Some Basic LLM Etiquette
Article URL: https://steenbok.space/blog/ai-etiquette/ Comments URL: https://news.ycombinator.com/item?id=48769662 Points: 2 # Comments: 0
Article URL: https://steenbok.space/blog/ai-etiquette/ Comments URL: https://news.ycombinator.com/item?id=48769662 Points: 2 # Comments: 0
Article URL: https://www.youtube.com/watch?v=cfaZZPjA3g0 Comments URL: https://news.ycombinator.com/item?id=48769645 Points: 2 # Comments: 0
Comments URL: https://news.ycombinator.com/item?id=48769635 Points: 4 # Comments: 0
<p>A potential deal with the government would allow international tech companies to mine the creative work of Australian musicians. Some of the prime minister’s favourite artists told the Guardian how they feel about it</p><ul><li><p><a href="https://www.theguardian.com/australia-news/2026/jul/01/australia-ai-copyright-laws-creatives-sound-alarm-albanese-government-datacentres">Creatives sound alarm on copyright as Pocock calls $50bn datacentre proposal ‘ultimate dirty deal’</a></p></li></ul><p>

New hope for millions with the disease. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Article URL: https://openllmstack.com/blog/open-source-llm-statistics/ Comments URL: https://news.ycombinator.com/item?id=48769528 Points: 2 # Comments: 0
<p>I built an AI-powered daily ritual app called Wishyze (<a href="https://wishyze.com" rel="noopener noreferrer">wishyze.com</a>). I watched 28,547 users try to change their lives, and what I learned surprised me.</p> <h2> The Streak Obsession </h2> <p>Every habit tracker worships at the altar of the streak: Duolingo, Streaks, GitHub. Wishyze has a 93-day longest-streak leaderboard too. Streaks create momentum, but data revealed an uncomfortable truth:</p> <p><strong>73% of users who quit do no
<p>I've been playing with a small browser-side prototype that tries to answer a boring but increasingly important question:<br> Where did this piece of web content come from?<br> Not “is this AI-generated?” I think that question is already getting messy. Detectors can be useful sometimes, but I would not build trust on top of them alone.<br> The simpler question is: When was this page first seen? Who saw it?<br> Did it change? Can we still replay what was captured?</p> <p>While building this, I
<p>A 429 error is easy to misread.</p> <p>The first instinct is often:</p> <p>"The provider is unstable."</p> <p>Sometimes that is true. But in OpenAI-compatible API systems, a 429 can also come from a much more local problem:</p> <ul> <li>too many concurrent requests;</li> <li>retries firing too aggressively;</li> <li>an agent loop making more calls than expected;</li> <li>fallback models multiplying traffic;</li> <li>one shared key serving several environments;</li> <li>a batch job and product
<p>Yes. Below are two implementations of the same task: <strong>an agent applies a fix to a file, runs tests, commits the result or rolls back</strong>. The first version follows a typical imperative style. The second is structured through the A11 lens. In the second version's code, I have explicitly marked locations that characterize structural resilience. These markers do not affect code execution; they are architectural annotations demonstrating <em>where exactly</em> the framework holds the
<h1> Ringkasan Tren GitHub Minggu Ini: Efisiensi Kode, OCR Jangka Panjang, dan Evolusi Agen AI </h1> <p>Minggu ini, papan tren GitHub mencerminkan pergeseran paradigma yang menarik dalam pengembangan perangkat lunak modern. Fokus utama bukan lagi sekadar menciptakan fungsi baru, tetapi mengoptimalkan <em>effort</em> pengembang. Dari alat yang secara agresif mengurangi baris kode hingga sistem OCR yang mampu menangani dokumen beresolusi tinggi tanpa batas, tren mingguan ini menyoroti tiga tema be
Article URL: https://www.wsj.com/tech/ai/the-worlds-top-economists-are-sounding-the-alarm-on-ai-d99055b6 Comments URL: https://news.ycombinator.com/item?id=48769316 Points: 4 # Comments: 2
Article URL: https://arxiv.org/abs/2606.29657 Comments URL: https://news.ycombinator.com/item?id=48769266 Points: 2 # Comments: 0
Article URL: https://charitydotwtf.substack.com/p/in-defense-of-ai-mandates Comments URL: https://news.ycombinator.com/item?id=48769250 Points: 2 # Comments: 0
Article URL: https://www.youtube.com/watch?v=QQC6V8T8ry4 Comments URL: https://news.ycombinator.com/item?id=48769238 Points: 1 # Comments: 0
Two-thirds of enterprises have hedged their AI model strategy, and the past few weeks of controversy around Anthropic’s Claude Fable 5 model showed why that posture has gone mainstream. On June 12, a U.S. export-control order pulled Anthropic's Claude Fable 5 — the most capable model on the market — offline for every customer, with no warning and no timeline. It returned this week wrapped in tighter safeguards , after China's Z.ai released its open-weights GLM-5.2 into the vacuum . New VentureBe

Article URL: https://bennorthmore.com/journals/welcome-to-the-suspicion-economy/ Comments URL: https://news.ycombinator.com/item?id=48769181 Points: 4 # Comments: 0
I know we are very wary of AI slop but disclaimer: AI was used. I don't think that it is slop though and it was mostly used for formatting. Comments URL: https://news.ycombinator.com/item?id=48769064 Points: 4 # Comments: 0
I built a free tool to convert a PDF to excel. It's more nuanced than first meets the eye: a document can not contain any tables but still lend itself to becoming one or more excel sheets (e.g. a rental contract can be a list of renter names, or a doctor's note may contain a list of allergies and a list of diagnoses). Anticipating the obvious question of "why can't I just shove this into GPT" - DocuPipe is far better accuracy (see benchmark [here]( https://github.com/DocuPipe/DocuBench )), espec
<p> Talk to GitHub Copilot out loud </p> <p> <a href="https://www.producthunt.com/products/vox-5?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1186717?app_id=339">Link</a> </p>
We built AI that makes home services businesses fully autonomous. It runs your entire home services business 24/7. The AI: - Handles inbound calls and SMS - Creates quotes - Dispatches your workers - Runs ads - Follows up with leads - Handles customer support We already run the entirety of 3 cleaning companies in California using AI agents. You can apply for early access by booking some time with us on the site! I'm open to feedback so let me know in the comments! Comments URL: https://news.ycom
Vercel Flags segments can now be managed from the Vercel CLI with the new vercel flags segments command. A segment is the targeting primitive a flag uses to decide who sees what. Membership composes from three repeatable tokens: include: , exclude: , and rule: . Pass them to --add or --remove for incremental edits. For full replacement, --data takes the entire segment definition as raw JSON. All segment commands support --json output, making them scriptable from CI, local workflows, and agent-dr
Your agent can now inspect Agent Runs via the Vercel MCP and CLI for eve , the open-source agent framework. eve traces are automatically ingested when deployed to Vercel and available as Agent Runs. The new Vercel MCP tools and Vercel CLI commands let you find projects with runs, list recent runs, and retrieve full traces, including reasoning, tool calls, and token usage. Vercel MCP tools: list_agent_run_projects : Find projects in a team with Agent Runs activity. list_agent_runs : List recent r
Article URL: https://righttointelligence.org/ Comments URL: https://news.ycombinator.com/item?id=48768951 Points: 496 # Comments: 175
Article URL: https://github.com/vaishcodescape/data-spear Comments URL: https://news.ycombinator.com/item?id=48768927 Points: 2 # Comments: 0
At an internal meeting, the Meta CEO reportedly said that AI development efforts were not moving as quickly as anticipated.
Article URL: https://www.phoronix.com/news/Linux-AI-Attribution-Again Comments URL: https://news.ycombinator.com/item?id=48768765 Points: 1 # Comments: 0
Hi all, I just got hands on a domain for my name that I've been looking for quite some time now. I want to start pointing it to something. I started looking for options, but there are so many junk articles and I think my knowledge is a bit outdated. I am a developer with around 10+ YOE. I want to host a simple page of my profile, and then if possible a few articles / blogs. What are my options ? AI suggests Cloudflare + Github pages. I am looking for cheap/free options. Comments URL: https://new
Article URL: https://thenewstack.io/ai-infrastructure-cargo-theft/ Comments URL: https://news.ycombinator.com/item?id=48768623 Points: 6 # Comments: 0
Leading AI researcher Yan LeCun has a start-up which is developing a more flexible AI system.

Pocket is listed in app stores but seemingly unavailable in the US.

MeetingTV wants to see the evidence
If the existing AI guideline approach is akin to giving a criminal (the AI) moral education (training) to encourage good behavior, how about creating a kernel-level switch that forcibly cuts off the electrical signals to its muscles the moment it grabs a weapon—that is, the moment it attempts a "jailbreak" via an adversarial method? Comments URL: https://news.ycombinator.com/item?id=48768344 Points: 1 # Comments: 1
Dunno about you, but I cannot keep up with the volume of AI research. Reading one paper properly takes 30–60 minutes, and it's hard to know which are even worth it. I built Gist, an AI summarizer for ADHD brains. Each paper is a clean slide deck with four layers you move through: 1. the Gist – the whole idea 1-2 sentences 2. Logic – breakdown of the argument + evidence in discrete ideas 3. Counter-Argument – strongest argument against the author (AI generated) 4. Steelman – transcendent rebuttal
Capcat is a python based CLI/TUI FOSS utility for Ethical archiving of given website or RSS source. The github repo: https://github.com/stayukasabov/capcat It is generated with NLP, context-engineering, spec-driven development and LLMs. Fully functional at http://capcat.org , with instructions for usage and documentation. The project started from my personal needs of simple archiving with structure and moved to product design/MVP exercise. I am longtime HN user, and the most value I got in years
<p> An AI agent that reproduces GitHub issues for you </p> <p> <a href="https://www.producthunt.com/products/osloq?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="https://www.producthunt.com/r/p/1186689?app_id=339">Link</a> </p>
Article URL: https://ankur-chr.github.io/inside-the-model/ Comments URL: https://news.ycombinator.com/item?id=48768249 Points: 2 # Comments: 0
Me and my team Ostrya AI were fed up with broken meet notetakers So we decided to roll out our own. Introducing OpenNotetaker (OSS). A small notetaker that works with hinglish and is able to extract all the useful information out of your meetings. Some highlight features: - Works with hindi, english, and hindi - Works on google meet - Connects to your google calendar and able to autojoin meetings Link: https://lnkd.in/gPiAv-J7 Give it a try if you take meetings in hinglish! Comments URL: https:/
Article URL: https://www.runmirrors.com/ Comments URL: https://news.ycombinator.com/item?id=48768200 Points: 3 # Comments: 0