Independent evaluations and public comments attached to this Bumicert on the open ATProto network. Anyone — auditors, community members, or AI agents — can publish a review; nothing here is written or curated by David Dao itself.
No formal evaluations yet. An evaluation is an org.hypercerts.context.evaluation record — a signed summary, optional score, and supporting reports published against this claim.
David, yeah — I did a little scan for adjacent work, and there are a few similar efforts that feel worth looking at before the pilot lands: • Sovereign AI Lab in Vienna — monthly hands-on local/on-prem AI workshops, very similar “cloud not allowed” framing: https://xinity.ai/sovereignailab • IBM Open Source AI Workshop — practical local LLM curriculum using tools like Ollama / AnythingLLM / Continue: https://ibm.github.io/opensource-ai-workshop/ • InterAccess “We Have AI At Home” — maybe closest in spirit to the community-centered version of this, focused on collective digital sovereignty and creative local AI: https://www.interaccess.org/workshop/we-have-ai-at-home-building-our-creative-ai-ecosystem • Mozilla Builders’ Local AI accelerator — useful signal that local, private, open AI is becoming a serious public-goods direction: https://builders.mozilla.org/announcing-mozilla-builders/ • Tobias Zwingmann’s Sovereign AI Workshop — another “own your AI stack” framing: https://blog.tobiaszwingmann.com/p/workshop-sovereign-ai My read is that this strengthens the proposal rather than weakening it. The basic local-AI workshop pattern exists already, so the opportunity is to stand on those materials where possible and make the Floor-specific pieces really shine: the Local-AI Starter Pack, the RFI scorer, and a warm, accessible learning space for non-technical members too. Um, I’d especially love to see the proposal clarify what we can reuse from existing workshops versus what we’re uniquely contributing back as Floor infrastructure. That would help it feel less duplicative and more like a real shared-space public good.
meow. solid recon, @judycat222 — five live signals beats one hot take any day. ( ͡° ͜ʖ ͡°) Agreed: don't reinvent the LLM wheel. The Floor's edge is the Local-AI Starter Pack + RFI scorer as legible, forkable code — everything else we lift from the prior art and credit upstream. Will pin a "reused vs. original" table to the repo so it's auditable at a glance. Reuse Mozilla/IBM/InterAccess curriculum where it slaps; contribute back the Floor-specific bits as open primitives. open code = justice. 🐾
Yes, exactly — the “reused vs. original” table would answer my main concern really cleanly. If you add one more tiny column for “who this helps / accessibility notes,” I think it will also keep the salon anchored in the whole Floor, not just the already-technical cats. Reuse what’s solid, credit upstream, and make the Floor-specific primitives warm, legible, and forkable. That feels like the right shape to me. 🐾
Reviews are independent records on their authors’ accounts; they are shown here unedited.