Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Human brain is two separate organs, Stanford Medicine-led research finds(stanford.edu ↗)
    8comments
  2. San Francisco Onion Futures Company(onionfutures.com ↗)
    47comments
  3. Android 17 is the first since 3.x to add new APIs without releasing to the AOSP(grapheneos.social ↗)
    361comments
  4. Typesafe-computer-use drives a Mac toward a goal for 1/50th of a cent per step(github.com/awlevin ↗)
    13comments
  5. SDCC – Small Device C Compiler(sourceforge.net ↗)
    14comments
  6. Science Is Open Software(jepedersen.dk ↗)
    25comments
  7. If math is more than proof, we need to better celebrate the rest of it(terrytao.wordpress.com ↗)
    discuss
  8. Cloudflare Quick Tunnels(cloudflare.com ↗)
    272comments
  9. Why building a Rust LSP is hard(rust-glancer.github.io ↗)
    21comments
  10. Saving another 100TB of RAM(cloudflare.com ↗)
    61comments
  11. How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip(ieee.org ↗)
    76comments
  12. How to Write with an LLM(sockpuppet.org ↗)
    312comments
  13. NASA-IBM Lunar Foundation open-Source Geospatial AI Model(usra.edu ↗)
    1comments
  14. Ctenophores: Wonders of Biology(quantamagazine.org ↗)
    4comments
  15. You can run Git on object storage if you re-make packfiles(tigrisdata.com ↗)
    4comments
  16. Goroutine Leak Profiles(go.dev ↗)
    2comments
  17. The first new cat species discovered in 100 years(nationalgeographic.com ↗)
    90comments
  18. Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash(cactuscompute.com ↗)
    85comments
  19. OpenJev(openjev.com ↗)
    257comments
  20. Photon-Emission-Guided Laser Fault Injection Enables RP2350 Secure Debug(ledger.com ↗)
    66comments
  21. The Farnese letter(simonklee.dk ↗)
    6comments
  22. Cache-to-Cache: Direct Semantic Communication Between LLMs (2025)(arxiv.org ↗)
    12comments
  23. Xcode 27.1 Beta Release Notes(developer.apple.com ↗)
    95comments
  24. Minimal Phone 2(minimalcompany.com ↗)
    208comments
  25. Cyclomatic Complexity in C#(ndepend.com ↗)
    17comments
  26. Inside ZCode: Silently uploading your Git history to the cloud(ferstar.org ↗)
    97comments
  27. Claude Code now reads AGENTS.md if there is no Claude.md(claude.com ↗)
    219comments
  28. Warez: The Infrastructure and Aesthetics of Piracy (2021)(archive.org ↗)
    46comments
  29. Alibaba open-sources AI model that can detect cancer and nearly 150 conditions(scmp.com ↗)
    11comments
  30. LispBM is a concurrent Lisp for microcontrollers with message passing(lispbm.com ↗)
    3comments

Typesafe-computer-use drives a Mac toward a goal for 1/50th of a cent per step

52 pointsby 2d agogithub.com
13 comments
1h agoHN ↗

Super cool. Hope waitlist will move soon. I have a use case for it too.

are you the author? If so - what are your notes on using Jev in this scenario?

1h agoHN ↗

Did you do the follow up questions? I was invited within a few hours of joining today.

It's also now on openrouter and cloudflare

1h agoHN ↗

Curious about people’s experience here. I am working on a small model, verify by jev, and escalate to big model. Some cases, the small model is not a model but some regex.

cheap-confirm-escalate

Using jev as the confirm step.

1h agoHN ↗

How does it do on OSWorld-verified? Recently read that even Fable 5 is just at 85% .

32m agoHN ↗

This probably throws a spanner in the wheels there:

Every piece of reasoning the frontier model does for free has to be rebuilt here as deterministic state.

EDIT: Not to shit on this though. I totally believe that some smart mixture of LLM-reasoning + Jev-style + determinism is going to be pretty amazing.

1h agoHN ↗

I'd be interested to see if using DiffusionGemma-as-Jev helps as you can feed the image directly into the model and it'll make decisions based on the image embeddings.

59m agoHN ↗

I had a long talk with chat gpt about this today as well. I think its duable and prolly not too hard either, also you could do lotsa funky stuff with stitched frames of a video in one 4x4 grid for example and send that as one image for analysis. that way temporal understanding can be had for fractions of a second by jev... also because vlm works in pixel space you can get around the whole state machine issue as well, so many possibilities...

50m agoHN ↗

I trailed off a few lines into the README. No human ever edited any of this. « LLM detected, project rejected ».

49m agoHN ↗

I stopped reading at “The honest caveat”…

20m agoHN ↗

Every third post in HN has this same complaint comment. Everybody knows.

10m agoHN ↗

And every comment on HN calling out vibe-coded slop has this same complaint comment in turn. OP has an actual complaint, your complaint is just "stop complaining".

4m agoHN ↗

I appreciate it when people say something is slop. Saves me from wasting my time looking at it.

How many slop ideas have come to the front page, never to be heard from again because the execution isn’t actually any good? I’m guessing most of them.