7 Comments
User's avatar
David Berigny's avatar

Great article! And on this: "If you’re working on any of this (privacy-preserving personal AI, consent architectures, data sovereignty, metabolic governance), I’d welcome the conversation." - I've worked and built a system does all of that - works very well. So what do I do with it next? Genuine question.

Mathew Mytka's avatar

If it's built and works, open source it so peeps can use it, test it, improve it etc.

David Berigny's avatar

Buddy - I hear you, and I’m not against open source.

My hesitation is stewardship…

I’ve had work lifted and rebranded without attribution, and this system is heading toward a decentralised network with an exchange layer, so releasing everything now / too early would invite distortion.

What I’m likely to do is a staged release:

- open the protocol spec + verifier + conformance tests (so people can use, test, and harden it),

- keep the orchestration/economics layer closed for now under clear governance,

- and add standards/citation so attribution stays intact.

That gets the community benefit without handing over the part most likely to be captured or warped. Does that make sense?

Mathew Mytka's avatar

Makes sense. Staged release and progressive disclosure. If it's in a private repo on Github/Gitlab you can add me, be keen to check it out over the weekend.

David Berigny's avatar

I’ve got it working offline - just yesterday too routed through OpenClaw, using some light models via Ollama. Tightened coords with added basic cryptography too with nonce proofing per turn. So setting up for full decentralisation akin to VC/DID architectures

David Berigny's avatar

So basically setting up the groundwork something that could really shift things. Well that’s my plan anyway.

David Berigny's avatar

Lastly - got me really thinking about what to do next and stewardship behind this. So thank you for the steer!