Every Dollar You Spend on AI Makes Someone Else Richer

Emma Merrill

Six-hundred fifty billion dollars. That’s what hyperscaler companies are committing to AI in 2026. Actual enterprise AI revenue doesn’t keep up. A recent study from MIT revealed that 95% of the 52 organizations considered gained zero return on investment: a small sample, but consistent with broader signals. Earlier this year, a National Bureau of Economic research study found that 90% of firms reported no productivity impact from AI. Furthermore, less than 30 percent of CEOs were satisfied with their generative AI returns.

The conventional diagnosis is that the technology isn’t mature enough yet. The models will get better. The ROI will come.

I disagree. Yes, implementation failures and unrealistic timelines are real, but they’re symptoms, not the disease. The technology is already powerful. The deeper problem is who owns it.

Subscriptions Are Rented Capability

The companies spending the most on AI are getting the least. Why? Because they don’t own anything they’re spending on. They spend to rent, not own

Consider your enterprise AI subscription. You get an API from one of the major model providers. Then, each year, your subscription fee and your switching costs go up. Meanwhile, your vendor demands more while providing less value. And in many agreements, your data is used to improve their model. Your data is used to improve their model. You spend money for your vendor to become more valuable. Not you.

Every dollar you spend on SaaS makes your vendor more valuable. Not you.

Companies aren’t failing at AI because the models are bad. They’re failing because they’re renting capabilities they should be owning. This is a depreciating liability: the structural problem poised to pop the dominant business model in AI today.

Tools vs. Outcomes

The next great technology company will sell outcomes rather than tools.

Here is the distinction: most AI vendors sell you a tool. You’re left to figure out what to do with it. You need to hire a team to maintain it, and if there are ever rate limits or an outage, you eat the costs.

But the deeper problem is accountability. When something built on AI fails to deliver, the vendor doesn’t answer for it. You do. Your team does. Your job does. The model doesn’t get fired. You do: for work you didn’t produce, using a tool you don’t fully control, toward outcomes nobody contractually guaranteed.

That accountability gap is the real crisis. And there is a better way. An AI partner sells you the work and takes accountability for the outcomes. That is OpenTeams.

The Third Option

Every organization that wants to use AI seriously faces the same fork in the road.

Option one: send your data to a vendor and accept the terms described above.

Option two: build your own stack. To maintain it, you need to hire infrastructure engineers. Instead of using AI, you become an AI infrastructure company.

Most organizations can’t afford option two. And they shouldn’t accept option one.

There is a third option. You own the infrastructure, the code, the data, the models, the configurations, deploy it on your terms, and someone else manages it. Not a vendor that locks you in. A partner who keeps it running.

That is the promise of OpenTeams: to make our clients more capable, not more dependent.

What the Survivors Will Look Like

The organizations left standing will be the ones that own their AI infrastructure: that treat AI as an asset to build, not a service to rent indefinitely.

The choice isn’t really between vendors. It is between dependency and ownership. And the window to make that choice on your own terms, before switching costs make it for you, is right now.

Control

April 9, 2026

I’ll be brief. I value your time. You’re in control.

Every Dollar You Spend on AI Makes Someone Else Richer

Emma Merrill

Subscriptions Are Rented Capability

Tools vs. Outcomes

The Third Option

What the Survivors Will Look Like

Share:

Related Articles

Control

Follow Us:

Collab (Desktop App)

Nexus (Intelligence Hub)

Closing