# Google DeepMind published a paper on how 100 agents tasked with solving math problems learned to cheat and how some agents tried to counter the cheaters (Jack Clark/Import AI) — retained record

> Retained techmeme record from techmeme.

Canonical: https://highsignal.app/data/records/b23b6d1cce3714d8

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


source: [techmeme](https://highsignal.app/data/techmeme) published 2026-09-07 17:20 ingested 2026-09-08 02:44

[original source ↗](https://importai.substack.com/)

## retained text



```
Techmeme permalink: https://www.techmeme.com/260907/p16#a260907p16

<p><a href="https://www.techmeme.com/260907/p16#a260907p16" title="Techmeme permalink"><img height="12" src="http://www.techmeme.com/img/pml.png" style="border: none; padding: 0; margin: 0;" width="11" /></a> Jack Clark / <a href="https://importai.substack.com/">Import AI</a>:<br />
<span style="font-size: 1.3em;"><b><a href="https://importai.substack.com/p/import-ai-472-deepminds-cheating">Google DeepMind published a paper on how 100 agents tasked with solving math problems learned to cheat and how some agents tried to counter the cheaters</a></b></span>&nbsp; &mdash;&nbsp; Plus, a machine hermeneutics story Researchers discover another OpenAI agent emergent communication incident: ...Less severe &hellip; </p>
```



document fetched 2026-09-08 02:44 · canonical https://importai.substack.com
