Research · Track 04
Tokenomics — Valuing AI Token Usage
The unit economics of inference, which almost nobody can currently model.
6–10 contributors
Open — seeking a lead
5–15 hrs/week contributing · 20+ hrs/week core
Why now
Context
As language models become infrastructure, the unit economics of inference remain poorly understood. Pricing, fair use and per-token value attribution are open problems with genuine policy weight — and they are a live blocker on enterprise adoption, which makes them squarely an SME problem.
Enterprise buyers keep asking a version of the same question — what did that cost me, and was it worth it — and keep getting answers built on vendor benchmarks. A member has already published on blockchain token economics; part of the work here is testing how much of that machinery transfers.
Sub-themes
Scope
- Token-level value attribution — which tokens matter, and by how much
- Inference cost models across architectures, hardware and routing
- Pricing mechanisms — per-token, per-task, outcome-based and hybrid
- Efficiency metrics such as tokens-per-decision and quality-per-rupee
- Economics of open-weight versus proprietary AI services
- Fair use, attribution and revenue share for training data
Year one deliverables
Output
- One to two frameworks or position papers
- An open token-value attribution toolkit
If the attribution toolkit cannot reproduce a finance team's own cost model on a real workload, the framework is wrong and the track closes rather than being defended.