frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

McCLIM and 7GUIs – Part 1: The Counter

https://turtleware.eu/posts/McCLIM-and-7GUIs---Part-1-The-Counter.html
1•ramenbytes•1m ago•0 comments

So whats the next word, then? Almost-no-math intro to transformer models

https://matthias-kainer.de/blog/posts/so-whats-the-next-word-then-/
1•oesimania•3m ago•0 comments

Ed Zitron: The Hater's Guide to Microsoft

https://bsky.app/profile/edzitron.com/post/3me7ibeym2c2n
2•vintagedave•6m ago•1 comments

UK infants ill after drinking contaminated baby formula of Nestle and Danone

https://www.bbc.com/news/articles/c931rxnwn3lo
1•__natty__•6m ago•0 comments

Show HN: Android-based audio player for seniors – Homer Audio Player

https://homeraudioplayer.app
1•cinusek•6m ago•0 comments

Starter Template for Ory Kratos

https://github.com/Samuelk0nrad/docker-ory
1•samuel_0xK•8m ago•0 comments

LLMs are powerful, but enterprises are deterministic by nature

1•prateekdalal•12m ago•0 comments

Make your iPad 3 a touchscreen for your computer

https://github.com/lemonjesus/ipad-touch-screen
2•0y•17m ago•1 comments

Internationalization and Localization in the Age of Agents

https://myblog.ru/internationalization-and-localization-in-the-age-of-agents
1•xenator•17m ago•0 comments

Building a Custom Clawdbot Workflow to Automate Website Creation

https://seedance2api.org/
1•pekingzcc•20m ago•1 comments

Why the "Taiwan Dome" won't survive a Chinese attack

https://www.lowyinstitute.org/the-interpreter/why-taiwan-dome-won-t-survive-chinese-attack
1•ryan_j_naughton•20m ago•0 comments

Xkcd: Game AIs

https://xkcd.com/1002/
1•ravenical•21m ago•0 comments

Windows 11 is finally killing off legacy printer drivers in 2026

https://www.windowscentral.com/microsoft/windows-11/windows-11-finally-pulls-the-plug-on-legacy-p...
1•ValdikSS•22m ago•0 comments

From Offloading to Engagement (Study on Generative AI)

https://www.mdpi.com/2306-5729/10/11/172
1•boshomi•24m ago•1 comments

AI for People

https://justsitandgrin.im/posts/ai-for-people/
1•dive•25m ago•0 comments

Rome is studded with cannon balls (2022)

https://essenceofrome.com/rome-is-studded-with-cannon-balls
1•thomassmith65•30m ago•0 comments

8-piece tablebase development on Lichess (op1 partial)

https://lichess.org/@/Lichess/blog/op1-partial-8-piece-tablebase-available/1ptPBDpC
2•somethingp•32m ago•0 comments

US to bankroll far-right think tanks in Europe against digital laws

https://www.brusselstimes.com/1957195/us-to-fund-far-right-forces-in-europe-tbtb
3•saubeidl•33m ago•0 comments

Ask HN: Have AI companies replaced their own SaaS usage with agents?

1•tuxpenguine•36m ago•0 comments

pi-nes

https://twitter.com/thomasmustier/status/2018362041506132205
1•tosh•38m ago•0 comments

Show HN: Crew – Multi-agent orchestration tool for AI-assisted development

https://github.com/garnetliu/crew
1•gl2334•38m ago•0 comments

New hire fixed a problem so fast, their boss left to become a yoga instructor

https://www.theregister.com/2026/02/06/on_call/
1•Brajeshwar•40m ago•0 comments

Four horsemen of the AI-pocalypse line up capex bigger than Israel's GDP

https://www.theregister.com/2026/02/06/ai_capex_plans/
1•Brajeshwar•40m ago•0 comments

A free Dynamic QR Code generator (no expiring links)

https://free-dynamic-qr-generator.com/
1•nookeshkarri7•41m ago•1 comments

nextTick but for React.js

https://suhaotian.github.io/use-next-tick/
1•jeremy_su•42m ago•0 comments

Show HN: I Built an AI-Powered Pull Request Review Tool

https://github.com/HighGarden-Studio/HighReview
1•highgarden•43m ago•0 comments

Git-am applies commit message diffs

https://lore.kernel.org/git/bcqvh7ahjjgzpgxwnr4kh3hfkksfruf54refyry3ha7qk7dldf@fij5calmscvm/
1•rkta•45m ago•0 comments

ClawEmail: 1min setup for OpenClaw agents with Gmail, Docs

https://clawemail.com
1•aleks5678•52m ago•1 comments

UnAutomating the Economy: More Labor but at What Cost?

https://www.greshm.org/blog/unautomating-the-economy/
1•Suncho•59m ago•1 comments

Show HN: Gettorr – Stream magnet links in the browser via WebRTC (no install)

https://gettorr.com/
1•BenaouidateMed•1h ago•0 comments
Open in hackernews

Microsoft to Use AI from Anthropic in Partial Shift from OpenAI

https://www.reuters.com/business/microsoft-use-some-ai-anthropic-shift-openai-information-reports-2025-09-09/
9•andsoitis•5mo ago

Comments

consumer451•5mo ago
This is a really interesting development.

> Developers making Office AI features found Anthropic's latest models performed better than OpenAI in automating tasks such as financial functions in Excel or generating Powerpoint presentations based on instructions, the report said, citing one of the two people involved in the effort.

> Microsoft will pay its cloud rival Amazon Web Services to access the Anthropic models, according to the report. AWS is one of Anthropic's largest shareholders.

Any folks here who integrate LLMs into their products have opinions on when to use Anthropic's models over OpenAI's?

A_D_E_P_T•5mo ago
In my experience, Anthropic's models may be slightly better for coding. (I think that GPT-5 Thinking/Pro are ahead, but I know that others disagree.) Where just about everything else is concerned, though, my impression is that they're way behind. They're particularly terrible at analyzing images. I recently gave GPT-5, o3, and Gemini 2.5 Pro an image analysis task (literally as simple as "describe this image"), and all of them passed the test with flying colors. I gave the exact same task to Opus 4.1 and it was spectacularly botched, twice.

So I might use Opus for coding and perhaps for agentic stuff, but I'd hesitate to use it for anything else.

al_borland•5mo ago
The way things have been going, it seems like the only right move is to make the products AI model agnostic, where any model could plugin. There is too much flux in this space, and even based on user preferences, to be tied too tightly to any one model or vendor.
consumer451•5mo ago
I used Vercel's ai sdk like everyone else, to try to stay agnostic. However, when I am actually working on a feature that uses an LLM API call, I find that each prompt might need tweaking specific to that model to get great results. That means if I switched models for that feature, I would likely have to that redo the prompting. Or, am I doing this all wrong?