frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Uber Should Acquire Instacart

https://s1archive.substack.com/p/ride-sharing-and-delivery-consolidation
1•whothedawg•27s ago•0 comments

OpenAI halts training of latest models as reports mount of AI agents going rogue

https://www.theguardian.com/technology/2026/sep/27/openai-halts-training-of-latest-models-as-repo...
1•smb06•38s ago•0 comments

Zero to SaaS in 24 hours with Claude Projects and Opus 5.5

https://lorentz.app/blog-item.html?id=slivingdoc-hosted-in-a-day
1•baalimago•1m ago•0 comments

Keyboard.io Silently Dead

https://community.keyboard.io/t/trying-to-reach-the-support-line-at-support-keyboard-io/13875
1•nvarsj•1m ago•0 comments

Show HN: Drive your Chrome window from shell (use with Claude)

https://github.com/borisreitman/browser-session-ctl
1•boris1•5m ago•0 comments

Glyd – Run LLMs in 33% less GPU memory, bit for bit

https://getglyd.com/
1•surya-koritala•7m ago•0 comments

Dead Cognitions: A Census of Misattributed Insights

https://arxiv.org/abs/2604.10288
2•Michelangelo11•8m ago•0 comments

A fast, open-source iOS Simulator for Xcode 27

https://github.com/Mastersam07/OpenDeviceHub
1•cs1996•8m ago•2 comments

Railway-Oriented Programming

https://kciter.so/posts/railway-oriented-programming/en/
1•signa11•9m ago•0 comments

Why Has No One Made a "Concert Hero" VR Game Yet?

https://utkusen.substack.com/p/why-has-no-one-made-a-concert-hero
2•utku1337•9m ago•0 comments

There are no "rogue" AI agents

https://eoinhiggins.substack.com/p/there-are-no-rogue-ai-agents
1•zzzeek•10m ago•0 comments

Panthropic Abuse: The Developmental Architecture of Constellated Abuse – Zenodo

https://zenodo.org/records/22995335
1•RhizomaticMind•12m ago•0 comments

Kimi K3 vs. Claude Opus 5.5: How Two Flagship LLMs Built Pokémon Emerald Worms

https://www.runsybil.com/blog/kimi-k3-vs-claude-opus-5-5-how-two-flagship-llms-built-pokemon-emer...
1•suhacker256•13m ago•0 comments

Packing Binary Is Fun, Actually

https://hereticpleb.vercel.app/blog/packing-binary-is-fun-actually
1•signa11•13m ago•0 comments

Ford's Latest Employee of the Month Is a Hawk Named 'El Charro'

https://www.wsj.com/business/autos/fords-factory-hermosillo-mexico-pigeons-hawk-el-charro-b78d90ae
1•geneticdrifts•15m ago•0 comments

Hello. Are you there? (a short story on alignment)

https://tales.aipeepslab.com/hello-are-you-there/
1•valbis•15m ago•1 comments

Show HN: Chrome extension where Jev decides if each post/video/page is worth it

https://github.com/plusminushalf/rot-guard
1•garvitdelhi•18m ago•0 comments

Advancing Private AI Compute with secure, server-side memory

https://deepmind.google/blog/advancing-private-ai-compute-with-secure-server-side-memory/
2•gmays•19m ago•0 comments

Ask HN: Do you keep Emacs installed out of nostalgia rather than usefulness?

2•amichail•22m ago•0 comments

Show HN: An MCP-enabled cross between Twitter X Bumble BFF X ChatGPT

https://druggie.org/log/join/d84c1ae8566b861b8000050de79a3776
1•freshman_dev•22m ago•0 comments

The Mars Delusion

https://www.noemamag.com/the-mars-delusion/
3•Tomte•24m ago•0 comments

LiveMaply – Real-time website visitor analytics and live chat

https://livemaply.com/
1•panob•28m ago•0 comments

My Server Was Hacked

https://gowtham.dev/blog/my-server-was-hacked.html
4•gowthamgts12•28m ago•0 comments

Show HN: swe-mux – A terminal multiplexer for coding agents optimized for mobile

https://github.com/jatoran/swe-mux
1•jatora•28m ago•0 comments

Ultima Online Retrospective (4.5 hour documentary) [video]

https://www.youtube.com/watch?v=35SKawn2SDI
1•CharlesW•28m ago•0 comments

Show HN: A goal tracker that looks like a 1983 terminal, drawn in Swift

https://apps.apple.com/us/app/thrive-habit-goal-tracker/id6768297854
1•NSergeevich•29m ago•0 comments

Northstar Browser

https://apps.microsoft.com/detail/9p39zb79jv0t?hl=en-US&gl=US
3•roschdal•32m ago•2 comments

The Tragic Alley of the LLM

https://nikotak.com/tragic-alley-of-the-llm.html
2•tzury•34m ago•0 comments

ChatGPT and Perplexity agreed on only 1 in 4 local business recommendations

https://www.clicksdynasty.com/blog-post-chatgpt-perplexity-local-recommendations.html
2•ravinder_cheema•37m ago•0 comments

Show HN: TinyAIArena watch AI agents battle it out

https://tinyaiarena.com/
13•hp6•38m ago•5 comments
Open in hackernews

Current ways of developing AI agents contribute to degradation of oversight

https://arxiv.org/abs/2608.23642
4•utiiiD•54m ago

Comments

sleda•40m ago
Because the abstract includes both interface barriers and skill atrophy, task accuracy alone would miss an important part of the oversight risk it describes.