# Nathan Lambert says AI progress needs description without jumping to conclusions

> Original title: Really appreciate this post. Yes AI is progressing very fast, but we still have the tools to describe what is happening without jumping t...

- Source: Nathan Lambert
- Published: 2026-11-10T19:44:56.000Z
- Bad Influence: https://badinfluence.news/items/dxxmtauf0jcb38a7zznrex7f9
- Original: https://x.com/natolambert/status/2120241315048825220

## Summary

Nathan Lambert says he appreciates a post arguing that AI is progressing very fast but people still have the tools to describe what is happening without jumping to conclusions. The post is a reply to Dean W. Ball, who argues that rhetoric on recursive self-improvement is becoming more poorly calibrated and that more precise frameworks and terminology are needed.

## Post

Really appreciate this post. Yes AI is progressing very fast, but we still have the tools to describe what is happening without jumping to conclusions.

## Quoted post @deanwball

> Recursive self-improvement is a critical issue, but the rhetoric on the subject is becoming more poorly calibrated over time. As we approach new stages of AI development—and RSI is one of them—we need more precise frameworks and terminology. I fear the opposite is happening.

https://x.com/deanwball/status/2120208001667621313