■ MODERATE RISK ■ Creative Arts
Partly, and the simulation half is already doing the heavy lifting. Computational fluid dynamics has genuinely displaced iterative field testing for performance kites, but somebody still has to decide what the thing should be and prove it survives real gusts.
“CFD simulation optimizes aerodynamics. Your wind testing was just a fun afternoon.”
Our AI replacement risk score — how we score jobs
The trade spans two very different worlds. Performance kite design — kitesurfing, snowkiting, traction, and increasingly airborne wind energy — is an engineering job: canopy profile, bridle geometry, leading-edge inflation pressure, depower range, ripstop panel layout, and seam placement, all tuned so a rider can relaunch from water and not get lofted. Artistic and festival kite making is a different craft entirely: spar tapering, sail tension, appliqué and dye work on nylon, and the finicky trimming that gets a sculptural kite to actually fly rather than merely look good hanging in a hall.
Simulation has taken the iteration loop. CFD and fluid-structure interaction models predict lift, drag, and canopy deformation before a prototype exists, generative and topology optimization propose bridle and frame configurations, and parametric CAD plus automated panel nesting turns a shape into cutting files instantly. Automated fabric cutting and offshore sewing handle production. Generative imagery handles graphics and colorways that a designer once sketched. For a company iterating a kitesurfing line, the number of physical prototypes per season has dropped substantially, and that used to be the designer's whole calendar.
What resists is validation and taste. Soft, highly deformable membrane structures in turbulent, gusty air are one of the genuinely hard simulation problems — models drift from reality precisely at the edge conditions that matter, so test pilots on the water remain the arbiter of whether a kite feels stable or scary. Feel is not a computable quantity: riders buy handling character, bar pressure, and predictability. Safety failure modes and manufacturability also demand hands-on judgement. And artistic kite making is craft with cultural roots, priced accordingly. Our 29 reflects a real transfer of engineering hours to software inside a role that still needs a designer with an opinion.
Automatability: our editorial assessment of current and near-term AI capability
The shift is already well underway — simulation-first development is standard at major kite brands and the prototype count per model has been falling for years. Through the 2030s expect further consolidation, with small design shops squeezed by brands that can afford compute and offshore production. Airborne wind energy pulls some designers toward serious aerospace engineering. Field validation and rider feel keep humans in the loop, so the role is transformed rather than eliminated by 2040.
It has replaced most of it, not all. CFD and fluid-structure models now filter concepts before anything is sewn, which has cut prototype counts dramatically at established brands. But soft membranes in gusty air remain a hard modelling problem, and models diverge from reality exactly at the extreme conditions that determine whether a kite is safe. Final validation still happens on the water.
It is viable but narrow and consolidating. The performance market is dominated by a handful of brands, so most roles sit inside them and demand real engineering credentials plus riding ability. The growth area is airborne wind energy, which needs people who understand tethered soft-wing aerodynamics. Independent artistic kite making is a craft practice supported by workshops, commissions, and festivals.
Decide what a kite should feel like. Bar pressure, turning character, how predictably it depowers when a gust hits — these are the qualities riders actually buy, and they are evaluated by experienced test pilots rather than computed. AI also cannot take responsibility for safety behavior at the edge of the envelope, where simulation confidence is lowest and the consequences are highest.
Become the person who runs the tools rather than the person the tools replace: learn CFD, parametric CAD, and structural analysis properly. Keep riding and keep a tight loop with test pilots, because interpreting vague rider feedback into a bridle change is the durable skill. Then consider adjacent markets — airborne wind, traction, paragliding — where the same expertise pays better.