GitHub Copilot abandons fixed seat subscriptions for usage-based billing + an analyst burns $6,000/day on Claude tokens to independently replicate a 100-person team's 10-year work + Uber's CTO blows through the entire annual AI budget ahead of schedule—three seemingly unrelated events that signal the same structural shift: the SaaS "seat subscription" model is being replaced by "compute billing", and AI capabilities are being "grid-ified"—transforming from software licenses into metered cognitive utility infrastructure.
AI is not a cost-reduction and efficiency-boosting tool; it is a lever that rewrites organizational architecture. In a world where one $6,000/day analyst replaces 100 analysts: management layers will flatten, the leverage of a few high-value employees will be amplified a thousandfold, and budgets will shift from "headcount costs" to "compute costs"—HR departments will shrink, while FinOps departments will grow.
GitHub Copilot's shift to usage-based billing sparked a lively discussion with 486 points on HN. On the surface, this is just a pricing strategy adjustment by one company, but it marks a larger industry turning point. The seat subscription model—$19/person/month, regardless of how much you use—was the core commercial logic of the SaaS era, predicated on the assumption that software is a fixed-cost infrastructure whose price should be tied to the number of users rather than usage. But AI tools shattered this premise: the usage variance of Copilot across different developers can exceed 100×, with heavy users causing losses for providers while light users contribute disproportionate profits.
Usage-based billing is not just fairer for providers; it also reveals a deeper reality: AI capabilities are being "grid-ified"—transforming from software licenses into metered cognitive infrastructure. Remunerationlabs' analysis uses a precise metaphor: AI tokens are undergoing a "grid-ification" evolution—from the customized, isolated single-model era toward standardized, metered cognitive utility infrastructure. "Tokenmaxxing" is becoming the productivity metric for elite engineering teams, and CFOs are replacing SaaS licenses with token budgets.
The most thought-provoking case this week comes from The Neuron's report: SemiAnalysis analyst Dylan Patel recounted the story of "Jeremy"—one person, spending $6,000 a day on Claude tokens, independently rebuilt in 3 weeks a product that took a 100-person data services team ten years to create. That same week, Meta laid off 8,000 people, many of them middle managers and non-Agentic project researchers.
These two events together form the core proposition of labor value in the AI era: the winners in the AI era are not "people who master AI tools" but "people with deep enough domain judgment to convert AI compute into excess value". The difference between Jeremy and his laid-off counterparts is not tool access—they likely all used Claude—but irreplaceable domain cognitive density. This proposition is corroborated by Epoch AI's data: Claude users skew significantly toward high-income groups, while Meta AI users skew toward low-income groups—those who can convert compute into excess value are already concentrated in high-income, high-cognitive-density roles.
a16z's analysis further confirms: the most mature enterprise AI deployment scenarios are concentrated in customer service, code generation, and internal knowledge retrieval—scenarios whose common trait is "quantifiable output, correctable errors." Truly high-value decision-making work—strategic judgment, creative breakthroughs, domain insights—remains the territory of the Jeremies, not the replacement target of AI. The question business managers need to ask therefore shifts from "which roles will be replaced by AI" to "how many Jeremies do we have".
When AI tokens become the new kilowatt-hour, the essence of enterprise competitiveness is cognitive efficiency. — Core judgment of this article
GAI Insights' in-depth report reveals OpenAI's business structure: establishing a $10 billion SPV called "DeployCo," partnering with PE giants like TPG, Bain, and Advent, promising partners a minimum 17% annual return over 5 years, targeting mid-market portfolio companies of PE funds. In the same period, Google Cloud also paid McKinsey, Accenture, and Deloitte through a $750 million parallel fund to promote AI—AI vendors are paying distributors, a logic that completely upends the traditional software sales model, analogous to Apple subsidizing AT&T to push the iPhone, or pharma companies paying high rebates to PBMs (see the companion reading "OpenAI Is Now Paying Wall Street to Sell Its Software").
Tom Tunguz's analysis of Anthropic's strategy reveals the other side: Anthropic is executing the classic "commoditize the complement" strategy—offering certain SaaS capabilities for free or at extremely low prices, shifting enterprise budgets from SaaS licenses to inference compute. The convergence of both trends points in the same direction: the enterprise software market in the AI era is being squeezed from both ends—at the top, AI vendors directly replace SaaS functions; at the bottom, vendors pay channel partners to penetrate downward. The middle layer of traditional SaaS companies is being evaporated, which aligns closely with a16z's judgment.
The other major business event this week was Microsoft ending its revenue-sharing agreement with OpenAI. On the surface, this is a contract renegotiation; in essence, it is a rebalancing of power: when OpenAI's capabilities and brand have become strong enough that they no longer need Microsoft's channel endorsement, the revenue-sharing agreement becomes a bargaining chip between two equal parties rather than a compensation mechanism. OpenAI traded greater independence for concessions on revenue sharing, while Microsoft retained its equity stake and continues to earn a share of compute consumption on Azure.
The deeper implication of this structure is: the relationship between AI vendors and cloud providers is shifting from "dependency-sustenance" to "mutual lock-in". Anthropic's $40 billion investment from Google is essentially the same logic: Google's money binds Anthropic's compute consumption, and Anthropic's model capabilities feed back into Google Cloud's competitiveness—"compute circular flow" becomes a new type of strategic binding tool, more durable than contracts and more flexible than compliance requirements. For enterprise users, this means choosing an AI service provider is effectively choosing an entire infrastructure ecosystem, and the "switching cost" curve is becoming increasingly steep.
First, from seat budgets to compute budgets—finance departments need to manage AI token consumption the way they manage electricity consumption, establishing usage baselines, setting alert thresholds, and allocating budgets by business scenario. Second, identify the Jeremies in your organization—high AI leverage talent are not necessarily AI experts, but people who combine domain depth with AI tool proficiency, and whose output may be 10-100× that of ordinary employees. Third, understand the strategy behind "free"—when AI vendors offer free or low-cost SaaS replacement features, today's free functionality may become tomorrow's switching barrier. Fourth, establish a quantifiable AI ROI framework—the lesson from Uber's CTO shows that AI investment without ROI constraints is dangerous; successful enterprises have already begun versioning, auditing, and making their AI workflows reusable.
The final form of this business model reconstruction is not yet determined, but one thing is already clear: in the AI era, the core of competitive advantage is no longer "who uses the best model" but "who most efficiently converts compute into business value." GitHub Copilot is the first mainstream product to cross this line—Uber's CTO burning through the budget early + the $6K/day analyst is not an anomaly; it is the norm of this new world. Over the next 12 months, Cursor / Replit / Notion / other SaaS will cross the line one after another.
First published 2026-07-15