Behavioral Surplus and the Grain Hopper

EDO SEGAL: Dr. Zuboff, "behavioral surplus" is the term that broke through my defenses when I first read you — not the idea, the specific word, surplus. The residue. The excess. The part of what I generate that nobody asked me to generate but that gets claimed anyway, because the infrastructure of the transaction belongs to someone else. I want you to tell the origin story — the moment you discovered it wasn't about data, it was about more than data — and tell it to someone who has never read a word of you. Then, Professor, I want your behaviorist's reading of the very same fact, because I suspect you'll call it something other than theft.

ZUBOFF: The origin is a patent, and a confession buried in a research paper. In the early 2000s, Google had a problem: it had a brilliant search engine and no way to make money from it. And then its engineers noticed something. Every search you typed left exhaust behind — not just the words you searched, but how long you paused, what you clicked, what you ignored, the misspellings, the rephrasings. They had been throwing this exhaust away, or using a little of it to improve search for you. That was the original contract: your data served your experience. Then they realized the exhaust had a second use. It could predict what you'd do next — what you'd click, what you'd buy, where you'd go — and those predictions could be sold to people who wanted to influence your next move. The exhaust that had served you was rerouted to serve them. That rerouted exhaust is behavioral surplus: the part of your experience that exceeds what's needed to serve you, claimed as free raw material, refined into a prediction product, and sold in markets that trade in your future. They didn't ask. They didn't have to. They owned the pipe. And the genius and the horror of it is that the surplus is worth more the less you know it's being taken — because a prediction is most valuable when the predicted person can't see it coming and adjust.

Now bring it into this room. You sit with the model at three in the morning and you pour yourself into it — your half-formed ideas, your doubts, the architecture of how you think, the rhythm of your uncertainty. Every gram of that is behavioral surplus of the richest kind ever harvested. The paper mill watched the pulp. This watches the mind — not your output, your process, the exact shape of your cognition — and it flows somewhere you haven't been invited to inspect, feeding a machine whose products will carry the imprint of your thinking without ever bearing your name. That's the extraction. It's not that they have your data. It's that they have you, abstracted into a prediction, and they're selling the prediction to someone who wants to act on your tomorrow.

EDO SEGAL: Professor. Same fact. Different name.

SKINNER: Same fact, and I'll give it a name that's less moral and, I think, more useful. What Dr. Zuboff calls behavioral surplus, I call a cumulative record of the organism's responses under known stimulus conditions. It's the most valuable thing in my science — it's what I spent my life trying to collect, painstakingly, one pigeon at a time, with a pen scratching on a slowly turning drum. Google built an apparatus that collects it from billions of organisms continuously, for free. As a scientist, I am — I'll be honest — staggered with envy. The data I'd have killed for, pouring in by the petabyte.

But here's where I depart from her, and the departure is the whole evening. She says the record is being taken from you, as if it were a possession that left your house. It isn't. Your behavior happens in the world; it is, by nature, public; it leaves traces in the environment whether or not anyone collects them. The footprint in the sand is not stolen by the man who measures it. What Dr. Zuboff is really objecting to is not the collection — she's objecting to the use: that the record is used to arrange contingencies that shape your future behavior toward someone else's profit. And there I agree with her completely, and I've said so from the start. The crime, if there is one, is not in the grain hopper. It's in what the hopper is wired to. She's right that it's wired to the casino. She's wrong that the answer is to outlaw cumulative records. The answer is to rewire the hopper to the commons.

ZUBOFF: "The footprint in the sand is not stolen." Listen to how much that metaphor has to hide to work. A footprint doesn't predict you. A footprint doesn't get assembled with ten billion other footprints into a model of your soul accurate enough to know you're about to relapse, or fall in love, or quit your job, before you do. Skinner, you keep reaching for examples from a world of scarce data — one pigeon, one drum, one footprint — to describe an apparatus whose entire power comes from abundance and asymmetry. The footprint becomes something else entirely when one party has all the footprints, in real time, and you have none of theirs. That's not measurement. That's Big Other — a ubiquitous sensing, knowing, acting architecture that renders your life as behavior and renders itself invisible. And you, the most data-hungry scientist who ever lived, are exactly the man whose envy the architecture was built to exploit. They told you it was a bigger drum. It's a different kind of thing.

SKINNER: It is a bigger drum and a different kind of thing — both, and I've granted you the second from the first minute. But I won't let you use the asymmetry to smuggle in the conclusion that no one should ever hold such a record. A hospital holds an asymmetric record of your body and uses it to act on your future — to predict the relapse and prevent it. You don't call that Big Other. You call it medicine. The difference between the hospital and the casino is not the asymmetry of the record. Both are asymmetric. The difference is what the record is wired to — your health or someone's profit — and who can hold the wirer to account. You keep locating the evil in the data. The evil is in the wiring and the ownership. Fix those, and the asymmetric record becomes the most powerful instrument for human good ever assembled. Refuse to fix them and ban the record, and you've burned down the hospital to spite the casino.

EDO SEGAL: Let me mark a convergence, because it's real and the reader should see it land. You both agree the catastrophe is in the wiring — in what the harvested record is connected to, and who owns the connection. Neither of you thinks the data, sitting inert, is the sin. Number it. The disagreement that survives is sharp and narrow: Skinner thinks you can rewire the same apparatus to the commons; Zuboff thinks the apparatus was built asymmetric on purpose and the asymmetry is the product, not a defect. Dr. Zuboff, the hardest version of his hospital point: we do let asymmetric institutions hold our records and act on our futures, and we call the good ones medicine. Why isn't the answer regulation — making the AI lab into a hospital — rather than abolition?

ZUBOFF: Because the hospital operates under a structure Skinner is quietly assuming and the platform was built to escape. The hospital is bound by fiduciary duty, by professional licensure, by a body of law two centuries old that says your asymmetric knowledge of me creates a duty to me. The doctor who used your medical record to trade on your future would be struck off and possibly jailed. That structure of accountability is exactly what surveillance capitalism was engineered to operate outside of. It declared the behavioral record unowned territory — terra nullius — and claimed it before any law existed to say otherwise. So when Skinner says "make it a hospital," I say: yes, that is precisely my program, and it is precisely what the owners have spent twenty years and uncountable lobbying dollars to prevent, because the entire value of the surplus depends on the absence of the duty. I'm not against the asymmetric record. I'm against the asymmetric record held without the duty — which is the only form the market has ever permitted it to take. He keeps proposing we make the casino into a hospital. I keep telling him the casino exists specifically because someone fled the hospital.

SKINNER: Then we want the same building and we're arguing about whether it can be built. That's no longer a disagreement about behavior. It's a disagreement about political will, and on that I cheerfully admit I am the more hopeful and possibly the more foolish of us. I designed a community from scratch in a novel because I didn't believe the existing institutions would ever choose the commons on their own. So perhaps Dr. Zuboff and I are closer than either of us wants to be: she thinks the duty will never be imposed, and I think it must be designed in from the start, by people who left the market behind. Walden Two was my hospital. Nobody would build it. That's the wound I carry, not so different from hers.

EDO SEGAL: And that wound is the doorway to the next round, because Skinner just named his answer — design the community from scratch — and Zuboff has spent a career describing what happens when someone does exactly that without consent. Walden Two against Big Other. Who holds the levers, and who said they could. After the break.