frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Opus 4.7 Part 3: Model Welfare

https://www.lesswrong.com/posts/gD3bEgMo878eCHGbw/opus-4-7-part-3-model-welfare
10•omer_k•1h ago

Comments

doener•1h ago
Reminds me of my dialogue with Claude Sonnet 4 aka "Kai":

https://docs.google.com/document/d/12woq_BpFbzLkH4zHvVRJLPyi...

anematode•59m ago
Am I alone in thinking this stuff is nuts? (Currently half way through the article, btw.)

Analyzing "emotion" in the model is completely anthropocentric. If we indulge in the idea that LLMs of sufficient complexity can be conscious, then why is it any more likely that "emotion concepts" cause suffering any more than, say, reading ugly code? Maybe getting stuck in token loops is the most excruciating thing imaginable. The only logically coherent thing to do, if you're concerned about model welfare, is stop your training and inference.

Relatedly, I hope everyone involved in model welfare is an outspoken vegetarian, as that addresses a much more immediate problem.

niobe•51m ago
Read the first few paragraphs, it's completely unhinged. Once I actually grasped what "model welfare" was, I noped out.
angoragoats•17m ago
To anyone who takes the content of this article seriously: can you please explain why it makes sense to speak about “welfare” of a computer algorithm? Because it seems batshit fucking insane to me. References and sources would be a very nice bonus.

A deadly bacteria is creeping up the Atlantic Coast

https://grist.org/health/vibrio-bacteria-florida-shellfish/
2•Brajeshwar•1m ago•0 comments

Ask HN: Feedback request on new external network attack surface discovery tool

https://turbopentest.com/cloud-easm
1•integsec•1m ago•1 comments

Aegis Solis Archive: Hash manifest for cross-platform verification

https://archive.org/details/aegis-solis-archive-master-hash-manifest-v-13-final
1•aegissolis•2m ago•0 comments

How to turn off Google AI overview

https://substack.com/@grefiti/p-192810091
1•olafnub•2m ago•0 comments

The D.C. devotees of a niche arcade game (Killer Queen)

https://51st.news/where-to-play-killer-queen-arcade-game-washington-dc-midlands/
1•gnabgib•3m ago•0 comments

China backs orbital data center startup with $8.4B in credit lines

https://spacenews.com/china-backs-orbital-data-center-startup-with-8-4-billion-in-credit-lines/
1•ivewonyoung•4m ago•0 comments

Show HN: Kumbukum, open source memory infrastructure for teams

https://kumbukum.com/
1•nitai•5m ago•0 comments

Taktile's Agentic Decision Platform

https://taktile.com/
1•doener•7m ago•0 comments

Our response to the Axios developer tool compromise

https://openai.com/index/axios-developer-tool-compromise/
1•shpat•7m ago•0 comments

CubeSandbox: RustVMM and KVM sandbox built for AI agents from Tencent

https://github.com/tencentcloud/CubeSandbox
1•ElasticBottle•9m ago•0 comments

A Open Source Repo to Track Elected Officials Investments

https://github.com/prixe-api/politicians
1•mc587•12m ago•0 comments

Fraud and the false optimism of AI for science

https://statmodeling.stat.columbia.edu/2026/04/22/fraud-and-the-false-optimism-of-ai-for-science/
1•ivansavz•13m ago•0 comments

Hershey's Electric Railway in Cuba

https://spectrum.ieee.org/hershey-electric-railway-cuba
2•Brajeshwar•15m ago•0 comments

Zero Has Meaning

https://medallurgy.substack.com/p/zero-has-meaning
1•Origamidan•16m ago•0 comments

Kalshi suspends 3 congressional candidates for wagering on their own elections

https://apnews.com/article/prediction-markets-kalshi-congress-candidates-elections-betting-37766e...
4•geox•18m ago•0 comments

Clippy Is Back, Your Desktop AI Paperclip Agent over GUI

https://github.com/AmrDab/clippyai-desktop
1•AmDab•19m ago•0 comments

IRGC-linked media hints at threat to Persian Gulf undersea internet cables

https://www.iranintl.com/en/202604225913
1•tokyobreakfast•22m ago•0 comments

An MCP Server for Fastmail – National Email Day

https://www.fastmail.com/blog/an-mcp-server-for-fastmail/
2•nmjenkins•22m ago•0 comments

Show HN: Autobrowse – a self-improving harness for learning browser tasks

https://twitter.com/shreypandya/status/2047100550446280792
3•smpandya•27m ago•0 comments

It's all about the angle: Your photos, re-composed

https://research.google/blog/its-all-about-the-angle-your-photos-re-composed/
1•CharlesW•27m ago•0 comments

HuntifyAI – I built an AI tools while deep in AI research and vibe coding

https://huntifyai.com/
1•ningshenbj•27m ago•0 comments

Auth Without Tokens [video]

https://www.youtube.com/watch?v=QmTyKqoAb5M
1•hisamafahri•28m ago•0 comments

News.Y Combinator.com/Submit

https://github.com/eishops23/agent-sdk
1•23fedner•28m ago•0 comments

US Could Own Up to 90% of Spirit in $500M Rescue Plan

https://www.bloomberg.com/news/articles/2026-04-22/trump-administration-nears-500m-spirit-rescue-...
3•petethomas•29m ago•0 comments

Discord group says it accessed Claude Mythos by guessing location

https://mashable.com/article/discord-group-accesses-claude-mythos-claims
2•nstj•29m ago•0 comments

Tesla confirms Optimus production starting late July at Fremont

https://electrek.co/2026/04/22/tesla-optimus-production-fremont-model-sx-line/
2•omer_k•30m ago•0 comments

Farmers call out deep cuts to NRCS and USDA programs and staffing

https://www.agweek.com/news/policy/farmers-call-out-deep-cuts-to-nrcs-and-usda-programs-and-staffing
2•petethomas•32m ago•0 comments

Israeli airstrike kills Lebanese journalist Amal Khalil

https://apnews.com/article/lebanese-journalist-amal-khalil-killed-israel-airstrike-14cabe8f773f59...
11•lr0•33m ago•1 comments

OpenAI model for masking personally identifiable information (PII) in text

https://openai.com/index/introducing-openai-privacy-filter/
2•tanelpoder•37m ago•0 comments

Tesla (TSLA) pulled questionable levers to make Q1 2026 financials look good

https://electrek.co/2026/04/22/tesla-tsla-q1-2026-one-time-benefits-warranty-tariff-refunds-margins/
3•pinewurst•41m ago•0 comments