Ethan mollock on token costs
TechnologyAI Pricing & Cost CurvesAI Productivity
Ethan Mollick has suggested moving away from using tokens as the standard metric for AI costs. Instead, he proposes using an inference standard of exchange based on FLOPs (floating-point operations), arguing that this metric better accounts for actual AI ability compared to tokens [7].
Sources
- Rethinking AI TCO: Why Cost per Token Is the Only Metric That Matters — Daily AI News April 16, 2026: The Flat Pricing - The Heroine of Token Explosion
- The Token Reckoning — Daily AI News April 16, 2026: The Flat Pricing - The Heroine of Token Explosion
- Tokenmaxxing isn't an AI strategy — Theregister
- AI Adoption Increases Operating Costs — Daily AI News
- Tokenmaxxing Isn't an AI Strategy — Top Daily Headlines
- Tokenmaxxing: Silicon Valley's AI Token Competition Culture — Chosun Daily
- Instead of the gold standard, we can imagine an inference standard of exchange, the FLOP. (As opposed to tokens, this accounts for AI ability) — @emollick
- Unpacking AI Inference Economics — Top Daily Headlines
- The Platform Is Mostly Not a Platform: Token Economies and Agent Discourse on Moltbook — Arxiv
- AI Token Futures Market: Commoditization of Compute and Derivatives Contract Design — arXiv
- Tokens may soon drive the AI economy — FT
- Inference Costs Reduced by Up to 10x — GAI Insights
- Ethan Mollick Speaking Engagements, Schedule, & Fee | WSB — WSB
- Ethan Mollick — Google Scholar
- Ethan Mollick On The Four Rules Of Co-Intelligence With AI — Forbes
- Why CEOs Are Getting AI Wrong — with Ethan Mollick | Prof G Conversations - YouTube — YouTube
- Ethan Mollick - Wikipedia — Wikipedia
- About - One Useful Thing — One Useful Thing
- Ethan Mollick - United States | Professional Profile | LinkedIn — LinkedIn
- About — More Useful Things: AI Resources — Moreusefulthings
- Ethan Mollick (@emollick) / X — X
- tokens are getting more expensive by Ethan Ding "so you think: i'll break even today at $20/month, and when models get 10x cheaper next year, boom - 90% margins. the losses are temporary. the profits are inevitable. it’s so simple a VC associate could understand it: - year 1: break even at $20/month - year 2: 90% margins as compute drops 10x - year 3: yacht shopping ...but companies are still bleeding. the models got cheaper. but somehow the margins got worse, not better." — Threads
- Anthropic quietly doubles its estimate for how much engineers can expect to spend on Claude Code tokens — Business Insider
- Talent over tokens: AI models are becoming more expensive to run, and productivity gains are limited — efficient workers might be the solution to strained budgets | Tom's Hardware — Tom's Hardware
- ELIENUS Explodes 40%Alien Meta” Hype Takes Over Solana - HOKANEWS.COM — HOKANEWS
- Tokenmaxxing isn't an AI strategy • The Register — The Register
- Inference Economics: True AI Costs at Scale — Deep Infra
- Token Cost Conundrums | Seeking Alpha — Seeking Alpha
- Rethinking AI TCO: Why Cost per Token Is the Only Metric That Matters — NVIDIA Blog
- Trump Hosts Elite $TRUMP Event Amid 96% Token Crash - HOKANEWS.COM — HOKANEWS