User Tools

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
committee_meeting_2026-09-14 [2026/09/14 19:53] – sarahcommittee_meeting_2026-09-14 [2026/09/14 21:30] (current) – [Agenda] sarah
Line 10: Line 10:
 ===== Agenda ===== ===== Agenda =====
   * Memberships process:   * Memberships process:
-    * How to add ppl to Matrix: Jade needs to write this down 
   * Everything Open Talks   * Everything Open Talks
   * AI Policy    * AI Policy 
Line 50: Line 49:
   * Peopling computing:    * Peopling computing: 
     * To what extent should we make the tech-based interactions human? Should people see the name and face of the person who's providing tech or will that lead to finger pointing?      * To what extent should we make the tech-based interactions human? Should people see the name and face of the person who's providing tech or will that lead to finger pointing? 
 +    * We can't fully throw a question like this over to designers: There are security implications which impact on risk, accessibility, compliance. Frameworks that we can work through e.g. Cade offers the ANXIETY framework. 
     * Security model?      * Security model? 
     * Auditable and transparency rather than control      * Auditable and transparency rather than control 
     * Wade: "embarrassingly transparent"     * Wade: "embarrassingly transparent"
 +
 +==== AI Policy ====
 +
 +  * For Jade the big AI issue is the oligarch cash grab and power grab. 
 +  * Economic consequences; circular debt and risk of bubble bursting for economic stability. 
 +  * Secondary concerns: energy and 
 +
 +On the other hand AI works: in some circumstances, can be used to generate 
 +
 +Solarpunk policy: carve-out for tinkerers to run agentic AI if code is up to our standard, but not:
 +   * pay big tech
 +   * if it runs in a data centre; must run locally (within 20kms), fully powered (actually, not just buy green power) 
 +   * alternatives: a more strict policy, i.e. without the loophole or targeting engineering quality: anti-slop to ensure the code quality is maintained. 
 +
 +Wade: it's not intelligent so we should put that framing in to the preamble. It's not refusal so we should consider carefully how this might be succeeded 
 +
 +Sarah: We should set a process timeline or overall framework for revisiting the policy. 
 +
 +AI enables us to work without collaboration. 
 +
 +Sarah: let's discuss whether to remove the rationale to e.g. a blog post. 
 +
 +
 +
 +
    
  

This website uses cookies. By using the website, you agree with storing cookies on your computer. Also, you acknowledge that you have read and understand our Privacy Policy. If you do not agree, please leave the website.

More information