frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

What If AI Is the End of the World?

https://www.theatlantic.com/ideas/2026/10/ai-end-of-world-meaning/688977/
1•7777777phil•2m 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...
1•delichon•2m ago•0 comments

Morph Map – Visualise your codebase, sketch a refactor, and export the prompt

https://github.com/dannynatha-max/morph-map
1•dannynatha•3m ago•0 comments

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

https://lamoom.com/
3•mikemilla•5m ago•2 comments

Under the hood, off the hot path: How D1 scales to databases

https://blog.cloudflare.com/d1-control-plane/
1•Icathian•14m ago•0 comments

Rustuna: A faster Optuna implementation in Rust

https://rustuna.readthedocs.io/en/latest/
1•jonbaer•15m ago•0 comments

Show HN: Lection – describe the data you want, get a table from any web page

https://www.lection.app
1•jlauf•16m ago•0 comments

Google's Fitbit Edge Has Health Features and a Replaceable Battery

https://www.engadget.com/2283307/googles-sleek-new-fitbit-edge-is-packed-with-health-features-and...
2•gmays•18m ago•0 comments

DuoDiscover: Scaling Autonomous Discovery Beyond the Evaluation Bottleneck

https://arxiv.org/abs/2610.17184
1•matt_d•18m ago•0 comments

AB Fab – On Anthony Burgess

https://www.lrb.co.uk/the-paper/v48/n18/ian-penman/ab-fab
2•feltsense•20m ago•0 comments

How to be more ambitious: A tutorial for humans with models

https://felixrieseberg.com/how-to-be-more-ambitious-a-tutorial-for-humans-with-models/
1•ot•22m ago•0 comments

Adobe Software Engineer Interview: Process, AI Rules and How to Prepare

https://nokku.payanai.com/interview-prep/adobe
2•payanai•22m ago•0 comments

Not Mutual but Assured

https://monroelab.com/2025/01/01/not-mutual-but-assured/
3•CGMthrowaway•22m ago•0 comments

Show HN: Ask a technical question, get an animated explainer you can interrupt

https://openlume.com/examples/3f65f6ce-87bf-4c81-a34b-3a1ec82f0ed1
1•shadaba•25m ago•0 comments

Helix World: A Real-Time Audio-Visual World Model

https://noiz.ai/lp/blog/meet-helix-world
1•gmays•27m ago•0 comments

Problems with the Rust Language

https://oxetene.net/problems-with-the-rust-language/
6•embedding-shape•28m ago•0 comments

Anthropic revenue jumps to $16.8B in third quarter

https://www.ft.com/content/11632d03-358c-454b-af95-d06d4072932b
7•orf•28m ago•1 comments

Benchmarking decision models for relevance judgments

https://iwhalen.com/relevance-bench/
1•iwhalen•30m ago•0 comments

What's a Nonius?

https://hackaday.com/2026/10/14/whats-a-nonius/
1•alberto-m•32m ago•0 comments

Strive Targets $500M Raise to Buy More Bitcoin

https://coinmarketcap.com/community/post/380133331/
1•joeymabia1•32m ago•1 comments

Show HN: Fantastic Canvas

https://github.com/aisixteen/fantastic
2•Alexadar•32m ago•2 comments

New York City Agrees to Profit-Sharing Deal with NYU Quantum Lab

https://www.nytimes.com/2026/10/14/nyregion/mamdani-nyu-quantum-partnership.html
1•appleappleapple•33m ago•0 comments

Nvidia's China AI-chip share falls to 34% as Huawei reaches 28%

https://technode.com/2026/10/14/nvidias-china-ai-chip-share-falls-to-34-as-huawei-reaches-28/
2•yogthos•35m ago•0 comments

Inference vs. Distillation

https://debarshibasak.github.io/readables/blogs/inference-vs-distillation
2•debarshri•35m ago•0 comments

Coding with AI stopped feeling like craft, so I built a tiny desk radio

https://radiowaveshare.vercel.app
1•mayilian•36m ago•0 comments

ESA targets Anbernic emulation handhelds as "superfake" piracy enablers

https://www.osnews.com/story/146148/esa-targets-anbernic-emulation-handhelds-as-superfake-piracy-...
1•iamnothere•37m ago•0 comments

Show HN: Walk or your virtual pet gets fat

https://apps.apple.com/au/app/terry-pedometer-step-counter/id6818785615
1•VatanaChhorn•39m ago•0 comments

Atelier – Watch AI Learn to Paint (Sometimes, Live)

https://atelier.randomdrake.com
1•randomdrake•39m ago•0 comments

ZenOS–A clean, minimal E-Reader experience

https://github.com/xZenLabs/zen-os
1•Cider9986•39m ago•0 comments

What MCP changes about enterprise software

https://www.deepl.com/en/blog/what-mcp-changes-about-enterprise-software
1•mooreds•40m ago•0 comments
Open in hackernews

8.8M records hacked from Danish national registry with password 123456

https://www.tomshardware.com/tech-industry/cyber-security/hacker-claims-123456-password-let-them-breach-the-danish-national-registry-accessing-8-8-million-records-it-and-data-services-company-with-weak-passwords-linked-source-of-the-hack
3•jameslk•57m ago

Comments

iamnothere•34m ago
Any time people question why government (or companies) shouldn’t be trusted with massive centralized databases of PII, point them to this.

Policies can only do so much. If you set a policy, there’s no way to ensure that future users/admins won’t allow violations for convenience, and yes this holds even if you have criminal penalties.

I hate to say it, but the “best practice” in an increasingly hostile environment may look something like Germany’s byzantine systems that don’t communicate directly with one another.