■ SAFE RISK ■ Sports & Entertainment
No. A machine can score a touch faster than any judge, but nobody earns a sixth dan from a scoreboard, and nobody learns to hold a blade correctly from a video that can't feel their grip collapse.
“Teaching kendo or fencing is about discipline, not data. Your bokken doesn't have Bluetooth.”
Our AI replacement risk score — how we score jobs
Strip away the romance and the daily work of a master swordsman is teaching: running keiko or piste sessions, drilling footwork until it stops being conscious, calling out the half-beat where a student's shoulder tenses before a cut, and grading kata against a standard that lives in a lineage rather than a rulebook. Between classes there is equipment work — replacing bamboo slats in a shinai before one splinters into someone's throat protector, checking that a foil's blade hasn't developed the hairline set that precedes a snap — plus club administration, grading panels, and the slow politics of any dojo or fencing federation.
Automation has genuinely landed in scoring and analysis. Electric scoring apparatus decided fencing touches decades before anyone said 'AI'; now video review, pose-estimation apps and instrumented targets can tell a student their elbow drifted or their tempo slipped, and wearable sensors give reasonably honest feedback on cut speed and distance. Competition officiating is drifting toward machine assistance for the same reason line-calling did in tennis: humans blink. A phone app is already a better metronome and a better mirror than most beginners have ever had.
What resists is everything the sword is a pretext for. Correction in a martial art is physical and consensual — you take a student's wrist, you feed them a strike at the exact speed they can barely handle, you decide in a half-second whether this person needs pushing or catching. That calibration of controlled danger is a safety judgment no vision model owns, and the liability of getting it wrong lands on a named human. Add lineage: rank in kendo, kenjutsu or classical fencing is conferred by people, transmitted body-to-body, and worth precisely as much as the tradition behind the person granting it. An AI-certified swordsman is a certificate nobody in the room will honor. Our low risk score reflects a craft where the tool is optional and the transmission is the product.
Automatability: our editorial assessment of current and near-term AI capability
Expect the officiating and analytics layer to keep automating through the late 2020s — sensor scoring, automatic point calls, app-based drill feedback that most clubs will happily adopt. The teaching relationship stays intact well beyond 2040. The real threats to a sword school are demographic and economic, not computational: shrinking youth membership, rent, and the cost of insuring a room where people hit each other.
It can teach the shape of a movement and nag you into practicing it, which is not nothing. What it cannot do is apply resistance, absorb a badly aimed cut, or notice that a student is compensating for an old shoulder injury. Beginners who learn purely from apps arrive at a dojo with confident bad habits that take longer to unpick than starting fresh.
Largely, yes, and it started long ago — electric apparatus has scored fencing for generations. Expect sensor and video systems to absorb more of the call-making in kendo and HEMA too, because they are more consistent than tired referees. That shifts human officials toward rules interpretation, safety stoppages and disputes rather than eliminating them.
Viable, rarely lucrative, and threatened by rent rather than robots. The schools that survive diversify: kids' programs, seminars, film and stage combat, and online membership tiers that fund the physical space. Treat the dojo as a community business with a teaching product, not a hobby that occasionally takes money.
Coaching pedagogy, sports-injury awareness, and enough media literacy to produce decent video — since students will judge you online before they walk in. Learning to use pose analysis as a teaching aid rather than pretending it doesn't exist puts you ahead of instructors who dismiss it. The blade work you already have; the business and safety layer is where most schools fail.