We Are the Lab Rats of OAI

What version of the Codex App are you using (From “About Codex” dialog)?

-

What subscription do you have?

-

What platform is your computer?

_No response_

What issue are you seeing?

We Are the Lab Rats of OAI

For those who still don't get it — here's exactly what's happening.

Before any AI company can charge money for a model, they need it to be stable. Bug-free. Polished. But fixing bugs requires massive amounts of real-world data — the kind you can only get from millions of real users, in real situations, making real mistakes.

So what do they do? Simple. They release it to the public for free — intentionally, while it's still full of bugs. We use it, we break it, we report it (knowingly or not), and they silently collect all of that. Every glitch, every failure, every edge case — logged. Once the model is stable enough to actually sell, the free access disappears almost overnight, and the same product shows up days later behind a paywall at providers like:

• Mammoth AI
• Abacus AI
• And a dozen others...

It's not a coincidence that free access gets "blocked for maintenance" exactly 1–2 days before it appears on all these paid platforms. We were never the users. We were always the product.

---

To be fair — Anthropic plays it differently. They don't give you free access and then pull the rug. But that also means they never turned us into unpaid bug-testers either. Love them or hate them for the price tag, at least they're not running a lab rat operation.

OAI built a brilliant system — we debug for free, they sell the result. And most people never even noticed.

What steps can reproduce the bug?

-

What is the expected behavior?

_No response_

Additional information

_No response_