A developer working on AI economics said that to me. That’s exactly the problem. Right now, AI is priced on tokens — but tokens don’t represent the work being done. Same token count → radically different compute → radically different cost So everything downstream breaks: * pricing * margin control * forecasting. We’re
Sen. Mark Warner (D-VA) just emailed constituents acknowledging growing fear that AI will eliminate jobs. He’s not wrong about the fear. But his note offered no mechanism, no model, no way to measure or control what’s actually happening. That’s the tell. Washington is reacting to visible effects—
Oracle didn’t just cut jobs. It moved the center of gravity of its entire business. Tens of thousands of roles gone. Not quietly, not incrementally — decisively. And not because one thing broke, but because multiple pressures converged at once. This is what a cost structure inversion looks like in
A company deploys an AI agent for customer support. Based on token pricing: $0.12 per interaction Finance signs off. It looks predictable. Then it runs. Each interaction triggers… * 3 retries * 2 tool calls * 1 retrieval pass * a branching decision path The system doesn’t fail. It works exactly as
Oracle has introduced “AI Units” to simplify pricing. But “AI Units” don’t stabilize cost. Because the problem isn’t the price of a unit. It’s how many units a real workflow actually consumes. In SaaS, you can model this: users → seats → predictable usage → stable margins In AI, you
If it looks like a duck, acts like a duck, and quacks like a duck… it’s probably a duck. AI looks like SaaS. Is sold like SaaS. But it doesn’t behave like SaaS at all. In SaaS: * cost is largely fixed * margins are modeled upfront * pricing holds as
Most conversations about AI governance start in the same place: “What happens if something goes wrong?” And that’s a valid concern. A single failure — a hallucination, a bad decision, a compliance breach — can be expensive. Sometimes catastrophically so. Like hitting an iceberg. That’s the world governance is built
By Alan Jacobson, Systems Architect & Analyst Can LLMs “scale” their way to profitability by charging a flat rate of $20/month for unlimited compute? * If every man, woman and child in America pays $20/month = $81.6B/year * Capex depreciation: $50B/year * Staff: $11.25B/year * Cost of electricity
By Alan Jacobson, Systems Architect & Analyst The first wave of AI adoption is over. Not because the technology failed, but because users hit its limits faster than the industry expected. Models forgot. Systems hallucinated. Outputs couldn’t be trusted. And when reliability drops, adoption at scale is imp. That
AI needs to make sense. But it also needs to make dollars. Dollars+sense. That’s how most AI governance frameworks are designed. * Policies that make sense. * Controls that make sense. * Audit trails that make sense. But AI systems don’t run on sense. They run on dollars and cents.
AI didn’t eliminate cost. It changed how cost behaves. What Finance expected The pitch was simple: * reduce headcount * automate work * lower fixed cost AI was framed like software: buy seats, deploy tools, realize efficiency. The assumption: cost goes down, margin goes up. What Finance is seeing The first few
AI does not behave like SaaS. It behaves like a utility — cost is incurred with every use. “AI economics done right” means aligning measurement, control, and pricing around the actual work being performed. * FLOP-Based Metering (FBM) measures the work. * Normalized Compute Units (NCU) standardize the unit. * G-PEP enforces control before
Elementary school teachers use a system called Lexile to determine the reading level of texts. It uses word length and sentence length as a proxy for difficulty. AI does the same thing. It uses token count as a proxy for compute. Different domain. Same mistake. Both systems rely on proxies
For the past two years, AI has felt like magic. * Costs looked manageable. * Capabilities kept improving. * And the narrative was simple: Build with AI, and the economics will figure themselves out. That narrative is starting to break. Not because the models stopped improving. But because the economics underneath them never
The AI industry keeps trying to solve a measurement problem by changing the model. That’s the wrong layer. The problem is not the LLM. And there is no need to change the model to solve the problem. The problem is the unit used to measure cost. Today, AI is
What is “FBM?” FLOP-Based Metering – which does not require changing the LLM. It is a layer between the user and the model. Jeff Bezos didn’t change TCP/IP to build ecommerce. Amazon was built with SSL as a commerce layer on top of the internet. That’s the analogy
By Alan Jacobson, AI Economics Strategist Digital pricing has always followed a simple pattern: move closer to the result, and pricing gets smarter. * CPM (cost per thousand impressions): pay for exposure * CPC (cost per click): pay for engagement * CPA (cost per action): pay for a result Each step reduced waste
Tokens don’t measure AI cost. They measure words. But AI doesn’t charge for words. It charges for compute. And in agentic systems, a growing share of that compute: * never becomes tokens * happens off-screen * and can’t be measured by word count So if you’re using tokens to
By Alan Jacobson, AI Economics Strategist Everyone is building dashboards. Beautiful ones. Clean charts. Real-time graphs. Observability everywhere. But what’s the point if what you are seeing is missing the point? The entire system is built on a unit that cannot see what matters. If you are counting tokens,
By Alan Jacobson, AI Economics Strategist Another week, another “step change in capability,” according to Fortune’s Jeremy Kahn. This time, it’s Anthropic’s reportedly more powerful model, Mythos—a system described as a meaningful leap beyond its current generation. The details surfaced in an unsecured draft blog post,
By Alan Jacobson, AI Economics Strategist This week, stories in The New York Times and The Wall Street Journal highlighted something that’s been quietly building inside companies: employees are deploying AI agents that generate massive volumes of tokens—and massive, unexpected costs. The assumption behind most AI systems is