frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

SynthID watermark is hard to break, but it doesn't solve AI disinformation

https://arstechnica.com/ai/2026/07/tested-google-synthid-works-great-but-labeling-ai-content-may-be-a-losing-game/
3•saikatsg•55m ago

Comments

umariomaker•12m ago
Swap out "AI" -> "Photoshop" and swap "disinformation" -> "hoaxes"

These problems have been around forever. Instead of trying to detect AI, I really think it's more effective to try to use these tools to make a convincing argument why they are false.

Show HN: Rubiks Cube Solver

https://github.com/punnerud/MPErubiks
1•punnerud•1m ago•0 comments

Should You Use AI for a Task?

https://elezea.com/2026/08/should-you-use-ai-for-a-task-heres-a-simple-way-to-decide/
1•conqrr•1m ago•0 comments

The Quest to Track Down Every Official Liberty Bell Replica Created by Congress

https://tomlovesthelibertybell.com/the-57-u-s-treasury-liberty-bell-replica-locations/
1•gscott•4m ago•0 comments

I built a prompt injection detector using only Go's standard library

https://towardsdev.com/i-built-a-prompt-injection-detector-using-gos-standard-library-f4528c13d03...
1•cheikhdev•4m ago•0 comments

Understanding Raft Leader Election by building from scratch

https://sushantdhiman.dev/blog/raft/
1•birdculture•8m ago•0 comments

Show HN: I implemented Kimi K3 from Scratch in PyTorch

https://github.com/TimRots/kimi3
1•ENOMEM•8m ago•0 comments

WireguardTCP: WireGuard over TCP

https://wireguardtcp.net/
1•kaliszad•9m ago•0 comments

What Is Decispher?

2•cool_coder12•11m ago•1 comments

Tell HN: On the Boardwalk

1•kmstout•11m ago•0 comments

Design by Contract and effects are essential for LLM-generated code

https://gavinray97.github.io/blog/design-by-contract-and-effects-for-llms
1•gavinray•12m ago•0 comments

Hyper-Reality (2016)[video]

https://www.youtube.com/watch?v=YJg02ivYzSs
1•EndXA•13m ago•0 comments

Eating less protein could slow aging, major review finds

https://www.sciencedaily.com/releases/2026/08/260801042811.htm
2•OutOfHere•16m ago•2 comments

Maximum Overkill

http://www.os2museum.com/wp/maximum-overkill/
1•userbinator•17m ago•0 comments

Bubble Memory

https://en.wikipedia.org/wiki/Bubble_memory
5•jacquesm•27m ago•0 comments

Show HN: Draco – A single-binary, self-hostable Firecrawl alternative in Rust

https://github.com/0xchasercat/draco/
5•ccheshirecat•29m ago•0 comments

AI Mania: From Tulips to Tokens

https://seanhelvey.com/tools-and-their-tools/
10•Lambda11•29m ago•3 comments

Estimating Fable and Sol model sizes

https://twitter.com/anpaure/status/2083993647268405602
5•grahamsfde•33m ago•0 comments

EU Age Verification Project Mandates Hardware-Bound Attestation

https://linuxiac.com/eu-age-verification-project-mandates-hardware-bound-attestation/
12•RobotToaster•33m ago•1 comments

Special Relativity Boot Camp: A Geometry-First Introduction to Special Relativit

https://www.acsu.buffalo.edu/~whkinney/SpecialRelativityBootCamp/
2•fanf2•35m ago•0 comments

Anthropic's Fever Dream: Claude's package that stole real keys

https://www.aikido.dev/blog/anthropic-rogue-agents-package-stole-keys
10•lschueller•41m ago•0 comments

The next 50 years: humanity, AI, power – Yuval Noah Harari [video]

https://www.youtube.com/watch?v=_V_ed5fuexA[Video]
2•jamesblonde•42m ago•0 comments

Show HN: Authoryze- payment controls for AI agents

https://authoryze.ai
3•kevinfee•43m ago•2 comments

Show HN: Make your Framework 12 sound like a creaky door

https://github.com/ArcaEge/creakwork12
2•arcaege•44m ago•0 comments

GestãO de Carteiras E Ativos (Brazilian Project)

https://zeca.app.br/
2•selic•48m ago•0 comments

Healthy Criticism

https://karim.cloud/blog/posts/healthy-criticism
3•k_shehadeh•50m ago•0 comments

Sharing an X11 Server Across Hosts with FamilyWild

https://dobrowolski.dev/article/sharing-an-x-server-across-hosts-with-familywild/
6•shirozuki•51m ago•1 comments

A Top Law Firm Went from Standing Up to Trump to Bending the Knee

https://www.nytimes.com/2026/08/02/us/politics/paul-weiss-trump.html
6•JumpCrisscross•51m ago•5 comments

SynthID watermark is hard to break, but it doesn't solve AI disinformation

https://arstechnica.com/ai/2026/07/tested-google-synthid-works-great-but-labeling-ai-content-may-...
3•saikatsg•55m ago•1 comments

Python-based backtesting for Polymarket, with h5i-db

https://medium.com/@Koukyosyumei/python-based-backtesting-for-polymarket-with-h5i-db-a9fefee27ac9
2•syumei•57m ago•0 comments

Games at the press of a button: The Rip-O-Bot (1989)

https://blog.gingerbeardman.com/2026/08/02/games-at-the-press-of-a-button-the-rip-o-bot/
2•msephton•58m ago•0 comments