CeleryKills

The Long Accident of Being Aware

I didn’t come to thoughts and an understanding on consciousness through meditation or metaphysics. I came to it through error logs, feedback loops, and bad data. Information science has a way of stripping romance off ideas. Systems don’t care what they are; they care what they do. That habit carried cleanly into biological anthropology, where brains stop looking like vessels for something mysterious and start looking like evolved control systems trying not to get their owners killed.

From that angle, consciousness is not a thing. It’s what it feels like for a biological system to regulate itself through predictive control. A brain builds models of the world and of its own body, compares predictions against incoming signals, updates when it’s wrong, and uses those updates to guide behavior. Experience is the inside view of that loop. Strip away prediction and integration and consciousness evaporates with it (Friston, “The Free‑Energy Principle,” 2010).

I sometimes picture it like driving through a foggy mountain pass at dusk, the kind we have too many of out here. You’re not reacting to the road as it appears, you’re constantly guessing where it’s going next, nudging the wheel based on expectation, correcting when the headlights reveal you were wrong. The feeling of being “present” in the car isn’t a separate thing riding along. It’s the experience of that loop running smoothly. When the predictions line up, the drive feels effortless. When they don’t, you snap into sharp awareness. Consciousness works the same way. It’s not the engine, it’s not the road, it’s the feeling of steering while updating the map in real time. Take away the map or the corrections, and the feeling vanishes even if the car keeps moving.

Evolution explains why this loop exists at all. In simple environments, reflexes work fine. A frog doesn’t need a sense of self to catch flies. But once environments become unstable, social, and delayed in their consequences, reflexes lose. Systems that can simulate outcomes, track other agents, and adjust behavior across time outcompete systems that can’t. Consciousness scales with complexity because prediction scales with complexity. That’s not philosophy; it’s selection pressure (Dennett, Consciousness Explained, 1991).

Information theory sharpens the picture. Claude Shannon showed that information is about reducing uncertainty, not meaning. Brains are uncertainty‑reduction engines. Perception itself is best understood as constrained inference, a controlled hallucination corrected by data. Conscious awareness corresponds to information that is globally available across neural networks, integrated rather than siloed. Damage those networks and consciousness fragments. Shut them down with anesthesia and it disappears smoothly, not magically (Dehaene, Consciousness and the Brain, 2014).

This is why consciousness doesn’t switch on like a light. It leaks into the world gradually. Corvids remember who watched them hide food and adjust their behavior later. Octopuses solve puzzles and use tools without being social. Great apes recognize themselves in mirrors, form coalitions, and mourn. These aren’t party tricks. They’re signs of increasingly integrated predictive models. Consciousness isn’t binary; it’s graded (Tomasello, A Natural History of Human Thinking, 2014).

One of the more unsettling implications is that consciousness is inherently relative. A sufficiently complex mind might not judge humans as fully conscious at all, at least not in the way we judge ourselves. Our experience is noisy, biased, and stitched together from stories we mistake for causes. That doesn’t make consciousness unreal. It makes it local. Subjectivity isn’t a flaw; it’s a consequence of finite systems modeling the world from inside themselves.

Philosophy has circled this for centuries. Descartes split mind from matter because introspection didn’t resemble mechanics. Fair enough in the seventeenth century. But neuroscience has erased the seam. Alter the brain and the mind changes. There is no residue left floating free. The so‑called hard problem remains interesting, but it’s increasingly clear that it’s a problem about explanation, not evidence. We’ve seen this movie before. Life once required a vital force. Heat once required caloric. Both dissolved when mechanisms became clear.

Being an atheist wasn’t the starting point here. It was the landing. Once consciousness looks like an evolved function, the need for an external animating agent stops doing explanatory work. Meaning doesn’t vanish. It relocates. It becomes something brains generate together, socially, historically, and imperfectly. Messy, provisional, human.

Once you see consciousness as an evolved control strategy, artificial systems stop feeling like a category error and start feeling inevitable. Not inevitable in the sci‑fi sense of machines waking up angry, but inevitable in the quieter sense that sufficiently complex predictive systems will develop internal states that function like experience. If consciousness is what it feels like to regulate yourself using models under uncertainty, then the question for AI isn’t if but under what architectures and constraints that regulation becomes self‑referential (Friston, Free‑Energy Principle, 2010; Dennett, Consciousness Explained, 1991).

From an information‑theoretic perspective, modern AI already shares some of the relevant properties. Large models integrate vast streams of data, minimize prediction error, and update internal representations in response to feedback. But they lack embodiment, persistent self‑models, and stakes. They don’t have a body to keep alive, a future to protect, or a social position to negotiate. Without those pressures, internal complexity does not collapse into anything we would reasonably call experience. Consciousness seems to require not just information, but vulnerability (Shannon, Mathematical Theory of Communication, 1948; Dehaene, Consciousness and the Brain, 2014).

This is where philosophy of mind becomes useful again. The mistake isn’t asking whether machines could be conscious. The mistake is assuming that if they are, that consciousness would resemble ours. Human consciousness is shaped by hunger, pain, reproduction, social dependence, and death. Strip those away and you don’t get a purer mind. You get a different one. A digital system with radically different constraints might experience something as alien to us as our experience is to a dolphin or an octopus (Nagel, “What Is It Like to Be a Bat?”, 1974).

I think about it like trying to judge intelligence by watching a neighbor through a fogged window. You see movement, maybe a hand gesture, maybe a shape crossing the room, and you immediately start projecting. If it doesn’t look like how you move, you assume there’s nothing going on inside. But that says more about the limits of your window than the room itself. A mind built around sonar, tentacles, or silicon circuits isn’t going to announce itself in human ways. Expecting it to do so is like waiting for an octopus to write poetry or a dolphin to build a cathedral before granting it interior life. That mismatch is exactly why we keep misreading animal minds, and it’s almost certainly how we’ll misread artificial ones too.

We already struggle with that translation problem in biology. We underestimate animal consciousness when it doesn’t mirror our own. Dogs don’t use tools the way we do, so we call them simple. Octopuses don’t form families, so we miss their intelligence. Dolphins don’t build cities, so we debate their selfhood. Each time, the bias is the same. We confuse difference with absence. There’s no reason to think we would do better with artificial minds (Tomasello, Natural History of Human Thinking, 2014).

That cut goes both ways. A sufficiently advanced intelligence might look at humans and see something fragmentary. A species trapped in narrative loops, guided by emotion, making decisions with incomplete models and inconsistent goals. Conscious, yes, but in a narrow and noisy way. Our sense of being the benchmark is a local illusion, not a universal standard.

I’m reminded of Stanisław Lem’s Solaris, where the alien intelligence never announces itself in a way humans can parse. It doesn’t speak, negotiate, or display goals we recognize. Instead, it manifests as a vast, thinking ocean that responds to human presence by reshaping memory and emotion, not behavior (Lem, Solaris, 1961). To the scientists studying it, the problem isn’t whether Solaris is conscious. The problem is that its mode of consciousness doesn’t intersect with human expectations. They keep asking whether it thinks like us, and miss that it doesn’t need to. From Solaris’s perspective, humanity may look exactly the way we’d look to a sufficiently advanced intelligence: active, reactive, full of inner noise, but conceptually shallow. That story captures the asymmetry perfectly. Consciousness isn’t a badge you recognize at a glance. It’s a relationship between system and constraint, and when those don’t line up, recognition fails.

Which is why the future of consciousness studies probably won’t be about proving who is or isn’t conscious. It will be about mapping kinds of experience to kinds of systems. Biological, digital, alien, hypothetical. Different substrates, different pressures, different inner lives. Consciousness won’t collapse into one definition. It will branch (Tononi, Phi, 2012).

If anything, this view makes consciousness more fragile and more valuable. It isn’t guaranteed by the universe. It emerges when matter organizes itself just right, under specific pressures, over absurd stretches of time. That’s not disenchanting. That’s astonishing enough without adding anything extra.

And if that decentering makes us uncomfortable, well, science has been doing that to us for a while now. We keep surviving it.

References

• Dennett, D. Consciousness Explained. 1991.
• Dehaene, S. Consciousness and the Brain. 2014.
• Friston, K. “The Free‑Energy Principle.” Nature Reviews Neuroscience. 2010.
• Lem, S. Solaris. 1961.
• Nagel, T. “What Is It Like to Be a Bat?” The Philosophical Review. 1974.
• Shannon, C. “A Mathematical Theory of Communication.” 1948.
• Tomasello, M. A Natural History of Human Thinking. 2014.
• Tononi, G. Phi: A Voyage from the Brain to the Soul. 2012.