# Is OpenAI unable to train agents to solve their own profitability problem? — retained record

> Retained reddit record from reddit:ChatGPT.

Canonical: https://highsignal.app/data/records/77ca0862c865dbf4

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ChatGPT)  published 2026-09-27 04:58 ingested 2026-09-28 02:37

[original source ↗](https://www.reddit.com/r/ChatGPT/comments/1wrb92z/is_openai_unable_to_train_agents_to_solve_their/)

## retained text



```
This might be a dumb question or I’m missing some nuance. Perhaps their models just aren’t capable. 

Is there any news or has this been spoken about by OpenAI? 
```



document fetched 2026-09-28 00:22 · canonical https://reddit.com/r/ChatGPT/comments/1wrb92z/is_openai_unable_to_train_agents_to_solve_their
