Weigh in on the Desirable Properties
If you represent American Security Foundation, confirm the source URLs, then open the Desirable Properties tab and leave a patch idea where the property text misses what your public work already argues.
These directions are generated from American Security Foundation's public corpus, not from private landing pads. Each one is a hypothesis. Follow the interest that feels like your work. The Hermes prompt and Discuss link are real contribution paths, not a brochure.
DP8 · hypothesis
Collaborative Environment and Meta-Communities
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP8's human issue: Online communities constantly restart. Guiding question: How do communities persist beyond platforms?
Direction to explore
How do Alliance members share a collaboration space without collapsing into one platform community?
Candidate patch idea
Define meta-communities as optional shared rooms with explicit consent, not an ambient social network that harvests the Alliance listserv.
DP20 · hypothesis
Community Ownership
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP20's human issue: Platforms own everything users create. Guiding question: Who should own the next layer of the Internet?
Direction to explore
This is the anti-capture test: can American Security Foundation's coordination space stay owned by people if commercial players arrive later?
Candidate patch idea
Strengthen anti-capture language: community rules and exit rights are written before large commercial platforms are invited to play.
DP11 · hypothesis
Safe and Ethical AI
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP11's human issue: AI hallucinations and harmful outputs. Guiding question: What does trustworthy AI actually look like?
Direction to explore
If Labs ships AI products, which safety claims belong in DP11 versus product policy?
Candidate patch idea
Require glass-box disclosure when AI drafts norms that later bind a community, including this briefing itself.
DP13 · hypothesis
AI Containment
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP13's human issue: AI manipulation and agent overreach. Guiding question: How do we keep AI helpful without letting agents overreach?
Direction to explore
What containment is needed before protocol-adjacent AI agents act in a shared layer?
Candidate patch idea
Sandbox agent write-access until a named human steward accepts the action against an ML-REQ.
DP14 · hypothesis
Trust and Transparency
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP14's human issue: Misinformation, propaganda, and deepfakes. Guiding question: How do we rebuild trust in an AI-powered Internet?
Direction to explore
How should American Security Foundation show when Hermes inferred from public pages versus when a human confirmed?
Candidate patch idea
Require provenance on AI-assisted governance artifacts: source URLs, confirmations, and remaining hypotheses.
DP4 · hypothesis
Data Sovereignty and Privacy
Your stated mission ("ASF Welcomes America's AI Action Plan ASF welcomes the White House's "America's AI Action Plan" to advance America's tech leadership in the AI age.") sits next to DP4's human issue: Companies own your data. Guiding question: Who should own your data?
Direction to explore
Where does American Security Foundation's “people own their data” language need to become an enforceable affordance?
Candidate patch idea
Require that personal data created in a layer can leave with the person, including social graph, without a platform petition.
