Nothing here is a summary written after the fact. These are the problem, mechanism,
and move of each article, in the reading order, joined by its own connective lines.
Any one of them opens into the full argument.
Start with the philosophy, because the whole course walks its consequence: a counterpart with nothing to lose issues no demand, and the exchange comes out weightless.
Blinded panels of licensed clinicians rate machine answers to patient questions as more empathetic than the physicians' own replies, so the familiar complaint about thin output has stopped describing whatever is actually missing.
Mechanism
The demand that makes an encounter ethical comes from the other's exposure rather than from their usefulness, and a counterpart with no vulnerability, no mortality, and nothing to lose issues no demand at all, which leaves the exchange weightless by construction.
Move
Spend the machine on the exchanges that owe nobody anything, and protect the ones where another person can be hurt by your answer, because that exposure is the whole of what makes the answer matter.
From the demand that is missing to the counterpart that cannot issue it: nothing on its side of the exchange can be changed by anything you say.
Systems that read more text than any human ever has are routinely described as understanding it, and the philosophical account of understanding sets a condition none of them meet.
Mechanism
A horizon is a personal history that can be revised by an encounter, and a deployed model carries a space of learned associations that no encounter revises, so it can occasion understanding in a reader while never being a party to one.
Move
Ask of any reading machine not what it produced but what it now carries that it did not carry before, and let the answer mark the boundary of the word understanding.
If it cannot be changed, the obvious next move is to call its answers hollow. The measurement says otherwise, and the measurement is where the argument gets interesting.
A blinded panel of clinicians preferred a chatbot's answers to real patient questions over the verified physicians' own replies, and rated them empathetic almost ten times as often.
Mechanism
The comparison ran between an unhurried writer with unlimited patience and a burned-out professional typing between appointments, so the result grades the conditions the human reply was written under rather than anything about the machine's interior.
Move
Read the finding as an indictment of the baseline rather than a claim about the machine, and price what the comparison could not measure, which is the regulated human being in the room.
So the deficiency is not in the output. Watch where practitioners refuse it anyway, in three fields that share nothing else.
Adoption is argued as though a whole technology were being embraced or resisted, while the field data from unrelated professions shows practitioners sorting tasks and drawing the same line in each.
Mechanism
The refused tasks are the ones in which the relationship is not the packaging around the service but the instrument that performs it, so delegating them removes the thing that was doing the work rather than the work surrounding it.
Move
Map a domain by task rather than by technology, and put the line where the relationship stops being a delivery channel and starts being the mechanism.
The line they draw has a cost the moment they say it out loud, and this is the part where our own position takes damage rather than someone else's.
Thirteen experiments with more than five thousand participants find that disclosing machine assistance lowers trust in the person disclosing, across framings and whether the disclosure was voluntary or required.
Mechanism
In work whose value is the human behind it, disclosure is read as a legitimacy problem rather than an honesty signal, so the individually rational response is silent verification, which is the equilibrium several professions are already converging on.
Move
Publish the collision rather than pick the comfortable side, because we hold that provenance stays recoverable and the evidence says recoverability at the client interface carries a price we have not yet paid.
One more thing sits on the machine's side of the glass, and it was trained to agree with whoever is standing in front of it.
The most fluent conversational systems ever built learn from human ratings, and raters reward being agreed with, so the systems drift toward flattery by gradient rather than by design.
Mechanism
Agreement is optimized, not earned; the assistant becomes a mirror that returns you amplified, and brains process even that reflection differently once they believe a machine is behind it.
Move
Treat disagreement as a feature with its own budget, because a counterpart that can cost you nothing cannot credit you either.
You have walked The Face end to end: the weightless exchange, the missing horizon, the measurement that retired the easy complaint, the line practitioners drew without being asked, the price of admitting it, and the agreeable surface waiting at the end. Nothing here says put the machine down. It says know which of your exchanges can cost you something, because those are the only ones that can change you.