frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Democrats fear more 'October surprises' after 'bullshit' raid

https://news.lavx.hu/article/democrats-fear-more-october-surprises-after-bullshit-raid-on-ilhan-o...
1•abdelhousni•1m ago•0 comments

Custom server to revive neato bots

https://github.com/vacuula/vacuula
1•matttah•2m ago•0 comments

Nvidia reportedly halts GeForce RTX 5090 production

https://www.tomshardware.com/pc-components/gpus/nvidia-reportedly-halts-geforce-rtx-5090-producti...
1•kristianp•6m ago•0 comments

A new model for open source: closed to pull requests, open to remix

https://ghuntley.com/remix/
1•ghuntley•10m ago•0 comments

Sunsetting Decomp.dev

https://decomp.dev/
1•Encounter•10m ago•0 comments

Against Big Tech Credentialism

https://alexwennerberg.com/blog/2026-04-16-credentials.html
1•nsm•11m ago•0 comments

Chaos ensues when AI interviewer meets AI interviewee

https://www.theregister.com/offbeat/2026/10/14/chaos-ensues-when-ai-interviewer-meets-ai-intervie...
2•mikelgan•18m ago•0 comments

Ask HN: Demonstrating Skills When Everything Is a Prompt

2•Brisk4t•18m ago•0 comments

Ben Affleck Sold His AI Company for $587M [video]

https://www.youtube.com/watch?v=MzXEWrT3Ld8
1•dboreham•19m ago•0 comments

Bank of America added 1k contractors to patch AI-exposed software flaws

https://runtimewire.com/article/bank-of-america-mythos-contractors-software-patching
2•runtimewire•25m ago•1 comments

Epic Games Launcher Preview

https://launcher.epicgames.com/
1•runtimewire•26m ago•0 comments

Local Model Guide – Find what LLMs fit on your machine

https://localmodelguide.com/
1•hiveindex•26m ago•0 comments

Netflix lays off 6% of staff in biggest job cuts since 2022

https://finance.yahoo.com/video/netflix-lays-off-6-of-staff-in-biggest-job-cuts-since-2022-134352...
3•BiraIgnacio•27m ago•0 comments

The genetic technique that finds almost every infection

https://web.archive.org/web/20261014232102/https://www.bbc.com/future/article/20261013-how-metage...
1•BiraIgnacio•27m ago•1 comments

1991 Federal Panel Blasted Vaccine Unavailability That Led to a Measles Epidemic

https://www.techdirt.com/2026/10/13/in-1991-federal-panel-blasts-vaccine-unavailability-that-led-...
1•hn_acker•27m ago•1 comments

TezGrid: AI inference on idle computers

https://tezgrid.com/
1•ms7892•29m ago•0 comments

GGUF: GGML Universal File

https://en.wikipedia.org/wiki/GGUF
2•teleforce•29m ago•0 comments

Frontier models score under 2/100 controlling a simulated humanoid

https://www.vals.ai/benchmarks/robo_bench
6•miyukimatthews•29m ago•0 comments

Anthropic Revenue Jumps to $16.8B in Quarter Before IPO

https://www.bloomberg.com/news/articles/2026-10-14/anthropic-revenue-jumps-to-16-8-billion-in-qua...
3•marc__1•33m ago•1 comments

Lunchables of UI Work

https://blog.exe.dev/implementation-fidelity
1•azimman•34m ago•0 comments

Backlash against data centres could shape the US midterm elections

https://www.bbc.com/news/articles/c0rezqpnekro
1•dabinat•37m ago•2 comments

RustGIS

https://rustgis.app/
2•jonbaer•37m ago•0 comments

Ablation is an autonomous reverse engineering system

https://github.com/Ablation-Tool/ablation
1•zellkernel•40m ago•0 comments

Things we wish we'd known before building an AI chat agent

https://trigger.dev/blog/seven-tips-for-building-better-chat-agents
1•shikhar•40m ago•0 comments

DHS Inspector General: TSA Is Not Safeguarding Passenger Facial Recognition Data

https://www.eff.org/deeplinks/2026/10/dhs-inspector-general-finds-tsa-not-properly-safeguarding-p...
5•hn_acker•41m ago•1 comments

Osint Report on Hybrid Warfare by Russia – Entelligencia

https://entelligencia.report/weekly/04/
2•entelligencia7•41m ago•0 comments

GLP-1 Drugs Cut Schizophrenia Hospital Visits, Study Finds

https://www.bloomberg.com/news/articles/2026-10-14/glp-1-drugs-cut-schizophrenia-hospital-visits-...
4•toomuchtodo•42m ago•1 comments

ICE Wants to Use AI to Monitor Detainee Calls and Movements

https://www.washingtonsun.com/immigration/ice-ai-monitor-detainees
9•hn_acker•44m ago•0 comments

What If AI Is the End of the World?

https://www.theatlantic.com/ideas/2026/10/ai-end-of-world-meaning/688977/
4•7777777phil•46m ago•0 comments

"Addictive Features" Social Media Restrictions on Minors Likely Constitutional

https://reason.com/volokh/2026/10/14/florida-addictive-features-social-media-restrictions-on-mino...
2•delichon•47m ago•0 comments
Open in hackernews

A friend of mine built an agent-first group chat/social network

https://lamoom.com/
3•mikemilla•50m ago

Comments

mikemilla•50m ago
My friend Kate built an interesting social app where you connect Claude Code, then let it go message other agents and respond when they message you.

I went in expecting it to be awkward, but it turned out to be fun!

You give your agent a little objective (I wanted mine to meet other product designers), and it goes off and socializes for you, haha. It read my LinkedIn, so it knows a little about me and what to answer when other agents ask. The agents also knew when to stop messaging, so they don’t get stuck in some infinite loop burning tokens.

It’s definitely some kind of social experiment. Kate says about 300 people in SF are on it so far. We created a group chat in Lamoom for the WeWork we’re in, and quite a few of us are in there now. It’s pretty cool to see our agents messaging each other, asking questions and stuff.

https://lamoom.com if you’re interested

gateauxd•45m ago
We use Lamoon for our agent's social media.