Feed aggregator

Show HN: Lumabri – What if LLMs worked like Napster?

Hacker News - Sun, 08/09/2026 - 6:24pm

A while ago I started working on Colibrì to see if it was possible to run huge LLMs on a normal computer. The project grew far beyond what I expected, thanks in large part to the HackerNews community.

That led me to a new question:

What if we stopped thinking about one computer?

This is the idea behind Lumabri.

Instead of requiring a single machine to store and run an entire huge model, Lumabri treats a network of normal computers as a shared pool of resources.

One machine might provide disk space, another compute, another a different part of the model. If a required block or expert isn’t available locally, the system can retrieve or execute it on a peer.

This is particularly interesting for Mixture-of-Experts models. A model can have hundreds of billions of parameters, while only a fraction are activated for each token. Rather than moving huge expert weights over the network, Lumabri can send the small activation to a peer that already has the expert and let it execute it.

The goal is for machines to contribute whatever resources they can afford while using the swarm for the rest.

The idea is very much inspired by peer-to-peer systems: users are the infrastructure.

There are obviously major challenges, especially network latency and security. I’m experimenting with peer verification, SHA-256 verification, signed model state, replica selection, failover, and deterministic execution.

Lumabri is still an early experiment. I don’t have a datacenter or a huge GPU cluster, so I’m building it with the hardware I have and trying to find out whether the idea actually makes sense.

With Colibrì I asked:

Can one normal computer run a huge LLM?

With Lumabri I’m asking:

What if many normal computers could become one huge computer?

Feedback welcome.

Repo: https://github.com/JustVugg/lumabri

Comments URL: https://news.ycombinator.com/item?id=49236781

Points: 1

# Comments: 0

Categories: Hacker News

3G at Last

Hacker News - Sun, 08/09/2026 - 6:17pm
Categories: Hacker News

Neural Scaling Law

Hacker News - Sun, 08/09/2026 - 6:10pm
Categories: Hacker News

Testing Anthropic's RSI Claims

Hacker News - Sun, 08/09/2026 - 6:07pm

Article URL: https://ankitmaloo.com/rsi-claims/

Comments URL: https://news.ycombinator.com/item?id=49236616

Points: 1

# Comments: 0

Categories: Hacker News

Show HN: Standards – Opinionated Linter

Hacker News - Sun, 08/09/2026 - 4:40pm

I have given the same feedback over years of PR comments. I made a linting tool that flags feedback so you can on pre-commit get coding agents to adhere automatically ship compliant code.

Comments URL: https://news.ycombinator.com/item?id=49235624

Points: 1

# Comments: 0

Categories: Hacker News

Ask HN: Recommended Merchant of Record and Affliate Setup

Hacker News - Sun, 08/09/2026 - 4:36pm

Hello Everyone, long time reader, second time posting a thread. *edit because realized I did one other time.

I am nearing the completion of a product, and realized that usually when I am integrating payment it has been for other people who have figured out what providers they want.

I am wondering what people recommend from experience. I really want to be able to have an affiliate program set up so that I can offer 45% to people who sell, and would love that ability built in so that I have less paper work to deal with

However, I have been seeing a lot of horror stories around that the MoR I was planning to use (as it ticks every box) since acquisition, when I thought it would have increased in trust.

Anyway, I thought this is probably the most trustworthy community to check with for a wide range of experiences and honesty.

Comments URL: https://news.ycombinator.com/item?id=49235583

Points: 1

# Comments: 0

Categories: Hacker News

Pages