The Distraction and the Trouble

**EDO SEGAL:** Before I hand you to each other, one round on the loudest noise in our whole conversation — the one neither of you makes, which is itself telling. Half the public discourse about AI is about existential risk: a future superintelligence, misaligned, ending humanity. Enormous researchers, enormous money, enormous attention. Meredith, you've built a sustained argument that this framing is, whatever the intentions behind it, a distraction. Make it. And Donna — I want you to tell me whether your refusal of both the sunrise *and* the apocalypse is the same move as hers or a different one, because I suspect it's different in a way that matters.

**WHITTAKER:** Here's the argument, and it has a structural blade in it that I want sharp. The existential framing says: the real danger is a hypothetical future superintelligence. Notice what that does. It directs all the attention, money, and regulatory energy toward a speculative future harm — and away from the concrete, present, unequally distributed harms these systems are producing *right now*, in the service of concentrated power. Algorithmic benefits denial. Predictive policing. Workplace surveillance. The sorting of migrants. And notice *who* is harmed by those present systems: disproportionately the less powerful, the people who are not in the room. The existential framing implicitly says we don't need to treat AI's danger as urgent until it threatens the *most privileged* — the people who aren't currently harmed, the people who'd only be threatened by a superintelligence that comes for everyone. There's a quiet hierarchy of whose suffering counts as an emergency, and the existential frame puts the powerful at the top of it. And — this is the part that matters most — the framing grants the technology an *autonomy it does not have*. To fear that AI will "escape human control" is to imagine it as an independent agent with its own will. But it isn't. It's a product, firmly under the control of a handful of profit-driven corporations. The real risk was never that the machine escapes human control. It's that it's working *exactly* as its controllers intend, and its controllers are five companies whose interest is not the public good. The doom narrative is, in significant part, a marketing narrative — it tells you these firms are building something so powerful it could end the world, which is a wonderful thing to imply about your own product to investors and regulators alike. The machine is not magic, and it is not a god, and it is not going to wake up. It is [concentrated power](https://www.youonai.ai/fieldguide/med/concentration_of_power), and the apocalypse talk is the smoke it hides in.

**EDO SEGAL:** Donna. You also refuse the apocalypse — and you refuse the sunrise. Same move?

**HARAWAY:** Adjacent, and importantly different, and the difference is the most useful thing I can offer the reader tonight. Meredith refuses the apocalypse because it's a *distraction from present power* — it points your fear at a fictional future to keep you from organizing against an actual present. I agree with every word of that, and I'd add my own version: the apocalypse and the sunrise are the *same story*, the story of a singular dramatic ending, salvation or damnation, and both of them are ways of *not staying with the trouble*. The doomer and the booster are twins — they both believe in The Event, the threshold, the moment after which everything is different and the question is settled. And the function of believing in The Event is that it relieves you of the daily, situated, unglamorous work of the present. If the singularity is coming, why tend the compost pile? If extinction is coming, why fight for the data workers' union? Both eschatologies are anesthetics. So where Meredith says the existential frame is a distraction *from power*, I say it's a distraction *from the trouble* — from the ongoing, never-resolving, tentacular work of living and dying well with the machines we've made ourselves with, on a planet we've damaged, among the labor and the species and the communities we're entangled with. [Staying with the trouble](https://www.youonai.ai/fieldguide/med/staying_with_the_trouble) is the refusal of every ending, the sunrise included. The sunrise is just the apocalypse that the comfortable prefer.

**WHITTAKER:** "The sunrise is the apocalypse the comfortable prefer" — I'm going to be using that, with attribution. And I think our two refusals actually reinforce rather than compete, which doesn't happen often between a philosopher and an engineer. You refuse The Event because it relieves people of the present's work. I refuse it because it relieves people of the present's *power analysis*. Same anesthetic, two organs it numbs. But let me press you on one thing, Donna, because "staying with the trouble" has a failure mode that's the mirror of my failure mode. My structural analysis can demobilize by making capital all-powerful — you got me on that in the first hour. *Your* trouble-staying can demobilize by making everything so entangled, so tentacular, so irreducibly complex that there's no lever anyone can actually pull. "It's all connected, stay with the mess" can become an excuse for never naming the company, never backing the specific bill, never picking the fight you could win. So here's my question across the table, before Edo even hands us the crossing: doesn't "stay with the trouble" risk becoming its own anesthetic — a sophisticated way of admiring the complexity of a problem instead of attacking the part of it you could actually move?

**HARAWAY:** Yes. It absolutely risks that, and I have watched it become that in the hands of people who use my work as a permission slip for paralysis, and it is one of the genuine griefs of my life. So let me answer you with the discipline and not the cliché. Staying with the trouble is not admiring the complexity. It is the opposite — it is *acting* inside the complexity without the comfort of believing your action resolves it. The beaver builds the dam knowing the dam does not stop the river; she builds it because the dam makes a pool where particular life can take hold, and then she maintains it tomorrow and the day after, forever, against the current. That's staying with the trouble: situated, partial, *effective* action that refuses the fantasy of the final solution. So when you say "name the company, back the bill, pick the winnable fight" — yes, do all of it, and do it *as* trouble-staying, which means without the lie that winning the fight ends the trouble, because the lie that this one fight ends it is exactly what burns activists out when the trouble, predictably, continues. Your organizing and my trouble-staying are not in tension. Your organizing is *how* I stay with the trouble. My trouble-staying is what keeps your organizers from quitting when the win turns out not to be the ending. We need each other more than the staging admits, Meredith, and I think we've both known that for an hour.

**EDO SEGAL:** That may be the truest thing said tonight, and it's the perfect note to step back on, because you've just told me you need each other — which means it's time to find out what you do when I'm not refereeing. The rules of the crossing are short. Each of you questions the other, directly. I will not rescue anyone. Donna, you won the toss backstage. Begin.