frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

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

https://github.com/JustVugg/lumabri
7•vforno•1h ago
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

nathan_compton•49m ago
Here is some feedback: don't put slop in your readme. It's unreadable.
drivingmenuts•21m ago
There needs to be some sort of safeguard to prevent people from abusing this and running queries to do something an owner objects to. I'd feel horrible if I discovered that someone used my AI to figure out how to kill someone. Another problem would be privacy. My highly personal queries would be available to whomever ran that query for me thru logging.

This is not to say it's a bad idea, but it needs some thought toward protecting users or at least acknowledgment and warnings that their information might be available to unintended eyes.

esafak•18m ago
LLMs are already slow on one machine. Now you are going to add network calls to it?
hazard•16m ago
I recommend reading some foundational materials on federated learning and distributed inference. The fundamental issues are (1) compute (2) latency and (3) security.

On the compute side, you can't load most individual experts into even high-end consumer-grade cards (e.g. RTX 5090) because the weights are simply too big.

On the latency side, you need to ship the activations and wait for the somewhat underpowered cards to actually do the matrix multiplications, then send the results back. This means each token takes hundreds of milliseconds or even more, which is borderline unusable.

Finally if you're running a P2P network, the only way to verify that peers aren't lying to you is to run duplicate calculations - I don't see how SHA-256 or signed model state help here unless I'm missing something (What are you calculating the hash of? What exactly is signing attesting to?).

This is one of the few places where borrowing ideas from cryptocurrencies actually makes sense - you could imagine a system where peers place bonds and forfeit them if they lie about calculations. You can look at projects like Bittensor and others for how these kind of things are currently implemented.

How I use LLMs to learn complex topics

https://laurentiugabriel.github.io/blog/articles/how-i-use-llms-to-learn/
318•laurentiurad•4h ago•180 comments

New Zealand lost its music media, and what we're building to replace it

https://propelmusic.co.nz/articles/the-sound-went-quiet-nz-music-media
54•berghoffer•3h ago•25 comments

"The Persian MâR-Nâmeh Or, the Book for Taking Omens from Snakes" (1892)

https://publicdomainreview.org/collection/marnameh/
10•Thevet•2d ago•0 comments

Mea Culpa – Dark Hours

https://blog.terrygodier.com/2026/08/09/mea-culpa-dark-hours.html
533•satvikpendem•10h ago•245 comments

Ask HN: What are you working on? (August 2026)

149•david927•6h ago•550 comments

The Hacker's Renaissance

https://phrack.org/issues/72/19#article
76•yu3zhou4•3h ago•51 comments

The tragedy of the commons, AI edition

https://www.economist.com/britain/2026/08/06/the-tragedy-of-the-commons-ai-edition
60•simonpure•4h ago•26 comments

Cool URIs Don't Change (1998)

https://www.w3.org/Provider/Style/URI
171•Klaster_1•9h ago•32 comments

Taxi drivers rarely die of Alzheimer's

https://theconversation.com/taxi-drivers-rarely-die-of-alzheimers-how-complex-mental-maps-and-spa...
158•jader201•8h ago•117 comments

Andrew Wiles on proving Fermat’s Last Theorem (1995) [video]

https://www.youtube.com/watch?v=GS7CxAtV5Ks
27•jackdoe•3d ago•14 comments

Tuxedo No. 2 – Cocktail recipes

https://tuxedono2.com
25•smartmic•3h ago•8 comments

Touring the Consensus, Six Months In

https://theconsensus.dev/blog/2026/08/05/touring-the-consensus-six-months-in.html
7•tosh•3d ago•0 comments

The climate benefits of retiring an operational combustion engine vehicle

https://www.science.org/doi/10.1126/science.adv5441
18•toomuchtodo•3h ago•20 comments

I made tinnitus my friend, then it disappeared [video]

https://mynoise.net/vlog.php?ep=20260803
50•gregsadetsky•5h ago•22 comments

OpenChamber: An Agentic Development Environment

https://openchamber.dev/
94•hexomancer•6h ago•53 comments

Everything you do is being recorded

https://www.theatlantic.com/technology/2026/05/ai-wearable-surveillance-countermeasures/687203/
182•ike_usawa•12h ago•139 comments

Georgia police officers fired after Flock camera misuse

https://www.wtoc.com/2026/08/07/savannah-police-department-fires-6-employees-over-flock-safety-sy...
48•chirau•2h ago•13 comments

How Golden Is Silence, Actually?

https://www.newyorker.com/magazine/2026/08/10/silence-kate-mcloughlin-book-review
29•tintinnabula•3d ago•16 comments

John C. Lilly on solid state intelligence and the elimination of man (1978)

https://kibotronics.net/unlisted/lilly-machines/
115•Kiboneu•10h ago•73 comments

Windows 11's built-in Weather app wastes more than 1 GB of RAM

https://www.notebookcheck.net/Windows-11-s-built-in-Weather-app-wastes-more-than-1-GB-of-RAM.1364...
329•akyuu•8h ago•272 comments

Show HN: A Project Oberon System version running on RISC-V instead of RISC-5

https://github.com/rochus-keller/OberonSystem/tree/op2-rv32
102•Rochus•11h ago•16 comments

Reviving a four year old reMarkable 2

https://oskrim.github.io/hardware/2026/08/09/remarkable-over-ssh.html
114•tremguy•12h ago•84 comments

Crickets as pets

https://en.wikipedia.org/wiki/Crickets_as_pets
75•Michelangelo11•4d ago•33 comments

When Compilers Disagree About UTF‑8

https://nemanjatrifunovic.substack.com/p/when-compilers-disagree-about-utf8
30•rbanffy•4d ago•5 comments

Show HN: Alphabet Soup, a multiplayer game, build the longest word to win

https://alphabetsoup.club
18•johnchinjew•3h ago•10 comments

There Are Magic Hexagons of Every Order

https://gukov.dev/math/2026/08/02/new-magic-hexagons.html
178•gukoff•16h ago•30 comments

The main way I've seen people turn ideologically crazy (2025)

https://blog.andymasley.com/p/the-main-way-ive-seen-people-turn
83•mef•3h ago•60 comments

To Recruit Teachers, School Districts Are Building Homes

https://www.nytimes.com/2026/07/28/business/affordable-housing-schools-teachers.html
17•mooreds•1h ago•6 comments

Human vs. AI – Diff-based line-level provenance for text under agentic editing

https://github.com/eighttrigrams/us-vs-them
42•eighttrigrams•8h ago•10 comments

The Alpha 21264 CPU: NT's Greatest RISC (1998)

https://halfhill.com/byte/1998-12_alpha.html
77•Lammy•13h ago•63 comments