frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Animation 10k Starlink Satellites

https://spaceweather.com/archive.php?view=1&day=18&month=03&year=2026
1•MeteorMarc•47s ago•0 comments

Catastrophe and Corporate Bonds: A Unified Approach to Pricing Disaster Risks

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=6165466
1•7777777phil•52s ago•0 comments

Show HN: Blocktools – a Rust-powered suite for the smart contract lifecycle

https://blocktools.dev
1•wazzaaaa•1m ago•0 comments

Show HN: QCCBot – Android in a browser tab, with AI agent control

1•Eastra•1m ago•0 comments

Show HN: ROMA runs multiple coding agents simultaneously – Claude, Codex, etc.

https://github.com/liliang-cn/roma
1•AISlop31415•3m ago•0 comments

Confer is bringing foundational AI privacy to Meta

https://confer.to/blog/2026/03/encrypted-meta/
1•jbegley•5m ago•0 comments

Echoo – Free, open-source macOS app for system-wide AI via shortcuts and voice

https://www.echoo.ai/
1•mike_elkabetz•6m ago•2 comments

Run any LLM on any hardware. Auto-detects your GPU, checks if the model fits

https://github.com/Julienbase/uniinfer
1•julien_base31•8m ago•0 comments

AI changed software cost structure not Value structure

https://testdouble.com/insights/quality-you-cant-generate-ai-is-only-as-good-as-your-constraints
1•kathir05•13m ago•0 comments

Arizona Charges Kalshi with Illegal Gambling Operation

https://www.bloomberg.com/news/articles/2026-03-17/arizona-charges-kalshi-with-operating-illegal-...
1•greg_w•20m ago•0 comments

Show HN: Open-source YouTube summary, transcript chat, and timeline sidepanel

https://github.com/EchoTide/QuickSummarize
1•oujiangping•24m ago•0 comments

British physics faces 'catastrophic' cuts

https://www.bbc.com/news/articles/czr0zmzzp84o
1•jb1991•24m ago•0 comments

Show HN: Vue-OTP-pro – Minimal OTP input for Vue 3

https://vue-otp-pro.netlify.app
1•pratik227•27m ago•0 comments

Celebrating Tony Hoare's mark on computer science

https://bertrandmeyer.com/2026/03/16/celebrating-tony-hoares-mark-on-computer-science/
1•benhoyt•32m ago•0 comments

The Hyperscale IPv4 Moat: Analyzing AWS's Latest 9M Address Acquisition

https://circleid.com/posts/the-hyperscale-ipv4-moat-analyzing-awss-latest-9m-address-acquisition
1•jruohonen•33m ago•0 comments

The Last Quiet Thing

https://www.terrygodier.com/the-last-quiet-thing
4•cylo•36m ago•0 comments

CVE-2025-71263: Unix fourth edition buffer overflow vulnerability

https://www.cve.org/CVERecord?id=CVE-2025-71263
2•pseudohadamard•37m ago•1 comments

The End of Political Hypocrisy

https://hedgehogreview.com/web-features/thr/posts/the-end-of-political-hypocrisy
2•rzk•38m ago•1 comments

Show HN: I wrote a macOS C++ audio driver to fix HDMI volume controls

https://github.com/chenjy16/SoundBridge
2•chenjy16•39m ago•0 comments

Contract-Centered Iterative Stability v4.7.3

https://ai.gtzilla.com/papers/contract-centered-iterative-stability-v4.7.3/
1•osconfused•41m ago•1 comments

DeepSeek by Hand in Excel

https://www.byhand.ai/p/deepseek-lecture-128
4•teleforce•42m ago•0 comments

Thrum – Agent coordination through messaging

https://leonletto.github.io/thrum/index.html
2•leonletto•42m ago•1 comments

How Israel is censoring reporting on the war

https://www.972mag.com/israel-media-censorship-iran-war/
6•0x54MUR41•44m ago•0 comments

100% private Mac dictation app

https://www.getvoibe.com/
1•ayushchat•46m ago•0 comments

Designing Defensive Databases for Agents

https://arpitbhayani.me/blogs/defensive-databases/
1•Noumenon72•46m ago•0 comments

LightningPDF – HTML to PDF API with a native renderer and Chromium fallback

https://lightningpdf.dev/
2•DigitalTrubador•49m ago•2 comments

Apache Fory C++: 10x faster serialization than protobuf

https://fory.apache.org/blog/fory_cpp_blazing_fast_serialization_framework/
3•chaokunyang•52m ago•1 comments

Solow's 1987 Computer Paradox Explains Today's AI Measurement Crisis

https://theaugmentedinvestigator.substack.com/p/echoes-of-1987-how-solows-computer
1•walterbell•53m ago•0 comments

Snap-confine and systemd-tmpfiles = root (CVE-2026-3888)

https://www.openwall.com/lists/oss-security/2026/03/17/8
3•rkta•59m ago•0 comments

Show HN: The Lottery of Life

https://claude.ai/public/artifacts/a62c4bac-3c05-4443-9d0a-50a9bd3f9d8d
5•atulvi•1h ago•2 comments
Open in hackernews

Commandments of AI (Mindfulness)

2•xaj•1h ago
In an age of AI slop, we need to set ground rules for how to properly engage with this beast of burden. If you are sick of hearing the "AI will save everything" or "AI is our doom" voices owning the room, let's try to find some common ground we can all agree on, without turning into luddites or dooming the planet to slop.

1. thou shalt not divulge the dump - read and interpret output for yourself, never send it to others uninterpreted

2. thou shalt use critical thinking - avoid becoming mushy brain

3. thou shalt not divulge personal information - avoid divulging any content, whether text, media, or live audio

4. thou shalt conduct ai mindfulness - correct mistakes, clarify points, review findings consciously and deliberately

5. thou shalt cite thy source - when AI supplies a source, you should pass along the source (and review it)

6. thou shalt not love ai before humanity

7. thou shalt not depend fully upon ai

8. thou shalt not shame others for usage of ai

9. thou shalt not judge others for their transgressions

10. thou shalt spread the good word of the ai commandments