The pricing page for OpenAI's API reads like a restaurant menu written in a foreign language.
Millions of developers now build on it, yet almost none can explain at a dinner party what they actually pay.
Here's the uncomfortable truth: most Americans interact with this pricing constantly without knowing it.
The customer service bot that answers your insurance question, the app that summarizes your meeting notes, the therapy chatbot your teenager confides in — nearly all of them rent intelligence by the token from a handful of companies.
You pay one rate to feed text in and a higher rate to get text out.
Thinking costs more than listening, and the meter never stops running.
The sticker prices have fallen dramatically, which sounds like good news.
Cheaper models, aggressive competition from Google and Anthropic, the rise of open-source alternatives — the race to the bottom has been real.
But falling unit prices hide a quieter problem: usage is exploding faster than prices are dropping.
We are consuming so much more that the monthly bill climbs even as the rate card shrinks.
Then there's the part nobody puts on the marketing page.
A chatbot that "remembers" your conversation resends the entire history with every new message.
Long chats get quadratically more expensive.
Companies have quietly started truncating memory, summarizing old context, or simply forgetting what you told them — not because they lack the technology, but because remembering you costs money.
The result is a new kind of customer experience: amnesia by accounting.
Your bank's assistant loses track of your issue.
Increasingly, you do, through subscriptions that creep upward.
The "unlimited" plan gets a fair-use clause buried on page nine.
A tool that felt magical in 2023 now feels like a toll road.
We were told AI would be abundant, cheap, and everywhere.
Instead we're getting metered, throttled, and nickeled.
When you call a company and reach a bot, no one discloses that the conversation is being billed by the syllable.
When an app summarizes your document, you don't see the invoice.
The cost is real, it's just invisible — shifted onto the product's price, the service's quality, or your patience.
Americans deserve to know when they're talking to a machine and roughly what that conversation costs.
Not because we need to audit every API call, but because a society that outsources its thinking to rented cognition should at least understand the rental terms.
We built an economy where attention was the currency.
Now the currency is attention plus context plus compute, and the meter runs whether you're watching or not.
My take: falling prices are not the same as affordability, and cheap tokens are not the same as a cheap society.
Final Thoughts
If companies won't explain what they're spending on AI, we should assume we're the ones paying for it — in dollars, in degraded service, and in machines that forget us the moment remembering gets expensive.