frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Ask HN: How are you using LLMs for coding?

2•mraza007•1y ago
What have been your tips and tricks when coding with LLMs

Comments

benoau•1y ago
Whenever it gets in a cycle of fucking up, I ask it to break the code down into smaller functions and write a test suite for each individually.

Tell it what language or packages to use or it might make selections that add dependencies or require installing stuff on your computer.

Tell it how you want your code written or it will be an extra chore to accommodate linting requirements.

Basically just being very explicit.

GianFabien•1y ago
I mostly write low-level code in Python and JS. When working with a new API, I would sometimes find it difficult to locate information to fix problems. With ChatGPT I simply rubber duck my problem. It usually only takes a couple of iterations to zoom into the core solution.
Flundstrom2•1y ago
Copilot in visual studio to fix compiler errors and let it type all the boilerplate code by TAB-ing. Sometimes it even manages to spit out an entire - and correct - function.

Mistral Le Chat for more advanced questions and figuring out things that copilot can't.

Show HN: CiteLadder – open-source AI search visibility tracker

https://github.com/abhij1306/CiteLadder
1•abhij1306•1m ago•0 comments

If it hurts, do it more

https://jnsgr.uk/2026/09/if-it-hurts-do-it-more
1•Momo_isk•1m ago•0 comments

Chinese phones with built-in malware sold in Africa

https://www.bbc.com/news/technology-53903436
1•sehw•2m ago•0 comments

Aviation superfan installs an Airbus A380 suite in his apartment

https://www.cnn.com/travel/aviation-superfan-a380-etihad-residence-suite
1•Geekette•3m ago•0 comments

Ask HN: Local AI product situation

1•ankurdhama•4m ago•0 comments

Neuralwatt says cache misses can use up to 140% more inference energy

https://runtimewire.com/article/neuralwatt-harness-cache-energy-chad-gibson
1•runtimewire•4m ago•0 comments

People Were Always Shitty

https://notoriousbfg.com/people-were-always-shitty/
1•trwhite•7m ago•0 comments

Show HN: I turned my kids' old Yoto Player into an Audiobookshelf player

https://github.com/tomekent/yotwo
1•beret4breakfast•8m ago•0 comments

Szczecin Paprikash

https://en.wikipedia.org/wiki/Szczecin_paprikash
1•jxub•10m ago•0 comments

DeepSeek-v4.1-Flash-1017 is here

https://twitter.com/deepseek_ai/status/2111375640717033777
4•theanonymousone•11m ago•0 comments

DeepSeek: DeepSeek v4.1 Flash 1017

https://openrouter.ai/deepseek/deepseek-v4.1-flash-1017
1•theanonymousone•13m ago•0 comments

DeepSeek users report a possible v4.1-Flash-1017 refresh across its API

https://runtimewire.com/article/deepseek-users-report-a-possible-v4-1-flash-1017-refresh-across-i...
1•runtimewire•15m ago•0 comments

Galois Finite Fields and the Advanced Encryption Standard (AES)

https://www.cs.uaf.edu/2015/spring/cs463/lecture/03_23_AES.html
1•ankitg12•16m ago•0 comments

Formally verified AES-XTS: The first AES algorithm to join s2n-bignum

https://www.amazon.science/blog/formally-verified-aes-xts-the-first-aes-algorithm-to-join-s2n-bignum
1•ankitg12•19m ago•0 comments

Show HN: CADrs – onshape rust clone

https://github.com/rvdende/cadrs
1•rouanvde•22m ago•1 comments

Geographical Fun: Being Humourous Outlines of Various Countries (1868)

https://publicdomainreview.org/collection/geographical-fun-being-humourous-outlines-of-various-co...
1•Thevet•26m ago•0 comments

DeepSeek-v4.1-Flash-1017

https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash-1017
13•theykk•29m ago•4 comments

Show HN: Sys2txt – transcribe system audio to text in the CLI with a local LLM

https://Joe-Heffer.github.io/sys2txt/
2•joeheffer•36m ago•0 comments

Meta to scrape content from business customers for model training

https://www.theregister.com/ai-and-ml/2026/10/16/meta-grants-itself-right-to-scrape-content-from-...
1•nephihaha•36m ago•0 comments

PDFCube: Present your PDF slides on a spinning 3D cube

https://github.com/baol/pdfcube
1•baol•39m ago•0 comments

Pooling OpenAI subscriptions and bypassing regional inference capacity limits

https://ghuntley.com/underclass/
1•ghuntley•41m ago•0 comments

Hallucination as Model Defense: On the Nature of Artificial Intelligence Errors

https://medium.com/@smoking2man/hallucination-as-model-defense-on-the-nature-of-artificial-intell...
1•Aldus551•46m ago•0 comments

MapQuest will now show you where Flock cameras are in your neighborhood

https://bsky.app/profile/mapquest.com/post/3mxx5hbo2rc23
1•DeepLogin•49m ago•0 comments

Show HN: Agentfork – Replay an LLM agent from any step and test fixes

https://github.com/bimal1023/agentfork
1•bkumal•57m ago•0 comments

Build or Surrender

https://twitter.com/arthurmensch/status/2110990492930314712
1•MrBuddyCasino•59m ago•0 comments

TinyFFR – A free realtime 3D rendering library for modern .NET

https://tinyffr.dev/
1•bj-rn•1h ago•0 comments

Palantir patient data tech rejected by many NHS trusts in England

https://www.theguardian.com/technology/2026/oct/17/palantir-patient-data-tech-rejected-by-many-nh...
3•chrisjj•1h ago•0 comments

Researchers Stranded an AI Agent on a Virtual Island

https://gizmodo.com/researchers-stranded-an-ai-agent-on-a-virtual-island-things-got-very-weird-20...
2•jonbaer•1h ago•0 comments

Show HN: FourCities – A multiplayer transport tycoon game in the browser

https://fourcities.app
1•emirhanozsoy•1h ago•0 comments

The Leap Second Is No More

https://www.wired.com/story/leap-second-is-abolished-bipm-says/
1•beardyw•1h ago•0 comments