In AI the threat is layer migration, not a competitor — work relocates across layers when any variable moves

Gupta names the new failure mode: “In SaaS, you were usually killed by another company… There was one exception: a company built on top of someone else’s platform could be wiped out by a pricing or terms change, no competitor involved. AI makes that exception the default.” The threat is not a faster team — it is layer migration: “Work moves when any variable changes enough: capability, cost, latency, deployment, regulation, trust. It can move into the model, an open-weight alternative, the customer’s data platform, an agent runtime, or the device itself.” Demand compresses the timeline too — “buyers… know what lock-in costs now, so the arbitrage in that lag is closing.”

This is the threat-model consequence of AI strategy is a self-rewriting equation — solving one constraint changes which constraint matters next: when variables are coupled and moving, value doesn’t get out-competed, it gets relocated. It sharpens Self-disruption follows the value chain downward — software companies must eat their own agent layer before someone else does — the value chain doesn’t only flow toward the model, it can migrate to the data platform or the device — and it is why Scaffolding is tech debt against the next model — the bitter lesson applied to product building and Model compensations become liabilities as capabilities advance — yesterday's fixes hobble today's agent: a capability gain migrates your function into the model. The defense is to own the layer that migration flows toward, which is the logic of LLMs selectively destroy vertical software moats — 5 fall, 5 hold.