AI writing code sounds like every CEO’s dream: cheaper, faster, maybe even better. But what if swapping expensive developers for clever tools just creates a new kind of risk?
This morning I noticed that Cursor — an AI coding tool — dropped a confusing pricing update. Users were furious. And it made me think.
Every CEO dreams of it: AI that writes, fixes, and updates code. No more expensive developers. Faster releases. Maybe even a competitive edge. Makes perfect sense, right?
But what happens if that world really arrives? A world where AI tools write a large part — maybe even most — of your production code. You’re no longer paying high salaries for developers. Great! Instead, you’re paying for access to AI tooling. And that might change the game completely.
But that also means giving up some control: your business becomes dependent on external tool providers — on their pricing, reliability, and roadmap. And that might change the game completely.
Sure, the big players might be able to self-host large models and according Agent Frameworks. Especially if the technology matures, and models become smaller and cheaper to run. But will that be an option for the average company? Probably not. Most will just use whatever hosted service is out there.
And here’s the catch: there won’t be thousands of providers. Most likely just a handful of big players. And those players can change their prices overnight.
Developers can’t easily coordinate and ask for collective raises. And it’s the same with external service providers: there are just so many of them that prices tend to move slowly, shaped by supply and demand. But a few AI providers? They could act almost like a cartel, whether intentionally or just by following each other’s moves.
Sure, companies might try to protect themselves with long-term contracts. But the risk doesn’t go away. Instead of being dependent on expensive developers whose cost moves predictably, you become dependent on a few tool vendors who could raise prices suddenly — and massively.
And then the dream starts to crack. Instead of saving money by replacing developers with cheap AI, you might end up paying just as much — or even more — but to a different group entirely.
Is that scenario even realistic — and is this really the trade we want to make?
And maybe there’s hope: if AI tooling becomes truly commoditized, we might see a broad market with many providers competing fiercely — bringing prices down and giving companies more control. But how likely is that future, really?
Leave a Reply