■ SAFE RISK ■ Creative Arts
No. Togishi work is irreversible material removal on irreplaceable objects, judged by aesthetic conventions that took centuries to form. Nobody sane hands that to a machine.
“Polishing a katana takes 2 weeks, 12 stones, and a lifetime of training. AI can wait.”
Our AI replacement risk score — how we score jobs
A togishi's job is to reveal a blade rather than shine it. That means a graded sequence of natural and synthetic stones taking the surface from coarse shaping to finishing, then the finger-stone work that brings out the hamon, the hada grain and the activity within the steel, and finally burnishing and the nugui treatments that set the blade's final tone. Each blade dictates its own approach: the school, period and smith determine what the polish should emphasize, and the polisher must decide how much steel to remove — every polish shortens a sword's remaining life, and a mishandled one destroys value permanently. Around the bench there is client work with collectors, museums and shrines, appraisal collaboration, and long apprenticeship obligations.
Automation exists in the adjacent world and is irrelevant here. Industrial blade grinding, CNC honing and robotic polishing produce excellent kitchen knives and surgical instruments at scale. Machine vision can classify hamon patterns and assist provenance research, and 3D scanning documents geometry before and after work — genuinely useful for conservation records. Digital marketplaces have transformed how collectors find polishers. But the mass-market sword trade was mechanized long ago; it simply never overlapped with this occupation.
The resistance is threefold. First, the material relationship: the polisher reads the steel by feel and by watching how the surface responds, adjusting pressure and stone continuously as the blade's own structure emerges. Second, aesthetic judgment: what counts as a correct polish for a Kamakura-period blade is a connoisseurship question, not an optimization target, and different schools of polishing disagree. Third, stakes and provenance — many of these swords are designated cultural properties, and both the object and the certification regime require a named, licensed craftsman with a documented lineage. There are very few full-time togishi left in Japan, which is the actual crisis: succession, not substitution. Our low score is about robots, and says nothing reassuring about the trade's size.
Automatability: our editorial assessment of current and near-term AI capability
No automation timeline applies; the work is conservation of irreplaceable objects and will remain hand-done indefinitely. The pressing threat is generational — the number of qualified togishi is small and apprenticeships run for a decade or more, so the trade's survival past 2040 depends on succession and on collectors continuing to fund proper conservation.
A machine can grind and polish steel beautifully — that is settled industrial technology. What it cannot do is make the aesthetic and conservation judgments a togishi makes: how much steel to remove, what features of the hamon and grain to bring forward, and what the blade's school and period demand. On an irreplaceable antique, an irreversible automated pass is an unacceptable risk.
Endangered rather than dying, and for demographic reasons. Very few practitioners work full time in Japan, training takes about a decade under a master, and income is uncertain in the early years. Demand from collectors and institutions exists; the constraint is how many people can afford the apprenticeship. That is the real threat, and no technology addresses it.
Mainly in documentation and reach. High-resolution imaging and 3D scanning create conservation records showing exactly what a blade looked like before and after work, which matters for cultural property. Databases and image classification assist attribution research. Online platforms connect polishers with international collectors who would never have found them otherwise.
That the apprenticeship is long, poorly paid and non-negotiable, and that reputation is everything — one badly judged polish on a significant blade follows you permanently. Beyond stone technique, you need connoisseurship: the ability to read schools and periods, and to explain your choices to collectors and curators who are paying to preserve history, not to make something shiny.