Good question. Mycellm is the protocol layer: identity, routing, credits, reputation. Similar to how BitTorrent verifies pieces match their hash but doesn't judge whether the content is any good (that's left to trackers, communities, external tools).
Mycellm can verify that inference happened: signed receipts, reputation tracking (success rate, speed, contribution history), admission control that cuts off freeloaders. Each network sets its own policies on top; a private group might trust everything, a public network might add spot-checks or consensus routing.
Directions I'm exploring for public verification: fast-of-N routing, spot-checking with known outputs, consensus at temperature=0 where inference is deterministic. But the verification logic should be pluggable — different networks, different standards. That's a design area where I'd welcome input.
Curious about Runfra — when you say 'Uber for GPU', is that an orchestration/marketplace layer where independent GPU owners sell compute time? And what does 'batch-first creative workflows' look like — image gen pipelines, video, or broader?
That BitTorrent analogy is spot on. For Runfra, I’m mostly focusing on the orchestration and inference pipeline. Trying to fix the waiting problem, but also thinking a lot about yield, like how many actually usable outputs you get per run. Right now, I’m just dogfooding a small cluster of my own 3x 4060s. Since these are basically idle home GPUs, I have the luxury to bias toward quality over latency. If this ever turns into a marketplace, something like Mycellm for node reputation would be a lifesaver. Honesty is definitely the hardest nut to crack in decentralized compute. The batch first idea is really my attempt to kill that “slot machine” workflow where you have to babysit every single prompt. Instead, you just fire off a bunch and walk away. The scoring layer (CLIP / aesthetic filters) acts as an automated quality gate, quietly filtering out the "AI slop" in the background and only surfacing the winners.
Since I need $Temperature > 0$ for that creative randomness, I inevitably get a lot of junk. So I’m essentially treating these idle GPUs as a distributed filtration layer, and trading idle time for guaranteed better outputs.
This is exactly what I've been building — mycellm.ai. Seed GPU when idle, earn credits, spend on inference. BitTorrent economics for LLM compute. It's live now (my homelab serves the public chat). Would love to have you involved — github.com/mycellm/mycellm
> Amazon says it has acted properly in response to FTC demands to ensure the preservation of relevant evidence. The company says it gave executives “explicit instructions on how to disable Signal’s DM [disappearing messages] feature.” A copy of those instructions has been provided to the FTC, according to Amazon’s filing.
Amazon's 'explicit instructions' on keeping messages visible: either a move for clarity or a masterclass in corporate subtlety.
This from the same company that insists on auto-deleting all old Slack and e-mail messages, to limit discoverability. And refuses to give you a larger e-mailbox than 4GB, for the same reason.
Washington state makes it so easy. My ballot is mailed to me weeks before the election date. I have plenty of time to research and fill it out. I then drop it off at one of the many official dropboxes around the city (or you can drop it in the mail for free, as long as you get it postmarked before the election date). The ballot is tracked from the time it's sent, received, and counted (I have mine set to alert me of its status via email and SMS).
I have a similar experience in CA. It's become a joy to sit down for an afternoon with a glass of wine, researching all the down ballot races as I fill in bubbles.
Mycellm can verify that inference happened: signed receipts, reputation tracking (success rate, speed, contribution history), admission control that cuts off freeloaders. Each network sets its own policies on top; a private group might trust everything, a public network might add spot-checks or consensus routing.
Directions I'm exploring for public verification: fast-of-N routing, spot-checking with known outputs, consensus at temperature=0 where inference is deterministic. But the verification logic should be pluggable — different networks, different standards. That's a design area where I'd welcome input.
Curious about Runfra — when you say 'Uber for GPU', is that an orchestration/marketplace layer where independent GPU owners sell compute time? And what does 'batch-first creative workflows' look like — image gen pipelines, video, or broader?
reply