The Scar Tissue: A Surgeon Identifies AI by Its Perfection
November 2034
Dr. Leila Sharifi was a plastic and reconstructive surgeon at Hôpital Saint-Louis in Paris. She was fifty-three. She had spent twenty-eight years looking at scars — not to remove them, though that was often the goal, but to read them. Scars, to Leila, were documents. They recorded the history of a wound and the history of its repair. They recorded the biology of the patient — how their tissue healed, how their immune system responded, how their collagen organized. And they recorded the surgeon.
Every surgeon left a signature in their scars. Leila had known this since residency — had learned to identify the attending surgeon on a revision case not from the chart but from the tissue. Dr. Moreau closed with a characteristic overtightness at the wound margins — his scars were slightly raised, a millimeter of excess tension visible as a faint ridge. Dr. Patel's suture placement had a rhythmic irregularity — every fourth stitch was fractionally deeper than the preceding three, a pattern so consistent it was almost certainly unconscious, a motor habit embedded in her hands the way a typist's rhythm is embedded in their fingers. Dr. Tanaka's incisions curved slightly at the terminus — a half-degree deviation that his conscious mind corrected for but that persisted in the tissue as a whisper, a directional signature, a ghost of the hand's natural arc.
These signatures were not flaws. They were not evidence of poor technique. They were the irreducible fingerprint of a human body performing a physical task — the same phenomenon that Priya Chakrabarti had documented in diagnostic errors, that Ines Achterberg had documented in orphaned gestures, that Constance Okafor had documented in her gesture archive. The body leaves a mark. The mark is individual. The individuality cannot be trained out because it is not a deficit. It is the body's way of being itself.
In 2032, robotic surgery systems with AI-directed suturing became standard in French hospitals. By 2034, approximately sixty percent of surgical closures in Leila's department were AI-assisted. The systems were excellent — the suturing was precise, the tension was uniform, the placement was consistent. The outcomes were measurably better: lower infection rates, less hypertrophic scarring, faster healing.
And the scars were anonymous.
The observation
Leila noticed it during a revision consultation in November 2034. A patient presented with a scar from an abdominal procedure performed eighteen months earlier. Leila examined the scar with her standard protocol — visual assessment, palpation, scar scale scoring. The scar was excellent. Well-healed, flat, minimal discoloration. A textbook result.
But it was empty. The scar had no personality. No signature. No evidence of the specific, irreducible individuality of a human hand. The suture placement was perfectly regular — not rhythmically irregular like Dr. Patel's, not characteristically tight like Dr. Moreau's, not subtly curved like Dr. Tanaka's. It was geometrically precise in a way that no human hand could achieve and no human hand had achieved.
Leila checked the chart. The procedure had been performed with AI-assisted robotic closure. The surgeon of record had supervised. The hands that closed the wound were not human.
She began looking. Over the following months, she examined scars from 200 surgical patients — 120 closed by AI-assisted systems, 80 closed by human surgeons. She developed a protocol: examine the scar blind, without access to the operative record, and determine whether the closure was human or AI-assisted.
Her accuracy was 94 percent.
The method was simple: she looked for the signature. Human scars had one. AI scars did not. The presence or absence of the surgeon's individual mark was, for a trained eye, as obvious as the presence or absence of a handwriting style. A letter typed by a machine and a letter written by hand convey the same words. The difference is visible at a glance.
The paper
"The Anonymous Scar: Identifying AI-Assisted Surgical Closure by the Absence of Operator Signature." Leila published in The Lancet in 2035. The paper was twelve pages and included photographs — close-up images of scar tissue, human-closed and AI-closed, presented side by side.
The human scars were not worse. Some were better than the AI scars by clinical metrics. But they were all identifiable — each carrying the specific marks of the surgeon's hands, the way a painting carries the marks of the painter's brushstrokes. The slight variations in stitch spacing. The characteristic tension profile. The directional bias of the incision. These marks were measurable but had previously been categorized as noise — the same classification Priya's diagnostic errors had received before she reframed them as signal.
Leila reframed them. The surgeon's signature was not noise. It was identity. It was the evidence that a specific human being, with a specific body and a specific training and a specific history of ten thousand closures, had touched this patient and left a mark.
"The AI-assisted scars are clinically superior," Leila wrote. "They are also forensically anonymous. No individual performed the closure. No individual left a mark. The scar was produced by a system. The system has no signature because the system has no body."
"This matters. Not for clinical outcomes — the outcomes are better without the signature. It matters for what the scar is. A scar is not merely a wound that has healed. A scar is a record of a meeting — the meeting between a wound and the person who repaired it. The human scar says: someone was here. Someone's hands touched this body. Someone made decisions — where to place the suture, how tight to draw it, when to tie — and those decisions were inflected by the specific way that person's hands moved through the world."
"The AI scar says nothing about anyone. It is a perfect closure that was performed by no one. It is a record without an author. A letter with no handwriting."
This is a supplemental entry in The Interregnum. For the diagnostic errors that first revealed human cognitive signatures, see The Archive of Errors. For the hospital that briefly ran on human signatures alone, see The Night the Power Went Out.