AI News Today

Article URL: https://github.com/Loxia-ai/onbuzz-community Comments URL: https://news.ycombinator.com/item?id=48678218 Points: 2 # Comments: 1

Article URL: https://promptowl.ai/resources/verifiable-context-governance/ Comments URL: https://news.ycombinator.com/item?id=48678194 Points: 2 # Comments: 1

<p>Over the past few months I've been building with Cursor, Claude Code, and Codex - same tools a lot of builders are running right now. The pace is genuinely wild. I've shipped MVPs for friends where we went from a rough PRD to a working demo in a single day: auth wired up, database in place, deployed, link sent before midnight. Some of it fully vibe-coded. Some of it structured. All of it fast.</p> <p>You're seeing this everywhere now. Builders spinning up entire products over a weekend, somet

Article URL: https://github.com/secondly-com/openphone Comments URL: https://news.ycombinator.com/item?id=48678160 Points: 3 # Comments: 0

Article URL: https://www.freightwaves.com/news/cargo-thieves-are-following-the-ai-boom Comments URL: https://news.ycombinator.com/item?id=48678143 Points: 3 # Comments: 0

<p><em>Originally published on <a href="https://aideazz.xyz/blog/i-built-a-131-test-eval-harness-before-writing-new-features-heres-the-silent-failure-it-caught" rel="noopener noreferrer">AIdeazz</a> — cross-posted here with canonical link.</em></p> <p>The agent passed every unit test and still gave a user financial advice it was explicitly instructed never to give. No exception thrown, no log line in red, no failed assertion. The function returned a clean 200 and a well-formed string. I only fou

Article URL: https://medium.com/open-ai/7-open-source-ai-tools-that-feel-like-cheating-2026-b6e1587bd0e6 Comments URL: https://news.ycombinator.com/item?id=48678137 Points: 2 # Comments: 0

Article URL: https://plansera.ai/ Comments URL: https://news.ycombinator.com/item?id=48678128 Points: 2 # Comments: 0

<p>I wanted to share a quick update on how O-AI is progressing.</p> <p>Over the past few weeks, I've been rebuilding and improving almost every part of O-AI. The goal has always been to create something that's genuinely useful—not just another AI chatbot.</p> <p>Here are some of the biggest updates so far:</p> <p>• O-AI Agent Extension – O-AI is evolving beyond simple conversations and can now work more like an AI agent, capable of handling more advanced and multi-step tasks.</p> <p>• O-Code – A

Article URL: https://github.com/adatarwa/operator-skills Comments URL: https://news.ycombinator.com/item?id=48678036 Points: 2 # Comments: 0

<p>We run <a href="https://auditvibecoding.com" rel="noopener noreferrer">Audit Vibe Coding</a> at Inithouse, a security audit tool built specifically for AI-generated projects. After scanning hundreds of vibecoded apps, the same seven vulnerabilities show up over and over. None of them are exotic. All of them are fixable in under five minutes each.</p> <p>Here they are, with a grep command you can run right now to check your own codebase.</p> <h2> 1. Hardcoded secrets in source code </h2> <p>AI

<p>Over the past few weeks, I’ve been experimenting with local language models. Like a lot of developers, I’ve used cloud AI assistants quite a bit, but I kept asking myself one simple question:</p> <blockquote> <p>Can a local LLM actually help me build a real, production-ready project?</p> </blockquote> <p>So I decided to find out for myself.</p> <p>I challenged myself to build an entire developer tools website using only a local AI model running on my own machine. No cloud assistant. No extern

Aptiv provides sensor fusion from radar and vision using AI to enable Robust.AI's Carter robot to move safely around people. The post Robust.AI chooses Aptiv PULSE sensor for Gen 3 Carter mobile robot appeared first on The Robot Report .

<p>There was something fulfilling that died within me when I no longer had to debug my code.</p> <p>I can't tell you the specific date that this feeling began to spread. But there's a growing chorus of experienced developers saying the quiet part out loud: coding isn't fun anymore, and AI might be the reason.</p> <h2> The Thread That Hit a Nerve </h2> <p>A viral thread in a popular senior developer community recently went boom. The premise was simple: “coding doesn't seem fun anymore.” It wasn't

The company said it is discontinuing its email inbox in favor of its AI agent offering as users are increasingly handing over the reins of their email to the agents.

Article URL: https://www.adweek.com/media/youtube-content-creators-are-winning-the-ai-search-game-jellyfish-data-finds/ Comments URL: https://news.ycombinator.com/item?id=48677946 Points: 2 # Comments: 0

<p>If you build AI agents, scrapers, or RAG pipelines, you have run into this: feeding raw HTML to a model is mostly waste. The <code><div></code>, the <code>class="..."</code>, the nav, the cookie banner. None of it is the content you wanted, and all of it costs tokens.</p> <p>I knew this was inefficient. I did not appreciate how inefficient until I measured it. Here is a single page of GitHub's documentation, run through an audit:<br> </p> <div class="highlight js-code-highlight"> <pre class="

Sometimes I’ll ask Claude to think through a problem with me, and instead of talking, it immediately starts editing files and writing code. Like… bro, I’m not asking you to do it yet. Let’s think first. Comments URL: https://news.ycombinator.com/item?id=48677917 Points: 3 # Comments: 3

The app is a traditional social media scheduler but has a fully featured API and I have personally used it with OpenClaw and Hermes ai agents with the included API keys Comments URL: https://news.ycombinator.com/item?id=48677897 Points: 2 # Comments: 0

Article URL: https://www.augmentcode.com/blog/ai-transformation-is-a-systems-problem Comments URL: https://news.ycombinator.com/item?id=48677877 Points: 3 # Comments: 0

<p>Like many developers, we started with the obvious idea.</p> <p><em>"Let's expose everything."</em></p> <p>Every model.<br> Every setting.<br> Every parameter.<br> Every generation endpoint.</p> <p>If our AI filmmaking platform could do it, the API should be able to do it too.</p> <p>Simple.</p> <p>Or so we thought.</p> <h2> The more power we exposed... </h2> <p>...the less interesting the results became.</p> <p>Not because the models weren't capable.</p> <p>Not because the API wasn't well des

Notion is "going all in on using agents to run your inbox."

<p>AI app builders crossed a line in the last year. They went from generating throwaway demos to scaffolding apps people actually deploy and charge money for. I've been shipping full-stack apps this way, and the honest engineering picture is more interesting than either the hype or the backlash. Here's the breakdown.</p> <h2> What the model reliably gets right </h2> <p>The boring 80% is genuinely solved:</p> <ul> <li> <strong>Scaffolding and routing.</strong> A coherent project structure, sane f

Article URL: https://www.cartai.ai/ Comments URL: https://news.ycombinator.com/item?id=48677647 Points: 1 # Comments: 0

<pre style='white-space:pre-wrap;width:81ex'>fix(github-copilot): bound model discovery and embeddings JSON response (#96499) * fix(github-copilot): bound model discovery and embeddings JSON response reads The GitHub Copilot embeddings plugin already bounds its error response bodies via readResponseTextLimited, but the success JSON reads for both model discovery and the embeddings call used unbounded response.json(). Route both through readProviderJsonResponse (16 MiB cap). Update isCopilotSetup

A novel experiment tested how AI ran a store, and the ability of artificial intelligence to complete more important tasks.

It took 20 years, but the Finance app arrives just in time to be packed full of AI.

I benchmarked raw chat history, vector-only RAG, and a context graph on the same multi-agent conversations. The results exposed a surprising weakness in relational retrieval. The post Vector RAG Isn’t Enough — I Built a Context Graph Layer for Multi-Agent Memory appeared first on Towards Data Science .

I run a small self-hosted website on a Raspberry Pi 4B at home. A few weeks ago I started wondering: who actually visits a website in 2026? Not just humans. Everything. So I built a public observability dashboard on top of GoAccess that separates traffic into four categories: human visitors, search engine crawlers, AI retrieval agents, and automated attacks. The numbers from the last 17 days surprised me: 4,523 human visits 6,409 automated attack attempts Thousands of crawler requests from searc

<pre style='white-space:pre-wrap;width:81ex'>fix(speech): bound TTS/STT voice-list and transcription JSON response reads (#96496) Route success JSON reads through readProviderJsonResponse (16 MiB cap) in azure-speech, elevenlabs, microsoft, minimax/tts, xai/stt, and openrouter/media-understanding to prevent OOM from oversized or hostile endpoint responses. Mirrors the response-limit campaign already applied to other provider paths. AI-assisted. Co-authored-by: Cursor <cursoragent@cursor.com></pr

Article URL: https://techtrust.ai/ Comments URL: https://news.ycombinator.com/item?id=48677437 Points: 1 # Comments: 0

Not just VSCode, but other pieces of software I always use, not sure if that's related to AI, but this genuinely upsets me. I would expect things to be better as AI helps accelerate development and bug reports and whatnot, but it looks like the opposite is actually happening... which is concerning. Comments URL: https://news.ycombinator.com/item?id=48677369 Points: 2 # Comments: 2

Alibaba allegedly used 25,000 accounts to mine Claude over 28.8 million exchanges.

They're not lost, they just didn't have a chance. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

A reproducible benchmark on latency, cost, and reproducibility, and where agents actually earn their keep. The post The Hot Path Belongs to GBDTs, Agents Own the Cold Path: A Payment-Fraud Benchmark appeared first on Towards Data Science .

In this technical collaboration between AWS and the authors, we present a pragmatic solution: agentic overlays. Agentic overlays are thin wrapper layers that transform traditional REST-based services into agents capable of participating in A2A interactions. They also expose REST APIs as tools compatible with the Model Context Protocol (MCP). Together, they let enterprises add A2A capabilities to existing REST services without rewriting business logic, without duplicating code, and without runnin

<pre style='white-space:pre-wrap;width:81ex'>fix(github-copilot): bound usage response (#96607) The Copilot usage read in extensions/github-copilot/usage.ts parsed its HTTP response with an unbounded await res.json(). A hostile or buggy api.github.com proxy (the proxy endpoint is derived from a user-supplied token) could stream an unbounded JSON body and drive the usage snapshot into OOM. Route the read through the shared readProviderJsonResponse (from openclaw/plugin-sdk/provider-http), which e

Did Full Self-Driving (Supervised), Tesla’s driver assistance feature, play a role in a woman’s death?

Article URL: https://zamechek.com/blog/the-sieve-growing-from-a-seed/ Comments URL: https://news.ycombinator.com/item?id=48676951 Points: 2 # Comments: 1