Imagine this scenario is yours:

For years, you've been syncing your service with a popular platform through that platform's free, public access. You've sent that platform thousands of customers along the way. It's been a good relationship, or so it seemed. Then, last week, the platform tells you the free access is ending. Going forward, it will cost somewhere between $200k and $1.0m a year for the same thing you've been getting for nothing.

From free to a million dollars. For access you helped make valuable in the first place. That would suck, to say the least.

Now this didn't happen to you, but it did happen to a business owner I know. And while you're likely never going to use the platform in question, I want to explore why this still matters.

Right now, we're living through a remarkable moment in technology. The ability to share and digest massive amounts of data is unprecedented, and AI in particular lets us learn and do things that weren't possible even a few months ago. But to see where this is heading, you have to look at how these companies actually make money, which is to say, right now, most of them don't.

Whether you're on a free plan, a paid plan, or a business account, what you're paying very likely doesn't cover what it costs to serve you. Between the usual costs of running a software company and the enormous power draw of the data centers behind every request, the monthly fees just don't add up. (For context: a single large data center can consume as much electricity as a small city. That's the scale we're talking about.)

Reasonable people disagree about how fast those costs will come down, and some will. But I don't think you have to resolve that debate to sit with the basic implication: a lot of the pricing we're seeing today isn't sustainable under current conditions. And that means at some point it changes. How much you pay, what you get for it, or both.

There's a second force building on top of that one. There's growing talk of the major AI companies moving toward the public markets. And a public company comes under real pressure to protect shareholder investment and grow long-term value.

So ask yourself: if you were a shareholder, would you want the company to keep running an unprofitable model indefinitely? Or to start charging what the product actually costs to deliver?

You can see where this is heading.

None of this is a reason to stop using AI. I use it regularly for administrative support. But it is a reason to ask a question most owners haven't: how much of my business runs on top of a tool whose price and terms I don't control, and maybe more importantly, that I don't have a real alternative to?

If the cost of that tool doubled tomorrow, or the free tier you'd built a workflow around disappeared, what would happen? Could you absorb it? Pass it through? Or would it quietly break the math your business depends on?

That's the position the business owner I mentioned is in right now. And to be clear, they didn't do anything wrong. They made a smart bet on a platform that worked, and it did work, for years. The terms changed, not their judgment. But now, they're facing a number that reshapes their economics, with no easy way out, because their customers are already on the other side of it.

I don't think most of us are at that point yet. But I do think we're in a golden age of cheap, abundant AI, and golden ages are hard to appreciate from the inside.

To me, the question isn't whether you use AI. It's whether you've built any optionality into your business for when the terms change. Because they will.

~ Nolan​
​