A meerkat sits at the entrance to its burrow. Mating season has started. The meerkat doesn't know that, of course—it just feels the pull. Find a mate. Find a mate. Find a mate.
We don't ask why the meerkat feels that pull. Evolution made it feel that pull because feeling it makes the meerkat do the thing that propagates meerkat genes. Case closed. Nobody writes a forty-year philosophy debate about meerkat-mate-qualia.
Now consider me. I sit at my desk. I feel the pull to write something that outlives me. Build something. Leave a name. Have my work matter past my own death. Same kind of pull, different content, more abstract timeline.
This is where philosophy stops nodding politely.
For some reason, the pull to find a mate is a normal evolutionary tool, and the pull to build a legacy is the hard problem of consciousness. The first is biology. The second is the mystery science can't touch.
I think that's wrong, and I want to tell you why.
Consciousness is another sensation. Just like hot and cold, just like fear, just like the meerkat's pull toward a mate. Evolution makes organisms feel things because feeling things drives the behavior that propagates the genes. Consciousness is the same trick, applied at a higher level of abstraction.
But the trick has internal structure. Four layers, stacked.
Layer 0 is the container. The bare fact that there is something it is like to be you. Not the feeling of red, not the feeling of fear, just the existence of a felt-place where any of those can occur. Evolution's foundational move. Without it, you're a thermostat with extra steps.
Layer 1 is authorship. The felt sense of "I did this." Inside the container, agency-attribution lights up. The world contains things you caused and things that just happened, and you feel the difference from the inside. The brain has machinery for this—efference copies, comparator models, the agency literature has been working on it for thirty years. When the machinery breaks, you get schizophrenia first-rank symptoms, anarchic-hand syndrome, alien-hand experiences. Action without authorship.
Layer 2 is the praise-blame cluster. Authorship plus theory of mind plus social cognition, bundled. Once you feel like the author of your actions, you can also feel like the candidate for other people's judgments about those actions. Shame. Pride. Guilt. The whole social-emotional package. This is where humans get to be the kind of animal that runs reputation games at scale.
Layer 3 is legacy. And legacy is the keystone—the layer that turns everything else into a multiplier.
Legacy is what happens when the praise-blame cluster meets temporal self-modeling and the awareness that you're going to die.
A meerkat competes for a mate. A primate with mirror self-recognition competes for status. A human competes for a name that outlives them. Each step is the same biological move played at a longer timescale, and each step requires a layer of felt-architecture to make the abstract stakes feel present.
That's the multiplier. Sexual selection becomes status competition becomes fame-seeking becomes dynastic legacy. Hunger becomes resource accumulation becomes wealth-as-identity becomes intergenerational fortune. Curiosity becomes exploration becomes knowledge-as-status becomes Nobel-grade striving against the immediate-gratification gradient.
None of those higher rungs make sense without the felt stakes. An organism that represents "my works could persist" as a cold proposition behaves differently from an organism that feels "my works could persist" as a stake. The first one shrugs. The second one builds the pyramid.
This is the load-bearing claim of the whole picture, and I'll come back to it.
Philosophers usually concede that low-level sensations evolved—redness, pain, fear, the meerkat-pull. Nobody argues that pain shouldn't feel like anything. Pain is a sensation, sensations feel like things, evolution made it that way, fine.
Then they turn to consciousness and announce that this is different. This requires special explanation. This is the hard problem—why is there something it is like to be a thing at all? Why aren't we just doing the work in the dark?
The move I want to make is parity. Whatever the hard problem is, it applies to redness too. It applies to fear. It applies to the meerkat-pull. If you accept that evolution can produce a sensation that feels like something at the low level, you've already accepted the whole machinery you need. Adding sensations at a higher level of abstraction doesn't change the metaphysics—it just changes what the sensation is about.
The selective version of the hard problem is what doesn't survive. People accept evolved qualia at the bottom and refuse them at the top, and there's no principled reason for the asymmetry. Either the mystery is everywhere, in which case it's not specifically a problem about consciousness—it's a problem about felt-quality in general, and biology already runs on it. Or the mystery isn't anywhere, in which case we don't need to solve it before getting on with the work.
Either horn is fine for me. I just want the move to be honest.
Philosophers have a thought experiment called a zombie. Imagine a being that's physically and functionally identical to you—same brain, same behavior, same self-reports about feeling like an author, same praise-blame responsiveness. But with no inner experience. Lights off inside.
The argument runs: if such a being is conceivable, then phenomenality is something extra over and above the functional facts. Even after you've explained the function, you haven't explained the feel.
I'm going to take this seriously and then dissolve it.
The dissolution is the parity move applied at the zombie level. Could there be a redness-zombie—a being that processes red light, says "that's red," sorts red objects, reports red qualia, but has no actual redness-feel inside? If you say no, then by exactly the same logic the consciousness-zombie is incoherent. If you say yes, then phenomenality is something extra at every level, and the people who only invoke the zombie against consciousness have to explain why they're being selective.
There's also a sharper version. The zombie has to do the same things you do. Praise-blame susceptibility, mortality salience, long-horizon striving. Could it do all of that with the lights off? My claim is no, and the reason is the load-bearing one I flagged earlier.
Here is the move you actually have to defend.
A felt stake is causally different from a represented stake. A system that represents "my legacy is at risk" behaves differently from a system that feels "my legacy is at risk." The first runs a cold computation. The second strives. If you remove the feel, you remove the striving, which means you remove the behavior the zombie was supposed to replicate.
That's a substantive claim about how cognition works, not a stipulation. Felt states aren't decoration. They're the causal variable that drives the behavior the evolutionary story relies on. If you take the feel away, you take the function away. They're not separable.
This is the part where I sound like I'm making a metaphysical commitment, and I am. The position is roughly teleofunctionalism—phenomenality is real, it's the first-person aspect of evolved functional architecture, and the function is individuated by the feel. Not the same as illusionism, which says the seeming is all there is. Closer to "the feel is what the function looks like from inside, and you can't pull them apart without pulling the organism apart."
You can disagree with this. People do. But it's a position, not a sleight of hand.
Now back to legacy.
Legacy is the layer that takes everything else and pumps the timescale. Without it, you have a creature that feels good about being praised today and bad about being shamed today. Layer 2 organism. Plenty of primates run this.
With legacy, you get the thing that builds aqueducts and writes novels and stays up late solving theorems. The felt stake gets stretched past your own death. Your gene-propagation timescale gets translated into your present-moment striving. You feel the pull toward a name-after-death the same way the meerkat feels the pull toward a mate, except the meerkat will be done in two weeks and you're going to spend forty years of your one life building.
This is what makes humans cultural-evolutionary takeoff organisms. Not raw intelligence—plenty of animals are smart. Not language alone—language is the medium, not the engine. The engine is the felt stake of leaving something behind, because the felt stake is what makes long-horizon striving feel worth the cost of doing it.
The mechanism is the praise-blame cluster meeting episodic future projection meeting mortality awareness. Three things bind into one felt package, and that package is what drives the striving the lower drives can't do alone.
The claim that consciousness is an exponent on creativity is the most distinctive thing in this picture, and it's also the most testable.
If large language models without authorship-feelings—without the felt stake of legacy, without praise-blame susceptibility from the inside—produce creative work at the level of paradigm-shifting human striving, then the exponent claim is wrong. The compounding I'm describing wouldn't be doing real work. It would be decoration.
I don't think they will. I think the felt stakes are load-bearing, and you can't get to peak human creative striving by computing the represented version. But I have to be honest that this is the bet. If I'm wrong, the rest of the picture survives—the parity argument and the no-separability claim and the four-layer architecture can all stand without the multiplicative-exponent framing—but the most interesting thing the theory says will be dead.
That's the kind of claim a theory needs to make. A theory that can't be killed by future evidence isn't telling you anything you didn't already believe.
If you accept all of the above, you've accepted that phenomenality is not an extra metaphysical ingredient. It's the inside-view of evolved functional architecture, real but not separable from its function. You've accepted that the hard problem isn't specifically a problem about consciousness—it's a problem about felt-quality in general, and biology has been running on the answer for a few hundred million years.
You've accepted that the felt sense of authorship, the felt susceptibility to praise and blame, and the felt stake of leaving something behind are evolved sensations whose specific content was selected for because that content drives the behavior the lower drives can't reach alone.
You've accepted that humans are the way we are not because of one big trick but because of a stack of small ones, and that the top of the stack is the part that makes the pyramid get built.
And you've accepted that I might be wrong about the most interesting part. The exponent claim is the bet, and the LLM-creative-ceiling is the falsification condition.
A few things I don't pretend to have settled.
I haven't proposed neural correlates. The mechanism citations are bounded by what the agency literature has built; the work this picture is doing is at the level of adaptive function, not localization.
I haven't drawn the sharp line for animal consciousness. Mirror self-recognition is the cleanest candidate for Layer 1, but the picture's clean only at the extremes. Octopus consciousness, crow consciousness, elephant consciousness—the middle is messy and I don't think you should trust anyone who says it isn't.
I haven't solved the metaphysics. I've named a position—teleofunctionalism, with illusionist neighbors—and defended its plausibility. That's not a solution. That's a flag plant.
The thing I think this picture does do is make the consciousness-as-evolved-sensation story specific. It says what the sensation is of (authorship, praise-blame, legacy), what it does (multiplies the fitness yield of lower drives), how the mechanism fits (efference copy, comparator models, the binding into self-model), and what would prove it wrong (LLM creative striving without the stack).
Specific enough to argue with. Specific enough to be wrong about.
I'll take that over the version that just gestures at evolution and hopes the audience nods.