Article URL: https://brennan.day/an-open-letter-to-compassionate-left-leaning-ai-hating-animal-loving-meat-eaters/ Comments URL: https://news.ycombinator.com/item?id=48691450 Points: 5 # Comments: 0
Article URL: https://epoch.ai/MirrorCode Comments URL: https://news.ycombinator.com/item?id=48691419 Points: 3 # Comments: 1
Article URL: https://gizmodo.com/duckduckgo-unable-to-resist-the-pull-of-ai-mistakenly-claims-trump-died-of-rabies-2000777863 Comments URL: https://news.ycombinator.com/item?id=48691368 Points: 6 # Comments: 0
NYT shifts OpenAI/Microsoft copyright claims after SCOTUS ruling against Sony.
<p>If you use Claude Code, Cursor, or any AI coding agent, you know the problem: you ask the AI to review its own work, and it says "looks good." Every time.</p> <p>The AI isn't being lazy. It's sharing the same mental model that produced the code. It literally can't see what's wrong — the blind spots are baked in.</p> <h2> Why "Review this as a senior engineer" Fails </h2> <p>Generic role-playing produces generic findings. The AI fills in what it <em>thinks</em> a senior engineer would say, whi
<p>I just published <strong>SuperCompress</strong> to PyPI! 🎉</p> <p><code>pip install supercompress</code> — that's all it takes.</p> <h2> What is it? </h2> <p>A tiny ~5K parameter CPU policy that scores every line of context for relevance before sending to the LLM. It keeps only what matters for the answer.</p> <h2> The Numbers </h2> <ul> <li> <strong>65% fewer tokens</strong> → same answers</li> <li> <strong>100% oracle recall</strong> → never drops the answer line</li> <li> <strong>~60ms CP
<p>I spent tonight building a methodology for turning personal tools into open source contributions. Before publishing it, I decided to let the methodology review itself.</p> <p>The results changed how I think about AI-assisted code review.</p> <h2> The Method: Named-Persona Adversarial Review </h2> <p>The core idea is simple: instead of asking an AI to "review this code as a security engineer" (which produces generic, shallow feedback), you <strong>web-search actual engineers' documented philos
Surely, we can have rules that allow us to continue doing what we're doing
<p>The paradigm of Artificial Intelligence is undergoing a fundamental shift. We are moving rapidly from the era of <strong>stateless chat completions</strong>—where an LLM simply acts as an advanced text-autocomplete engine—to <strong>stateful, autonomous AI agents</strong>. These agents don't just talk; they <em>do</em>. They plan multi-step workflows, execute tools, read and write files, run test suites, and react to background triggers.</p> <p>At the center of this revolution is a powerful s
<p>I've been working on a side project called <strong>SuperCompress</strong> — an intelligent prompt compression system for LLMs. The idea is simple: most tokens you send to an LLM never need to be processed. They're padding, boilerplate, irrelevant context. But they still burn GPU cycles.</p> <p>I wanted to fix that.</p> <h2> The Problem </h2> <p>Working with LLM agents, I noticed something: every agent loop was sending massive context through the GPU. 10K tokens. 50K tokens. Sometimes more. Mo
Article URL: https://github.com/royashbrook/hush Comments URL: https://news.ycombinator.com/item?id=48691038 Points: 3 # Comments: 1
Article URL: https://github.com/vicmaster/framesmith Comments URL: https://news.ycombinator.com/item?id=48690983 Points: 2 # Comments: 0
Article URL: https://opencomputer.dev/blog/open-ava-bdr-agent/ Comments URL: https://news.ycombinator.com/item?id=48690926 Points: 2 # Comments: 0
Hey HNers - Riz here. I got together with a few guys and we built an LLM gateway. It's designed for small teams working on early-stage products, and can be deployed to AWS using a single command (i.e. `mantis deploy`). It's self-hosted, and is designed to belong to you. Comments URL: https://news.ycombinator.com/item?id=48690749 Points: 3 # Comments: 0
Article URL: https://github.com/team-telnyx/telnyx-code-examples/tree/main/build-conversational-workflow-nodejs Comments URL: https://news.ycombinator.com/item?id=48690730 Points: 3 # Comments: 0
Article URL: https://devops.com/akrites-the-latest-attempt-to-protect-open-source-from-ai-attacks-has-arrived/ Comments URL: https://news.ycombinator.com/item?id=48690692 Points: 2 # Comments: 0
AI agents have changed a lot in the last two years. The first could only answer one question at a time. Then came multi-turn chat, where the model could keep...
With AI-generated code, moving fast, and hitting PMF, what are some ways startups test their changes to deliver high-quality and what are some struggles? I find that no matter how much automated tests and such ... the reag bugs are still caught by customers or internal team(manually). Comments URL: https://news.ycombinator.com/item?id=48690570 Points: 2 # Comments: 0
General Intuition is using video game clips with embedded action labels to speed up AI training for robotics. The post General Intuition raises $320M to use video game data to train robots appeared first on The Robot Report .
<p><strong><a href="https://www.fernandoi.cl/posts/hackmyclaw/">What happened after 2,000 people tried to hack my AI assistant</a></strong></p> Fernando Irarrázaval ran a challenge on <a href="https://hackmyclaw.com/">hackmyclaw.com</a> to see if anyone could leak secrets held by his OpenClaw test instance by sending it email.</p> <p>Surprisingly, after 6,000 attempts (and $500 in token spend and a Google account suspension triggered by too many inbound emails) nobody managed to leak the secret.
“We don’t believe this kind of government access process should become the long-term default,” says OpenAI. “It keeps the best tools from users, developers, enterprises, cyber defenders, and global partners who need them.”
OpenAI's new flagship GPT-5.6 Sol beats Anthropic's Claude Mythos 5 in coding benchmarks, but the US government is forcing a restricted rollout. OpenAI isn't happy about it. The article OpenAI's GPT-5.6 Sol launches to rival Claude Mythos under government access rules it calls unsustainable appeared first on The Decoder .
Machine Intelligence
Article URL: https://vacationrentalprotocol.com/ Comments URL: https://news.ycombinator.com/item?id=48690144 Points: 2 # Comments: 0
The hire marks OpenAI's latest push into India, expanding offices, partnerships and hiring.
Article URL: https://agentbrush.dev/ Comments URL: https://news.ycombinator.com/item?id=48689984 Points: 3 # Comments: 0
Building software is becoming easy. Building a business is not. AI coding tools have dramatically lowered the barrier to creating SaaS products, AI applications, online services, marketplaces, and e-commerce stores. Entrepreneurs can now generate working products in hours instead of spending months writing code. But for most founders, development is no longer the bottleneck. Launching a successful business still requires validating demand, researching competitors, defining positioning, creating
Getting a head start on disease. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
<p><strong><a href="https://nesbitt.io/2026/06/26/incident-report-cve-2026-lgtm.html">Incident Report: CVE-2026-LGTM</a></strong></p> Spectacular hypothetical incident report by Andrew Nesbitt.</p> <blockquote> <p><strong>Day 2, 16:00 UTC</strong> --- Two AI review agents from competing vendors, both attached to a downstream pull request bumping <code>foxhole-lz4</code>, enter a disagreement loop over whether the package is malicious. After 340 comments and $41,255 in inference spend, Finance re
Article URL: https://florio.dev/dont-use-llm-communication/ Comments URL: https://news.ycombinator.com/item?id=48689561 Points: 5 # Comments: 5
Nvidia has dominated the AI chip market for years, but the era of total dependence might be ending. OpenAI just shared its plans to spice things up with Jalapeño, its custom inference chip built with Broadcom, joining Google, Apple, and SpaceX in a growing list of companies building their way out of single-supplier risk. The goal is less of a […]
Running an AI agent in production and curious how others handle this. Cost and latency monitoring is solved (Helicone, Langfuse, etc.) but I can't find a good answer for: how do you get alerted when the quality of your agent's responses drops before a customer tells you? What are you doing today? Comments URL: https://news.ycombinator.com/item?id=48689447 Points: 1 # Comments: 0
An endpoint agent cannot report its own absence. The 2026 Axonius Actionability Report , conducted with the Ponemon Institute and surveying 662 IT and security professionals, put a number on a gap SOC teams have worked around for years. Across the Axonius customer base , 12.7% of devices in a 298,000-device median inventory are missing their expected security agent. If a device has no agent, no management console shows it. If a CMDB record is stale, no reconciliation flags it. An employee who in
Epoch AI's new MirrorCode benchmark tests whether AI models can recreate complete programs without access to the original code. Claude Opus 4.7 leads with a 56 percent solve rate, rebuilding a 16,000-line toolkit in just 14 hours. But every model tested still fails on the most complex tasks. The article An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run appeared first on The Decoder .
I'm outside the AI industry, outside of academia, and no easy contacts into relevant areas. My background lends itself to exploring AI & reasonable level of care checking results. I was focusing on building practical useful things for a portfolio to help change industries mid-career, but that has become something a little different now. The exploration has led to an analytical framework that appears useful more broadly for looking at neural representational models. (It's not a model architecture
<blockquote cite="https://openai.com/index/previewing-gpt-5-6-sol/"><p>We're beginning a limited preview of the GPT‑5.6 series: Sol, our flagship model; Terra, a balanced model for everyday work; and Luna, a fast and affordable model. Terra has competitive performance to GPT‑5.5 while being 2x cheaper and Luna brings strong capability at our lowest cost. [...]</p> <p>We believe in broad access, and we plan to make GPT‑5.6 Sol, Terra, and Luna generally available in the coming weeks. As part of o
Article URL: https://openai.com/index/previewing-gpt-5-6-sol/ Comments URL: https://news.ycombinator.com/item?id=48689028 Points: 63 # Comments: 42
