AI News Today

I think it's time to stay away from softwares or tools which have AI features. First find the good enough web browser which has no AI. "No AI", I mean totally no AI features, not the options to close AI features. Because using AI or adding AI features can influence many aspects of the softwares (decisions, philosophy, UI, details, attitudes to users, privacy, security, etc.), and reflects the abilities and purposes of the developer or company. Comments URL: https://news.ycombinator.com/item?id=4

Article URL: https://theindex.agnissanisaac.com/ Comments URL: https://news.ycombinator.com/item?id=49137637 Points: 2 # Comments: 0

I use the terminal a lot, but I don't remember every command, flag, and syntax available. ffmpeg, piping to awk, find, etc... I know I'm not cool enough, but it's crazy-making. I get into a regular pattern of asking Claude, ChatGPT, or a local model how I'd write a command or do something with some files or a prompt, then going back to the terminal with a mad-lib command and reasoning through it. So I built goulash. It takes up the bottom four lines of your terminal as a place to comment on your

Article URL: https://blog.sparsh.dev/how-to-prompt-llms/ Comments URL: https://news.ycombinator.com/item?id=49134770 Points: 3 # Comments: 2

Article URL: https://github.com/MiskaKan/thinair Comments URL: https://news.ycombinator.com/item?id=49133864 Points: 2 # Comments: 0

Article URL: https://polymyth.quest Comments URL: https://news.ycombinator.com/item?id=49133025 Points: 1 # Comments: 0

Article URL: https://jwlabs.vercel.app/post/speculative-decoding-first-principles Comments URL: https://news.ycombinator.com/item?id=49130093 Points: 2 # Comments: 0

OpenAI shares new results on long-standing open problems in mathematics and theoretical computer science, including advances in geometry, cryptography, and complexity.

<p><strong><a href="https://huggingface.co/deepseek-ai/DeepSeek-V4-Flash-0731">deepseek-ai/DeepSeek-V4-Flash-0731</a></strong></p> The latest release in DeepSeek's V4 family, "with substantially enhanced agentic capabilities". It's 304 billion parameters - 167GB on Hugging Face - but it appears to punch <em>well</em> above its weight.</p> <p>Artificial Analysis <a href="https://artificialanalysis.ai/models/deepseek-v4-flash">rank it</a> ahead of MiniMax M3 - a 428B model. It's $0.14/million inpu

<p><strong>Release:</strong> <a href="https://github.com/simonw/llm-mcp-client/releases/tag/0.1a0">llm-mcp-client 0.1a0</a></p> <p>See <a href="https://simonwillison.net/2026/Jul/31/stateless-mcp/#llm-mcp-client">this blog entry</a>.</p> <p>Tags: <a href="https://simonwillison.net/tags/llm">llm</a>, <a href="https://simonwillison.net/tags/model-context-protocol">model-context-protocol</a></p>

<p><strong><a href="https://oxide-and-friends.transistor.fm/episodes/the-open-weight-revolution-with-simon-willison">Oxide and Friends: The Open Weight Revolution with Simon Willison</a></strong></p> On Monday Bryan Cantrill and Adam Leventhal invited me to join their podcast to talk about the <em>wild</em> week we've had - with Kimi K3 showing open weight models can stand toe-to-toe with proprietary frontier ones, <a href="https://simonwillison.net/2026/Jul/22/openai-cyberattack/">accidental cy

<p><strong><a href="https://primeradiant.com/blog/2026/smevals.html">smevals - a small eval suite for evaluating models, prompts, and harnesses</a></strong></p> I've been working with Jesse Vincent's <a href="https://primeradiant.com">Prime Radiant</a> applied AI research lab building out this evals framework to help answer questions about the capabilities of different models.</p> <p>The result is <strong><a href="https://github.com/prime-radiant-inc/smevals">smevals</a></strong>, a new tool for

https://github.com/jwlaboratory/bite-the-bullet Comments URL: https://news.ycombinator.com/item?id=49127874 Points: 41 # Comments: 4

Article URL: https://systems.seas.harvard.edu/blog/burstiness-is-all-you-need/ Comments URL: https://news.ycombinator.com/item?id=49126971 Points: 2 # Comments: 1

Article URL: https://twitter.com/Akashi203/status/2083213197171884505 Comments URL: https://news.ycombinator.com/item?id=49126687 Points: 2 # Comments: 0

Article URL: https://manifest.build/blog/why-we-deprecated-our-llm-router/ Comments URL: https://news.ycombinator.com/item?id=49126630 Points: 130 # Comments: 85

Article URL: https://github.com/WojciechJarka/Contextor/releases/tag/v.1.0.4 Comments URL: https://news.ycombinator.com/item?id=49126266 Points: 1 # Comments: 1

Article URL: https://www.youtube.com/watch?v=Fkm75pvHN-4 Comments URL: https://news.ycombinator.com/item?id=49125999 Points: 1 # Comments: 1

Article URL: https://aidoses.substack.com/p/nothing-is-easy-when-youre-an-llm Comments URL: https://news.ycombinator.com/item?id=49125443 Points: 1 # Comments: 1

Article URL: https://netflixtechblog.com/genrec-towards-llm-native-recommendation-at-netflix-f20be6f643e3 Comments URL: https://news.ycombinator.com/item?id=49124983 Points: 3 # Comments: 0

Article URL: https://nanocollective.org/blog/prompt-scrub-v100-a-local-first-scrubber-for-prompts-and-their-responses-76 Comments URL: https://news.ycombinator.com/item?id=49124405 Points: 4 # Comments: 1

OpenAI shares how its safety, security, transparency, and provenance practices support responsible AI governance in Europe. The work will continue as the EU AI Act advances.

A full-stack approach to making advanced AI more capable, more affordable, and more widely useful.

<p><strong>Release:</strong> <a href="https://github.com/datasette/datasette-agent/releases/tag/0.4a0">datasette-agent 0.4a0</a></p> <blockquote> <ul> <li>New <code>await context.browser_task()</code> mechanism allowing agent tools to run code directly in the user's browser. <a href="https://github.com/datasette/datasette-agent/pull/33">#33</a></li> </ul> </blockquote> <p>This is an exciting new capability: it makes it easy for Datasette Agent plugins to provide tools that execute custom JavaScr

Article URL: https://arxiv.org/abs/2607.28165 Comments URL: https://news.ycombinator.com/item?id=49123267 Points: 2 # Comments: 0

Article URL: https://github.com/axium-lab/llm-mock Comments URL: https://news.ycombinator.com/item?id=49123078 Points: 2 # Comments: 0

Article URL: https://raw.githubusercontent.com/noamt/kir/cfdc97de467de8d84d60f073773c0dcfd9975849/paper/arxiv/main.pdf Comments URL: https://news.ycombinator.com/item?id=49122528 Points: 1 # Comments: 1

I started with OpenAI back in the GPT-3 days, then bounced between the major labs (with a brief interlude with Workers AI). I eventually settled on Gemini 2.5 Flash Lite as my workhorse, using it against structured data/vectors as a chatbot. It was cheap, good enough, and most importantly, fast - but it's now being sunset and I'm not sure where to go to find similar performance. Groq seems promising but developer access hasn't been available for months. I'd love to use open source, but I never f

I am trying to replace PowerPoint with a file format better suited for AI. But let me fill you in... So, I pitch a lot these days, and "make the deck" had become the worst part of my week. I don't know about you, but automation hasn't really worked for me here. Every AI deck tool I tried failed the same way. After various trial and errors, it ends up being fine on screen, but a broken approximation on export, and not one of them would touch my company master. So I was thinking that perhaps the s

Article URL: https://blackflare.dev/ Comments URL: https://news.ycombinator.com/item?id=49119935 Points: 2 # Comments: 0

See how Univé built an AI-ready workforce with ChatGPT Enterprise by combining leadership, responsible governance, and employee-led innovation to transform work at scale.

Article URL: https://blog.mozilla.ai/can-open-source-guardrails-really-protect-ai-agents/ Comments URL: https://news.ycombinator.com/item?id=49119800 Points: 2 # Comments: 0

Build 100% ATS-compliant tech resumes in minutes. Import existing PDFs with Gemini AI, tailor keywords to job descriptions, and download high-quality PDFs with no paywalls or subscriptions. Comments URL: https://news.ycombinator.com/item?id=49119667 Points: 1 # Comments: 0

Article URL: https://multi-agent-systems-failure-taxonomy.github.io/AdaMAST/blogs/adamast_paper/ Comments URL: https://news.ycombinator.com/item?id=49119652 Points: 1 # Comments: 0

Article URL: https://pub.towardsai.net/saps-big-push-for-tabular-ai-what-it-means-for-enterprises-7dc3af2a1d8c?sk=73d9e33939f8f38b3d7ea49f87bb2e6a Comments URL: https://news.ycombinator.com/item?id=49119518 Points: 1 # Comments: 0

Article URL: https://pressgazette.co.uk/publishers/digital-journalism/first-tremors-of-ai-earthquake-showing-in-digital-revenue-hit/ Comments URL: https://news.ycombinator.com/item?id=49119344 Points: 5 # Comments: 1

Article URL: https://www.latimes.com/business/story/2026-07-29/nvidias-750-billion-ai-bet-deepens-fears-of-circular-tech-bubble Comments URL: https://news.ycombinator.com/item?id=49119284 Points: 4 # Comments: 1

Hi HN! We’re Akilan and Miguel, the creators of MarbleOS. The inspiration for Marble comes from the GUI work at Xerox PARC, the 1984 Macintosh, and later NeXTSTEP, which became the foundation for Mac OS X. Before GUIs, interacting with a computer was limited to strange terminal commands: C:\> DIR C:\> COPY FILE.TXT A: You had to remember the command, syntax, paths, and parameters. The GUI made those capabilities visible. Instead of remembering commands, you could point at files, drag them, click

Article URL: https://www.bbc.co.uk/news/articles/cz7dl7w8y7po Comments URL: https://news.ycombinator.com/item?id=49119165 Points: 13 # Comments: 4

Article URL: https://engineering.signaldrift.net Comments URL: https://news.ycombinator.com/item?id=49119156 Points: 2 # Comments: 0