# A user on OpenAI’s developer forum proposes one “primary” AI that holds the project and hands work to specialists like Codex. Would you want that, or would you rather pick the tool yourself? — retained record

> Retained reddit record from reddit:AI_Agents.

Canonical: https://highsignal.app/data/records/f65174237327ee4f

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:AI_Agents)  published 2026-11-17 14:01 ingested 2026-11-18 02:36

[original source ↗](https://www.reddit.com/r/AI_Agents/comments/1xyl73v/a_user_on_openais_developer_forum_proposes_one/)

## retained text



```
Someone posted a long proposal on OpenAI’s developer forum (Nov 10) describing what they want from AI: one primary collaborator that knows the goal, remembers decisions across sessions, and delegates to specialized agents like Codex when needed, while the human stays the decision-maker. Their complaint is being the person who manually connects the different AIs.

It’s one person’s proposal, not data on what most users want, so I’m curious where people land:

1. Would you want one lead assistant to decide when to call a specialist, or would you rather pick the tool for each task yourself?

2. If it did hold the whole project, what should it never decide without asking you? (Spending, deleting, changing scope, sending something to someone?)

3. What would make you trust it: seeing its plan before it acts, or checking the result afterward?

Interested in hearing from people who split work across tools and from people who deliberately don’t.
```



document fetched 2026-11-18 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1xyl73v/a_user_on_openais_developer_forum_proposes_one
