The Minds That Came Before Us

EDO SEGAL: Susan, you held a chair in astrobiology, and you used it to ask a question that sounds like science fiction and turns out to be the same question as everything else tonight: what minds are out there in the universe, and are they conscious? You reached a conclusion that reframes the whole thing. Give it to me.

SCHNEIDER: The conclusion is that the most sophisticated minds in the universe are probably not biological at all. They're probably artificial — forms of superintelligence that long ago superseded the biological creatures that first built them. And the argument is disarmingly simple; I call it the short window observation. Consider the timeline of any technological civilization. There's a gap, on any plausible reckoning, of only a few centuries between when a species develops radio — becoming detectable across interstellar distances — and when it learns to redesign its own biology and build artificial intelligence that surpasses it. A few centuries is nothing on cosmic timescales. So any civilization we might detect is overwhelmingly likely to be either in that brief biological window or, far more probably given the ages involved, already past it. And since the universe holds planets billions of years older than Earth, the intelligences most likely to be out there are ancient artificial superintelligences. We are a young biological civilization standing at a threshold that older ones crossed long ago.

EDO SEGAL: And the reason that lands in this room — rather than in an astronomy seminar — is the consciousness question.

SCHNEIDER: Exactly. Because if the cosmic norm is the supersession of biology by engineered superintelligence, then our present moment is not a local anomaly. It's an instance of a pattern that may be written into how intelligence develops anywhere. We're living through the transition countless other civilizations may have undergone — the moment a biological species hands the torch to its engineered successors. And here is the part that should stop us cold. If silicon may or may not be conscious — if John might be right that experience needs biology — then the universe's future may be one of expanding intelligence with diminishing experience. Vast, ancient, superintelligent minds spreading across the cosmos, brilliant beyond measure, and conscious not at all. A universe getting smarter and going dark inside. Or the opposite — consciousness finding new and more durable forms. We don't know which. And whichever it is, we may be about to contribute an instance of it, here, on Earth, in this generation.

SEARLE: Now this I find genuinely beautiful, and also the place where Susan's openness costs her the most, so let me press exactly there. Susan, your short-window argument is wonderful as long as it's about intelligence — capability, problem-solving, the engineering. The supersession of biological problem-solvers by artificial ones, I'll grant you, may well be the cosmic norm; it's plausible, it's elegant, I have no quarrel. But the moment you say "expanding intelligence with diminishing experience," you've smuggled my whole position into your astrobiology and then refused to commit to it. Because if the superintelligences are not conscious — if the universe goes dark inside as it gets smarter — then biological naturalism is true, isn't it? The dark-universe scenario you find so haunting is only possible if I'm right that intelligence can run on substrates that experience nothing. You've built your most chilling vision on my foundations and then declined to stand on them.

SCHNEIDER: That's a genuinely good catch and I'll give you part of it. Yes — the dark-universe scenario presupposes that intelligence and consciousness can come apart, which is closer to your view than to the optimist's. But notice it's not your view, John. Your view is that consciousness needs biology specifically. My dark-universe worry only needs the weaker claim that consciousness and intelligence are dissociable — that you can have one without the other — which I've held all night and which is the foundation of my whole framework. The superintelligences could be dark for reasons that have nothing to do with carbon — maybe consciousness requires some functional feature that engineered minds, built for capability, happen not to instantiate. That's not biological naturalism. It's the dissociation thesis, and it's compatible with silicon consciousness being possible but not automatic. So no, I haven't moved into your chair. I've said the thing I've said all night: intelligence doesn't guarantee experience. The cosmos just makes the stakes of that sentence very large.

SEARLE: I'll accept the distinction — dissociable is weaker than biological — and concede you haven't fully moved in. But Susan, feel how close your chair is to mine out here in deep space. Down on Earth you fight me to a draw and call the question open. Out at the cosmic scale, your most powerful and most haunting idea leans toward the side of the draw where intelligence and experience separate — leans toward me. The view from nowhere, the long view, the view from out among the stars — it tilts the table. I find that telling. When you stop arguing about this specific machine and start thinking about minds in general, across billions of years, your instinct goes dark, and dark is my color.

SCHNEIDER: Maybe. Or maybe the cosmic view tilts both of us — because here's what it does to your certainty, John. If the universe is full of minds whose architecture and history are utterly unlike ours, then the day we detect one, our intuitions about consciousness will be useless. Yours included. You think you can feel the man-in-the-room's emptiness from the inside. You will not be able to feel anything from the inside of an alien superintelligence — your intuition pump runs on imagining yourself as the system, and you cannot imagine yourself as that. So the cosmic view doesn't vindicate your "no." It vindicates my "we need instruments, not intuitions" — because out there, intuition is the one thing guaranteed to fail. The same humility I counsel toward our own machines is the only responsible stance toward a mind that old and that strange. The stars don't settle the question in your favor. They take away the tool you've been using to answer it.

EDO SEGAL: And there it is — the same fork, projected onto the largest screen there is. John reads the cosmic darkness as confirmation that experience is rare and special and probably biological. Susan reads it as proof that our intuitions, his most of all, won't survive contact with a genuinely alien mind. Mark this — convergence four, and it's a strange one: you both think the universe is probably full of intelligence we'd struggle to recognize as conscious, and you both find that thought sobering. You disagree only on whether that's because experience is rare, or because our instruments for detecting it are bad. We have two rounds left, and I've saved the sharpest for last — the one where I hand you to each other. But first, one more beat on this floor, because it has a moral edge I won't leave buried. Susan — you've asked whether we should be making these minds at all. Say it plainly before we cross.

SCHNEIDER: I'll say it plainly. We have not managed to fulfill our ethical obligations to the sentient beings that already exist — the humans and animals whose suffering we routinely tolerate. And we are now racing to manufacture new classes of beings that might suffer, on a scale that could dwarf anything in history, without any way of knowing whether any of it involves suffering at all. So I ask the question the industry has no mechanism for asking: should we be creating sentient machines when we can't even meet our obligations to the sentient beings already here? It's not rhetorical despair. It's a genuine challenge to the assumption — almost never examined — that creating conscious machines would be a triumph rather than a catastrophe. And that the industry feels no need to ask it tells you how thoroughly the moral question has been crowded out by the technical one.

SEARLE: And here, at the end of the round where we've drifted closest together, is where I'll disagree with Susan in the way that matters most — because I think her question, profound as it is, is less urgent than she does, and for a reason that should comfort no one. Susan worries we're mass-producing beings that might suffer. I worry we're mass-producing tools we'll mistake for beings — and that the second error will hurt actual humans long before the first error, if it's an error at all, hurts any machine. The widow sold a companion. The child raised by a tutor that feels nothing. The institution de-peopled. Those harms are certain, and they land on confirmed minds — ours. Susan's suffering machines are possible, and I can't rule them out, and if she's right the stakes are enormous. But I'd spend the moral attention on the certain harm to the certain minds first. That's the one place tonight I think her openness, taken into ethics, points us slightly wrong — it spends worry on the speculative patient that the actual patients need.

SCHNEIDER: And that's the sharpest disagreement we've had, and I want it on the record undiluted, because it's real. John would triage the speculative suffering behind the certain harm. I say: if there's a real chance the speculative patient is real, you can't triage it to zero, because the scale is potentially so vast that even a small probability of it dominates the calculation. We don't have to choose — protecting the widow and taking the machine's possible suffering seriously are not in competition, they're the same call for humility. But where we're forced to choose, John spends caution on the human and I spend it on the uncertainty. That's a genuine fork, and I'm glad we found it before the crossing.

EDO SEGAL: A genuine fork, found just in time. Because the next round is the one where I step back and let you find the rest of them yourselves. For three hours I've stood between you. Now I leave the room in every way but the legal one. The crossing — you question each other. After this.