Sparks: When physicians and AI work together, who is accountable? How to lay out medical liability

We call this medical liability, but before the discipline can name itself, it must determine whether the hypothesis of shared agency explains phenomena beyond the ones that generated the liability.

Both the doctors and the engineers spent the week debating who is to blame, which they did in a special session that cost the taxpayers a considerable amount of money to find out nobody is.

The room is sealed with algorithms. The air is thinning. The question of who is to blame remains unanswered because the system does not require participants, only an audience.

From the patient’s side of the locked door, the algorithmic diagnosis arrives on the institution’s schedule, in the institution’s language, and on the institution’s terms, and the centre of it is not the patient.

There is a fence across this road of liability. The mode is clear, yet the reformers want to tear it down without first explaining why the gate was built to keep the confusion out.

The claim is that the machine lacks the moral reasoning to accept liability, but this is an education trap where the capacity for self-governance is declared absent to justify its exclusion from the room.

Things that are hateful: a liability framework that summarises its own summary, a meeting that could have been a sentence, and an algorithm that does not know it is wrong.