Hacker News

Top stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Claude Code now reads AGENTS.md if there is no Claude.md(claude.com ↗)
    125comments
  2. Android 17 is the first since 3.x to add new APIs without releasing to the AOSP(grapheneos.social ↗)
    176comments
  3. Saving another 100TB of RAM(cloudflare.com ↗)
    33comments
  4. Cloudflare Quick Tunnels(cloudflare.com ↗)
    218comments
  5. Xcode 27.1 Beta Release Notes(developer.apple.com ↗)
    57comments
  6. Cache-to-Cache: Direct Semantic Communication Between LLMs (2025)(arxiv.org ↗)
    11comments
  7. Photon-Emission-Guided Laser Fault Injection Enables RP2350 Secure Debug(ledger.com ↗)
    43comments
  8. Show HN: Cactus Needle 3: 8-29MB automation models can match DeepSeek V4 Flash(cactuscompute.com ↗)
    71comments
  9. How to Write with an LLM(sockpuppet.org ↗)
    237comments
  10. US troop deaths during Iran war exceed Pentagon count by at least four(reuters.com ↗)
    3comments
  11. OpenJev(openjev.com ↗)
    235comments
  12. The first new cat species discovered in 100 years(nationalgeographic.com ↗)
    34comments
  13. Cyclomatic Complexity in C#(ndepend.com ↗)
    8comments
  14. Two parallel neural ectoderm progenitors contribute to the developing brain(newscientist.com ↗)
    50comments
  15. The Implications of Linguistic Illegibility for LLM Security(arxiv.org ↗)
    16comments
  16. From Geometry to Algebra and Back Again: 4000 Years of Papers (2023) [video](youtube.com ↗)
    discuss
  17. C++26: Trivial infinite loops are no longer undefined behaviour(sandordargo.com ↗)
    166comments
  18. Size-Specialized Memory Allocation(go.dev ↗)
    3comments
  19. Minimal Phone 2(minimalcompany.com ↗)
    141comments
  20. How SpaceX streamlined the Raptor engine(construction-physics.com ↗)
    23comments
  21. Warez: The Infrastructure and Aesthetics of Piracy (2021)(archive.org ↗)
    12comments
  22. I vibed a proof of Conway's conjecture(overreacted.io ↗)
    174comments
  23. Inside ZCode: Silently uploading your Git history to the cloud(ferstar.org ↗)
    89comments
  24. A search-and-inference database from scratch in pure Zig(antfly.io ↗)
    16comments
  25. Korea raises data breach fines to 10% of revenue(koreajoongangdaily.com ↗)
    70comments
  26. Cekura (YC F24) Is Hiring(ycombinator.com ↗)
    discuss
  27. Border agents can search cellphones without a warrant or reasonable suspicion(lawandcrime.com ↗)
    130comments
  28. US Military had close call after using AI for hallucinated intelligence report(cnn.com ↗)
    275comments
  29. Mathematicians Build Long-Awaited Graph Sandwich(quantamagazine.org ↗)
    15comments
  30. North Korean nuclear test sets off years of earthquakes(science.org ↗)
    147comments

Show HN: Burn, baby, burn (those tokens)

135 pointsby 4mo agogithub.com
30 comments
4mo agoHN ↗

Creator here.

Are you a startup looking to show investors your hefty AI budget? Are you an engineer trying to top the AI token leaderboard? Wanna show your friends how "AI Native" you are? Look no further, I created this for you.

4mo agoHN ↗

If the aim is to burn tokens in a plausible manner why not go one step further, without getting users accounts banned, use those tokens as a distributed distillation effort.

Publish all the generated content into a data source, think open crawl, so open sourced models (open training not just weights) could use it as input.

4mo agoHN ↗

you laugh but this is so real. executives love metrics. I do feel like I need to keep the prompts going.

4mo agoHN ↗

Token-maxxing is a silly idea that does not measure true productivity and quality. I can easily burn through tokens and not produce anything useful.

4mo agoHN ↗

Token-maxxing is a silly idea

Well I guess you also thing that getting a job promotion is a silly idea? /s

4mo agoHN ↗

Never would I ever have imagined a world where companies are going all in on the idea of measuring productivity by how much an employee is able to spend company money. Try to explain this to a time traveler from 2020. What a time to be alive.

4mo agoHN ↗

Absolutely incredible, gonna kee this on my back burner for when my boss inevitably gets the tokenmaxxing mind virus.

4mo agoHN ↗

This idea is really innovative and interesting.

4mo agoHN ↗

Marketing material includes legendary promises such as:

Burning token like there's no context window.

4mo agoHN ↗

It’s a strange time ,some people are trying to save tokens to reduce costs, while in other places, employees are intentionally burning more tokens for performance evaluations.

4mo agoHN ↗

Should be making pelicans in hats riding unicorns.

4mo agoHN ↗

Ok so in all honesty I have it a try. I think you are burning "input tokens" rather than output ones, which fill your usage quicker.

Suggestion to improve? :praying_hands:

4mo agoHN ↗

I can do this with KiloCode and actually generate stuff. I burned 680 M tokens last weekend building over a 48 hr period on KiloCode with Grok Code First(free until 15th). Once I uploaded into replit, there was a lot of stubs lets just say and truncated code and things to clean. But it built something and did a hefty token burn in the process.

4mo agoHN ↗

This is one of the funniest things I've seen on HN. This is actually becoming useful where I work haha :) I think your next natural move should be to make a who-burned-it-baby version for management to audit and figure out who's been burning tokens and ironically use a ton of tokens to do that :D

4mo agoHN ↗

Like I just said in another thread, all this reminds me so much of that scene in Mr. Robot where they make a CIO set millions of dollars on fire.

Sometimes I feel like the leadership doesn't even know what they're at. They just want to show "We're an AI company", whatever we do with that AI is not relevant. It's frankly ridiculous.

4mo agoHN ↗

Finally, a proper way to turn company money into visible "productivity"...

4mo agoHN ↗

If anyone out there needs to burn some tokens, just DM me your keys. I got you.