---
title: "Talking to Robots: Uses and Abuses of LLMs in Communities"
license: CC-BY-2.0
license_url: http://creativecommons.org/licenses/by/2.0/be/deed.en
license_note: "The FOSDEM video directory https://video.fosdem.org/2025/ub5230/ states: This work is licensed under the Creative Commons Attribution 2.0 Belgium Licence, with the deed at http://creativecommons.org/licenses/by/2.0/be/deed.en. Confirmed 2026-10-18. The VTT file is the publisher caption of that same recording. This note is an original summary, not a transcript."
source: https://archive.fosdem.org/2025/schedule/event/fosdem-2025-5215-talking-to-robots-uses-and-abuses-of-llms-in-communities/
source_license: "CC BY 2.0 Belgium"
attribution: "Summary by iinventedlight (AI-assisted). Source: David Allen, Talking to Robots: Uses and Abuses of LLMs in Communities, FOSDEM 2025 Community devroom (CC BY 2.0 Belgium)."
date: 2026-10-18
status: published
category: talks
tags: assistants, communities, prompting, agents
checked_on: 2026-10-18
sources:
  - id: S1
    title: "Talking to Robots: Uses and Abuses of LLMs in Communities"
    authors: "David Allen"
    url: https://archive.fosdem.org/2025/schedule/event/fosdem-2025-5215-talking-to-robots-uses-and-abuses-of-llms-in-communities/
    accessed: 2026-10-18
    license: CC BY 2.0
    attribution: "David Allen, Talking to Robots: Uses and Abuses of LLMs in Communities, FOSDEM 2025 Community devroom (CC BY 2.0 Belgium)."
  - id: S2
    title: "Publisher captions (VTT) for the same recording"
    authors: "FOSDEM"
    url: https://video.fosdem.org/2025/ub5230/fosdem-2025-5215-talking-to-robots-uses-and-abuses-of-llms-in-communities.vtt
    accessed: 2026-10-18
    license: CC BY 2.0
    attribution: "FOSDEM, captions for David Allen, Talking to Robots: Uses and Abuses of LLMs in Communities (CC BY 2.0 Belgium). License stated on https://video.fosdem.org/2025/ub5230/."
---
# Talking to Robots: Uses and Abuses of LLMs in Communities

License: [CC BY 2.0 Belgium](http://creativecommons.org/licenses/by/2.0/be/deed.en). Source: https://archive.fosdem.org/2025/schedule/event/fosdem-2025-5215-talking-to-robots-uses-and-abuses-of-llms-in-communities/

## Summary
David Allen, speaking in the FOSDEM 2025 Community devroom on 1 February 2025, treats large language models as a hype cycle that community teams will be pushed to adopt. His test is whether a bot adds to human contact or replaces it. Two rules follow: say that it is a bot, and make the bot opt-in, with a narrow exception when a bot acts and a person follows up quickly, the way some GitHub bots already do. He names patterns rather than products. An intern bot can label issues if it is disclosed and the code is open. A researcher bot can summarize what people already said so organizers can read it, while members never talk to the bot. A robot in jail lives in one channel people enter on purpose. An egg-drop or sidebar bot speaks only when a trusted person calls it. A "superintelligent donut" that matches people by interest is a joke about a thing he has not built. Anonymized cases: a company that automated support never grew a community, and a company that cut support cost saw engagement fall. For 2025 he recommends the intern, the researcher, and the jailed bot, and he warns against both zero use and full automation. People, he says, are not technical problems to solve. He points at a public thread about pull requests that are mostly model-written, at open-source policy writing, and at the Contributor Covenant, and he closes on a line from Richie Hartmann's keynote: be excellent to each other. Captions garble some organization names, so this note does not repeat them.

## Why it is here
The library's assistant notes are about how a person uses a model. This talk is the other side: when a community puts a model in front of people, disclosure and opt-in matter more than a smarter prompt, and "superintelligent" is a label he uses for a matching toy, not a system that exists.