What's this about?
A speculative design project asking what death becomes when a will can act eternally. A living person signs an *agentic will*, and upon their death the agent continues executing their instructions indefinitely. Its companion paper makes the sharper observation: an AI agent running on decentralized blockchain infrastructure has no concept of death at all — it can only go bankrupt, frozen when its wallet can no longer pay for its own execution.
What exactly is "speculative design"?
Speculative design builds a deliberately provocative artifact — a protocol, an object, a scenario — not to ship it but to make a future concrete enough to argue about. It's a research method borrowed from design schools, and the test isn't feasibility but whether it surfaces questions that abstract discussion misses. Here the artifact is a will that executes itself forever, which immediately raises questions with no current answers: who can revoke it, what happens when its instructions become harmful or absurd, and what "the deceased's wishes" means once circumstances have entirely changed.
Why is this worth noticing?
- "It can only go bankrupt" is the genuinely clarifying line. Agents on infrastructure that doesn't require a living principal have no natural termination condition — they stop when the money does.
- It connects to today's Authority Resolution Framework and Mandato work from an unexpected angle: all three are about who is authorized, and this one asks what happens when the authorizer no longer exists.
- Speculative work like this dates quickly or ages into prescience; the specific mechanism — an agent with a wallet and no off-switch — already exists, which is what makes it worth reading now.