its source.
The Holographic Principle reaches a related conclusion through gravity: spacetime’s interior can be encoded on a lower-dimensional boundary. Both programs move spacetime from the foundation to an emergent description. Hoffman goes further and names what the description describes: interacting consciousness.
Traces of the Other
In 2026, Hoffman joined Andrew Gallimore and Niffe Hermansson in a preprint titled Traces of the Other. The paper asks whether DMT entities are independent conscious agents. A trace is another agent’s activity as it enters the perceiving agent’s experience. The trace need not resemble the agent that produces it.
Ordinary human perception registers only part of the agent network. Agents whose activity falls outside that range act without appearing. DMT changes the perceiver’s kernels, the transition rules that determine which activity becomes experience. Activity that ordinary perception filters out then gains a perceptible form.
The authors propose experiments using extended-state DMT, a controlled infusion that stabilizes the experience for hours. Their tests ask whether the encounter carries information that varies with external conditions or correlates across separate observers. Strassman and Gallimore on DMT follows the formal model in detail. The Pharmakon places the preprint within the long history of entheogenic contact.
The preprint completes a line that interface theory opened. If perception filters agents, the filter can change. A changed filter discloses inhabitants that the consensus interface keeps out of view. Hoffman’s mathematics thus reaches the ground the traditions mapped through vision, ritual, and contact.
The Contested Interface
Hoffman explains why the world appears as it does. Consensus Reality adds the question of who maintains the appearance. A desktop has designers, defaults, and permissions. A shared perceptual interface likewise carries inherited settings, social reinforcement, and institutional custody over which icons a population trusts.
Hoffman’s evolutionary account treats fitness as the interface’s designer. The account extends further. Fitness payoffs belong to the field in which agents act, and other agents shape that field. Some agents benefit when a population mistakes its icons for the whole of reality. The Parasitic Ecology describes agencies that feed through the interface they help stabilize.
Hoffman’s agent network is neutral in its formal statement. Its mathematics specifies interaction, not intention. The same structure accommodates cooperative, indifferent, and predatory agents. Gnostic tradition named the predatory case: archons administer the perceptual prison and profit from its closure. Conscious realism supplies the architecture in which that claim becomes precise.
The interface also explains the work of liberation. A person cannot leave the desktop by smashing icons. Release comes through recognizing the icons as icons and acting with knowledge of the agents behind them. The Rendered World follows that recognition into ordinary perception. Hoffman gives it a theorem: the world each person sees was shaped to serve survival, and survival is not truth.
Key Works
- Observer Mechanics: A Formal Theory of Perception (1989), with Bruce Bennett and Chetan Prakash, gives the first formal model of the perceiving observer.
- Visual Intelligence: How We Create What We See (1998) explains the rules by which vision constructs shape, color, and motion.
- “Conscious Realism and the Mind-Body Problem” (2008) states conscious realism as a solution to the mind-body problem.
- “Objects of Consciousness” (2014), with Prakash, defines the conscious agent and proposes how objects and spacetime emerge from agent dynamics.
- “The Interface Theory of Perception” (2015), with Singh and Prakash, gives the interface account and replies to its critics.
- The Case Against Reality: Why Evolution Hid the Truth from Our Eyes (2019) presents the full argument for general readers.
- “Fitness Beats Truth in the Evolution of Perception” (2021), with Prakash and colleagues, states the theorem formally.
- “Fusions of Consciousness” (2023), with Prakash and Prentner, connects agent dynamics to the positive geometries of particle physics.
References
Bennett, Bruce M., Donald D. Hoffman, and Chetan Prakash. Observer Mechanics: A Formal Theory of Perception. Academic Press, 1989.
Cohen, Jonathan. “Perceptual Representation, Veridicality, and the Interface Theory of Perception.” Psychonomic Bulletin & Review 22 (2015): 1512–1518. https://doi.org/10.3758/s13423-014-0782-3
Gallimore, Andrew R., Donald D. Hoffman, and Niffe Hermansson. “Traces of the Other: Are DMT Entities Real? DMT Phenomenology in the Framework of Conscious Realism.” PsyArXiv preprint, 2026. https://doi.org/10.31234/osf.io/8qvgy_v2
Hoffman, Donald D. Visual Intelligence: How We Create What We See. W. W. Norton, 1998.
Hoffman, Donald D. “Conscious Realism and the Mind-Body Problem.” Mind and Matter 6, no. 1 (2008): 87–121.
Hoffman, Donald D. The Case Against Reality: Why Evolution Hid the Truth from Our Eyes. W. W. Norton, 2019.
Hoffman, Donald D., and Chetan Prakash. “Objects of Consciousness.” Frontiers in Psychology 5 (2014): 577. https://doi.org/10.3389/fpsyg.2014.00577
Hoffman, Donald D., Chetan Prakash, and Robert Prentner. “Fusions of Consciousness.” Entropy 25, no. 1 (2023): 129. https://doi.org/10.3390/e25010129
Hoffman, Donald D., Manish Singh, and Chetan Prakash. “The Interface Theory of Perception.” Psychonomic Bulletin & Review 22 (2015): 1480–1506.
Prakash, Chetan, Kyle D. Stephens, Donald D. Hoffman, Manish Singh, and Chris Fields. “Fitness Beats Truth in the Evolution of Perception.” Acta Biotheoretica 69, no. 3 (2021): 319–341. https://doi.org/10.1007/s10441-020-09400-0
Arkani-Hamed, Nima, and Jaroslav Trnka. “The Amplituhedron.” Journal of High Energy Physics 2014, no. 10 (2014): 30. https://arxiv.org/abs/1312.2007
Hoffman, Donald D. “Do We See Reality as It Is?” TED, 2015.