Search papers, labs, and topics across Lattice.
This paper proposes framing agentic AI systems as marginal token allocation economies across routing, agentic planning, serving, and training layers, rather than as isolated text generators. It argues that each layer implicitly solves a marginal benefit equals marginal cost equation, but with different parameters, leading to global misallocations when layers optimize locally. By adopting marginal token allocation as a shared accounting object, the authors identify common failure modes and suggest a research agenda focused on token-aware evaluation and pricing.
Treating agentic AI systems as token economies reveals that current designs, which optimize token usage locally, lead to predictable global misallocations and inefficiencies.
This position paper argues that agentic AI systems should be designed and evaluated as \emph{marginal token allocation economies} rather than as text generators priced by the unit. We follow a single request -- a developer asking a coding agent to fix a failing test -- through four economic layers that today are designed in isolation: a router that decides which model answers, an agent that decides whether to plan, act, verify, or defer, a serving stack that decides how to produce each token, and a training pipeline that decides whether the trace is worth learning from. We show that all four layers are solving the \emph{same} first-order condition -- marginal benefit equals marginal cost plus latency cost plus risk cost -- with different index sets and different prices. The framing is deliberately minimal: we do not propose a complete theory of AI economics. But adopting marginal token allocation as the shared accounting object explains why systems that locally minimize tokens globally misallocate them, predicts a small set of recurring failure modes (over-routing, over-delegation, under-verification, serving congestion, stale rollouts, cache misuse), and points to a concrete research agenda in token-aware evaluation, autonomy pricing, congestion-priced serving, and risk-adjusted RL budgeting.
