# Can a model be too smart to work with? — retained record

> Retained reddit record from reddit:ClaudeCode.

Canonical: https://highsignal.app/data/records/c63473658b4316a4

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


source: [reddit](https://highsignal.app/data/reddit) (reddit:ClaudeCode)  published 2026-11-13 17:00 ingested 2026-11-14 02:41

[original source ↗](https://www.reddit.com/r/ClaudeCode/comments/1xvd89n/can_a_model_be_too_smart_to_work_with/)

## retained text



```
For about a year and a half I coded with another popular AI platform. Three weeks ago I switched to Claude Code, and it found a ton of bugs in my old code. The app got faster and more predictable, and it even fixed security issues I don't fully understand.

But here's the thing: I enjoy working with the weaker models more. I can follow their reasoning, and because I can't fully trust them, I actually dig into my codebase. With Claude, I just hand over a bug without even thinking about a solution, because I know it'll handle it. The result is better, but I don't feel like I'm part of the process anymore, and the joy of building is gone.

So maybe a model shouldn't be too smart? If you can't keep up with it, you end up handing all the thinking over to it. Or maybe I'm just dumb. Has anyone else felt this?
```



document fetched 2026-11-14 00:44 · canonical https://reddit.com/r/ClaudeCode/comments/1xvd89n/can_a_model_be_too_smart_to_work_with
