Hacker News

New stories

Live mirror
30 storiesupdated just nowView source snapshot
  1. Europe Must Choose Between Power and Dependence(slavoj.substack.com)
    discuss
  2. Mathematics in the Age of AI Oracles [pdf](colala.berkeley.edu)
    discuss
  3. CVE-2026-91766: PHP had the redirect credential leak curl fixed in 2018(daubois.dev)
    1comments
  4. LinkedIn wins court order blocking mass scraping of user data(therecord.media)
    discuss
  5. Megafauna Extinction and Startups(johnjwang.com)
    discuss
  6. Discovering and exploiting a remote code execution vulnerability in OpenCode(datadoghq.com)
    discuss
  7. Chicago No. 1(dynamicallytyped.org)
    discuss
  8. Oracle invokes force majeure on New Mexico AI data center(qz.com)
    discuss
  9. Making the case for Ed Zitron as the last of the great bloggers(observationalepidemiology.blogspot.com)
    discuss
  10. I Have a Confession: I Built This Site with AI – Please Forgive Me(dynamicallytyped.org)
    discuss
  11. Show HN: Timebar, a Mac timer shown as a thin line below the menu bar(github.com/velvet-shark)
    1comments
  12. Pixel Composer: Node based VFX compositor for pixel art(pixel-composer.com)
    discuss
  13. The Medicaid Work Requirement's Gambling Loophole(aei.org)
    discuss
  14. Null, a Skill and a Blog Post(probablyfine.dev)
    1comments
  15. Nikon Made a $1,400 Full-Frame Camera for Gen Z and I Might Be Too Old for It(pcmag.com)
    discuss
  16. The Geography of Bear Attacks [video](youtube.com)
    discuss
  17. QR Codes That Route to the Appropriate App Store(matthuggins.com)
    discuss
  18. Why Influence Is Deeply Human(unchartedpathbreakthroughs.com)
    discuss
  19. GitHub has not removed malicious imitation software after 3 weeks(successfulsoftware.net)
    discuss
  20. Nine countries without internet that never disconnected(observedstate.com)
    discuss
  21. The United States of Walmart, According to WiFi Data(ipinfo.io)
    1comments
  22. Show HN: Scalar.dev – Python/SQLite server-rendered starter app, no build step(getscalar.dev)
    discuss
  23. My obsession making native resolution PS2 emulation look good(themaister.net)
    discuss
  24. Show HN: Trify3D – Compare AI 3D engines on one input and keep the best mesh(trify3d.com)
    discuss
  25. AI Evals: Everything You Need to Know(hamel.dev)
    discuss
  26. Flip Fluid on Flip Dots(mitxela.com)
    discuss
  27. Nephew Tommy prank calls compilation and transcripts(railway.app)
    discuss
  28. Why is the human body so crap except for the liver?(dynomight.net)
    discuss
  29. What I learned from joining the Kubernetes Blog team(kubernetes.dev)
    discuss
  30. Show HN: Potluck – Run local AI across the computers you own(trypotluck.ai)
    discuss

Ask HN: How do you read, sign off, verify, manage the volume of generated slop?

1 pointsby 1h ago
5 comments
We gotta move fast and we are seeing the generated slop day by day. If you are lucky and don't see it, then please don't participate in this conversation. We don't need to be belittled down. We know where that path leads too.

However, for those who have to read 2000 lines of code every single hour, (2000 is exaggeration. But at the end of the PR, it adds up to be 2000.)

1. how are you managing the mental bandwidth?

2. How are you verifying these code and signing them off?

3. Any tips, tricks, ideas that will keep us relevant here?

The biggest concern is understanding the maths and research code behind it. The shitty abstraction and duplication just puts off reading the code.

Regarding context collapse, the model doesn't have any idea of across different sessions. The token inefficiency is off the charts.

4. The problem is that what are we going to do about this token inefficiency? I find myself rewriting same instruction because the fix from last patch broke the system so this patch it needs to write a new fix, then proceed with feature implementation. The same command, function, calls, doesn't work across two session.

5. TDD is not going to help when half of the tests are meaningless?

tldr; Problem is not that we can code by hand, problem is there is no way to code this sheer volume (feature or design, sometimes entire product) by hand in the amount of time that we have. That is the problem.

1h agoHN ↗

why are you even attempting to read 2000 loc per hour? that won't work.

if ai generated the code, ai can review it. I can review perhaps pieces that are deemed priority. manual and automated testing are both required.

for the rest, please number your Qs so they're easy to address.

50m agoHN ↗

But how are you scoping the features these days? Unlike back in the day, these days a feature is equivalent of research paper

26m agoHN ↗

if you prompt differently, you can instruct them to work on small things

they have a strong tendency to over engineer (volume), prompt against this - in permanent files like AGENTS.md or per-session in follow on messages

1h agoHN ↗

However, for those who have to read 2000 lines of code every single hour, how are you managing the mental bandwidth?

Do you really think that's what's happening?

28m agoHN ↗

staying in the loop and being methodical about it (at least trying)

I more often find the walls of markdown to be more onerous than the code changes