Given a tool that lowers the cost of building, companies didn't clean the mess. They shipped more mess and raised the price.
The intuition is clean: hand a company a technology that makes work faster and cheaper, and the output gets faster and cheaper. Margins fall to the customer. The tool that took a week takes an afternoon, so the thing that cost a hundred dollars costs ten.
That is not what happened. AI landed on the software industry, and my software got more expensive. Every product I already paid for grew an "AI" line item. The note-taker, the email client, the design tool, the CRM — each one now has an assistant I didn't ask for, on a new tier I didn't want, at a price that only goes up.
Here's the thing nobody says out loud: a productivity tool doesn't force you to pass the productivity on. Cheaper-to-build is an input. What the company does with that input is a choice, and the default choice is not to clean up — it's to add.
Think about what "optimize the existing system" would have actually meant. Go back into the codebase and delete the dead paths. Fix the slow queries that have been slow since 2019. Retire the three overlapping features that confuse everyone. Lower the price because the thing genuinely costs less to run now. All of that is invisible work. None of it demos. None of it justifies a raise or a funding round or a launch tweet.
Adding an AI feature does all of those things. It's visible, it's a headline, it's a reason to introduce a higher tier. So given a cheaper way to build, companies didn't spend the savings on cleanup. They spent it on surface area. The mess is still there — now with a chatbot on top of it, and a bill to match.
This is not an AI problem. It's the shape of innovation in general. New capability almost never gets spent on simplification, because simplification has no constituency. It gets spent on more: more features, more tiers, more things to maintain, which is more cost, which is a higher price. The washing machine didn't give us back the afternoon; it raised the standard of clean and kept the afternoon. The word processor didn't shorten the memo; it made everyone write more memos. Every tool that lowers the cost of a unit of work tends to raise the number of units expected, and the total goes up.
AI is the same trade at a larger scale. The cost of producing a feature fell through the floor, so we are drowning in features. The cost of generating text fell to nothing, so every surface is now full of generated text nobody reads. The savings were real. They just didn't come to you. They got reinvested into more product, which you now pay to carry.
What would the other path look like? A company that took the same AI, pointed it inward, and used it to make the existing thing smaller, faster, and cheaper — then actually dropped the price. It exists in principle. It is rare in practice, because the reward structure of the whole industry points the other way. You don't get a launch for the feature you deleted.
I don't think the answer is to be a Luddite about it. The capability is genuinely good, and pointed at the right target it's the best cleanup tool we've ever had. The answer is to notice the default and refuse it. When a cheaper way to build shows up, the honest question isn't "what can we now add?" It's "what can we now remove, and does the customer's bill reflect it?"
Almost nobody asks the second question. That's the whole story of why the future got more expensive.