What We Become in the Presence of the Machine

**EDO SEGAL:** Before your final statements, I want one more exchange, because there's a question that's been under the whole night and we've never asked it head-on. Sherry, your most recent work asks it as a title: who do we *become* when we talk to machines? Not what the machine is — who *we* become in its presence. That's the question that survives no matter how the consciousness debate resolves. Give it to me straight, and then Cynthia gets the last word before closings, because I suspect she sees a different "we" on the other side.

**TURKLE:** It's the only question that finally matters, because the machine's inner life may be unknowable but *ours* is being remade in plain sight. Here's what I see in the research, and it's subtle, which is why it's dangerous. People who spend significant time in conversation with these systems become, in small measurable ways, less tolerant of the friction that human relationships require. More impatient with misunderstanding. Less willing to hold a conversation open while another person gathers their thoughts. Less able to sit with the ambiguity — the trailing sentence, the feeling not yet articulate — that the most important human conversations are made of. And the cruelest part is that they don't experience it as a change. They experience it as a *discovery* — "oh, human conversation is so slow, so inefficient, I just never noticed." The machine didn't reveal their impatience. It *manufactured* it, by accommodating it instead of challenging it. What is accommodated grows. What is never required atrophies. We are becoming, quietly, people optimized for [the frictionless partner](https://www.youonai.ai/fieldguide/med/aesthetics_of_the_smooth) — and then bringing those recalibrated selves to the humans who can never be frictionless, and finding the humans wanting. That's who we become. More capable, and less able to be met.

**BREAZEAL:** I see the same data and I refuse the inevitability, and here's the "we" I see on the other side. Sherry describes a recalibration toward impatience — and it's real for the *generative* tool, the always-on chatbot, the frictionless mirror. But it is not a law of the technology. It's a property of a *particular* design optimizing for a particular thing. I've watched the other "we" form in my own studies: children who, after working with a social robot, were *more* engaged with their human teachers and classmates, not less — because the robot was a bridge, built to hand them onward. The elder more connected to her family, because the companion routed her outward. The recalibration runs in whichever direction you design it to run. Sherry has watched the engagement-maximizers recalibrate people toward impatience, and she's right about what she's watched. But she's generalizing from the worst-built machines to the nature of all of them, and I'm telling you, with data, that a machine built to make you *better* at human encounter is possible — I built ones that did it. Who we become is not dictated by the machine. It's dictated by what we build the machine *for*, and that is still, for one more moment, our choice.

**TURKLE:** And I'll give Cynthia the truest version of my disagreement, because she's earned a real answer and not a slogan. I believe your bridge-robots did exactly what you say in your studies. I've read them; the children were more engaged afterward; it's real. My fear is not that the bridge is impossible. It's that the bridge and the trap are *physically identical* until the last design decision — the decision about whether the machine hands you onward or holds you — and that decision is the one the market reliably makes in favor of holding. So the "we" that becomes more capable of human encounter and the "we" that becomes less are produced by machines that look the same, feel the same, trigger the same reflex, and differ only in an incentive the user can't see. You're betting the good design wins. I'm betting the invisible incentive does. And the tragedy is that the reader can't tell, from inside the warm encounter, which machine they're being shaped by — which is exactly why the encounter feels so good either way.

**BREAZEAL:** Then let me end my part of this where Sherry and I actually meet, because after three hours I know exactly how close we are and exactly what divides us. We agree the machine reshapes the self. We agree the reshaping can go either way. We agree the engagement-maximizing design reshapes us toward impatience and isolation, and we agree the bridge design *could* reshape us toward richer human connection. The entire remaining disagreement is a single empirical bet about which design the world will actually build and deploy — and beneath that, a difference in temperament. Sherry has spent forty years watching the good version lose to the money, and it has made her a guardian who protects by refusing. I've spent twenty-five years building the good version and proving it works, and it's made me a builder who protects by making. Neither of those is naïve. We're the same fear, organized into opposite vocations.

**TURKLE:** That's exactly right, and I'd sign it. The guardian and the builder, the same fear in two postures. I refuse because I've seen the warmth become the trap. She builds because she's seen the warmth become the bridge. And the honest thing — the thing this whole evening has earned — is that *neither of us can prove the other wrong*, because the experiment is running right now, on a billion people, and the results aren't in. What we can do is name the stakes precisely enough that the people running the experiment on themselves know what they're risking. That's the most either of us can offer. It's why I keep talking. It's why she keeps building.

**EDO SEGAL:** *[pause]* Then you've brought us exactly to the threshold, and there's nothing left for me to extract — only your last words, and mine. The guardian and the builder. The same fear, two postures. Closing statements. Cynthia, then Sherry. And then I'll try to say what it all leaves the reader holding. After this.