frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Saving another 100TB of RAM with math (and Rust)

https://blog.cloudflare.com/saving-100-tb-of-ram-with-math/
1•f311a•1m ago•0 comments

North Korea's fake job interviews infected 30k devices

https://www.theregister.com/security/2026/09/18/north-koreas-fake-job-interviews-infected-30000-d...
1•rndsignals•1m ago•0 comments

An update to how we count public views across YouTube

https://support.google.com/youtube/thread/433409976/an-update-to-how-we-count-public-views-across...
1•haunter•2m ago•0 comments

25 years of contributing to Postgres with Peter Eisentraut

https://talkingpostgres.com/episodes/25-years-of-contributing-to-postgres-with-peter-eisentraut
1•clairegiordano•3m ago•0 comments

A Third of Americans Still Can't Name the Three Branches of Government

https://studyfinds.com/americans-cant-name-3-branches-government/
2•mdp2021•3m ago•0 comments

Fathom: All in one Jellyfin client with Seer and YouTube

https://github.com/Fathom-Media/fathom
1•traceapps•4m ago•0 comments

Anthropic's CEO says that safety hinges on understanding how AI "thinks."

https://www.wired.com/story/if-the-ai-industry-followed-its-own-research-it-might-have-paused-alr...
2•rndsignals•5m ago•0 comments

Who's a Better Writer: A.I. Or Humans? NYTimes Quiz

https://www.nytimes.com/interactive/2026/03/09/business/ai-writing-quiz.html
2•theopsimist•5m ago•1 comments

Show HN: In Meeting – A native macOS Mic/Cam observer for home automation

https://github.com/fellowgeek/in-meeting
1•pcbmaker20•7m ago•0 comments

Show HN: Repairo – Auto fixing breaking changes in your codebase

https://github.com/adityacs50-lab/Repairo
1•Adityasays•9m ago•0 comments

Incrementally Computing 7B Permission Checks

https://www.feldera.com/blog/auth0-and-feldera-incrementally-computing-7-billion-permission-checks
1•gz09•9m ago•0 comments

Apple releases iPhone Duo simulator and Xcode 27.1 beta

https://developer.apple.com/documentation/xcode-release-notes/xcode-27_1-release-notes
4•CameronBanga•13m ago•1 comments

The new AgentCore runtime: Elastic, optimized, and consistently fast starts

https://aws.amazon.com/blogs/machine-learning/the-new-agentcore-runtime-elastic-optimized-and-con...
1•qainsights•13m ago•0 comments

Love vs. Wiley Book Writers: A Publishing Scam Gets a Rare Comeuppance

https://writerbeware.blog/2026/09/18/love-v-wiley-book-writers-a-publishing-scam-gets-a-rare-come...
3•ilamont•14m ago•1 comments

Record 38% of new vehicle sales to be EVs in 2030, study finds

https://news.harvard.edu/gazette/story/2026/09/record-38-of-new-vehicle-sales-to-be-evs-in-2030-s...
2•geox•15m ago•0 comments

Putin Uses Unlicensed Windows

https://twitter.com/christogrozev/status/2100955969404408311
1•mudil•16m ago•1 comments

PC ports of old console games are the new AI vibe coding battleground

https://www.pcgamer.com/gaming-industry/pc-ports-of-old-console-games-are-the-new-ai-vibe-coding-...
1•haunter•17m ago•0 comments

Classified F-35 stealth parts diverted to Hong Kong by shipper, gone missing

https://www.politico.com/news/2026/09/18/congress-f-35-china-investigation-01083647
6•cf100clunk•20m ago•0 comments

Stratasys awarded 27.6M in damages in first of 2 cases against bambu lab

https://all3dp.com/6/stratasys-awarded-27-6-million-in-damages-in-first-of-two-cases-against-bamb...
3•iancmceachern•20m ago•0 comments

Have You Seen Shannon

https://news.ycombinator.com/submit
1•susanboop_09•21m ago•0 comments

AI Lab building tycoon game

https://bedroomlabs.net
1•o_g•21m ago•0 comments

An Undercover Google Analyst Infiltrated a Notorious Supply-Chain Hacking Gang

https://www.wired.com/story/an-undercover-google-analyst-infiltrated-a-notorious-supply-chain-hac...
2•MC995•22m ago•0 comments

Before you build it, share it

https://coidea.co
1•moeidheidari•22m ago•0 comments

I vibecoded a tool to shrink Golang codebase automatically

https://github.com/dhilst/sx
1•geckones•23m ago•1 comments

SSD-Llama: SSD-Native Inference for Trillion-Parameter Moe on a Consumer PC

https://arxiv.org/abs/2609.18110
1•AlmostCosmo79•24m ago•0 comments

ChatGPT's Built-in Browser now supports Chrome extensions

https://twitter.com/JamesZmSun/status/2100995281097769216
2•twapi•25m ago•0 comments

Keeping vLLM's Prefix Cache Warm Between Agent Turns

https://doug.sh/posts/vllm-kv-cache-agents/
1•dougcalobrisi•26m ago•0 comments

Near-real-time vision using DiffusionGemma in Jev mode

https://twitter.com/mmastrac/status/2101011110132601054
1•mmastrac•26m ago•0 comments

Anthropic finally adds AGENTS.md support to Claude Code

https://twitter.com/trq212/status/2101009392611278961
3•deaux•26m ago•1 comments

The Highway Tutorial

https://fastcode.substack.com/p/announcing-the-highway-tutorial
1•behoppe•27m ago•1 comments
Open in hackernews

Grok Voice Transcribe 2.0

https://x.ai/news/grok-voice-transcribe-2
16•vertigoruntime•59m ago

Comments

bigyabai•57m ago
No weights, sadly. Bit of a nothingburger release.
nomel•20m ago
> Word-level timestamps. Each word has precise start/end times and confidence scores.

Neat. This is missing from many models.

exizt88•16m ago
I don't understand how these benchmarks get things so wrong. Nova-3 is by far the best model on the market in WER. Azure's STT is atrocious. And yet every benchmark rates Azure's STT higher than Nova-3's.
mistic92•4m ago
Mai 2 transcribe is giving me better results than nova
mistic92•4m ago
If only they had ZDR