Mark Cuban said something last week that I can’t stop thinking about.
“There are 33 million companies in this country. Aren’t going to have AI budgets. Aren’t going to have AI experts.”
He’s not talking about startups. He’s talking about the shoe store. The regional trucking outfit. The 12-person accounting firm. These businesses have no CTO, no ML team, no one reading AI Twitter. But they are the physical economy, and the wave we’ve been watching build is about to hit them directly.
The SaaS contract is broken
The SaaS model ran on a simple premise. Build one generic product. Force millions of businesses to bend their workflows around it. Charge rent forever. That worked because software was expensive to build and genuinely hard to customize.
Neither of those things is true anymore.
Cuban pointed to Satya Nadella saying software is dead because everything will be customized to unique utilization. That’s a Microsoft CEO describing the obsolescence of Microsoft’s own business model. When the people selling the product tell you the product is ending, pay attention.
The old relationship was: your business conforms to the software. The new one is: the intelligence conforms to your business. That flip is not incremental. It is a complete reversal of the power dynamic that built the $650 billion SaaS market.
Who actually wins this
Cuban’s real question, the one I think most people missed, was: “Who’s going to do it for them?”
Hundreds of billions are being burned right now at OpenAI, Anthropic, Google, and Meta to build the foundation models. The smartest engineers alive are in a capital war over who owns the base layer. Good. Let them. Because the wealth does not collect where the brain is built. It collects where the brain meets the business.
Think about the electricity analogy. The biggest winners of that era were not the engineers who built the generators at Westinghouse. They were the people who walked into dark factories and figured out where to plug in. The generator builders mattered. But the fortunes were made one factory floor at a time.
Thirty-three million companies are standing in the dark right now. Most of them will not figure this out on their own.
Where I think this gets mispriced
Every ambitious person in tech right now thinks survival means a seat at a frontier lab. The prestige is real. The compensation is real. But Cuban is pointing at the other 99 percent of the economy. The third-generation manufacturer. The county hospital. The regional distributor with 40 employees and a spreadsheet-based operation that a competent AI implementation would transform completely in two weeks.
The skill set that matters here is not ML research. It is understanding how a messy, unglamorous 50-person company actually operates, and then wiring intelligence directly into their revenue. That combination, domain empathy plus AI implementation, is genuinely rare right now. And it is not a job title. It is an entire economic class being born.
The uncomfortable part
I will be direct about where this gets uncomfortable. The same dynamic that creates opportunity for skilled implementers creates real disruption for the workers inside those 33 million companies. When the AI conforms to the business instead of the other way around, the business needs fewer people to run it. Cuban frames this as wealth transfer to small businesses. That is accurate at the owner level. What happens at the employee level is a different and harder conversation.
I don’t have a clean answer to that. But I would rather name it than pretend the upside exists without the displacement.
Where that leaves us
The fortunes will belong to whoever teaches the models a trade. The factories, the trucking routes, the accounting workflows, the insurance intake, the dental back office. Right now the gap between “AI exists” and “AI works inside my specific business” is enormous.
That gap is the opportunity, and it will not stay open indefinitely.
It is also why I built GRC the way I did. Productized engagements, fixed fees, industry-specific knowledge. The businesses we serve don’t want another sales-driven consultant. They want someone who will sit in their operations for two weeks, point at the three things that matter most, and either build them or tell them who should.