AI News Today

Epic Games is making generative AI a big part of upcoming versions of Unreal Engine.

Article URL: https://github.com/acunningham-ship-it/leakproof Comments URL: https://news.ycombinator.com/item?id=48576074 Points: 1 # Comments: 1

Hey HackerNews! My name is Tyler and I'm the solo founder of Quanta Intellect. I'm excited to share an update to Vessel Browser with you today! You can now either import an existing skill or create a new one in the Skills tab. Then, invoke your skill in agent panel with a "/" command! Vessel Browser is open source and BYOK, allowing you to bring your existing AI stack with you. Thanks for checking it out! Comments URL: https://news.ycombinator.com/item?id=48576073 Points: 1 # Comments: 0

<p> Clone yourself. Let AI do the work before you ask </p> <p> <a href="https://www.producthunt.com/products/vida-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/1174638?app_id=339">Link</a> </p>

Archaeologists say the motivation is "very difficult to interpret". ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.

Article URL: https://github.com/bohdan1288-dotcom/ai-audit-orchestrator Comments URL: https://news.ycombinator.com/item?id=48575897 Points: 1 # Comments: 0

Article URL: https://factory.ai/product/router Comments URL: https://news.ycombinator.com/item?id=48575781 Points: 1 # Comments: 0

How GitHub Copilot is making more of each session go toward useful work, so your credits go further. The post Getting more from each token: How Copilot improves context handling and model routing appeared first on The GitHub Blog .

The Agents SDK is now a runtime any agent framework can build on. Today we're opening up the Agents SDK primitives, with Flue as a first framework targeting Agents SDK, and rolling out agents in the dashboard.

Researchers urge developers to see that less is more when it comes to instructions

Article URL: https://github.com/CEM888AI/CEM888.AI-Site Comments URL: https://news.ycombinator.com/item?id=48575589 Points: 1 # Comments: 0

Article URL: https://www.bbc.com/news/articles/cwy24v72n19o Comments URL: https://news.ycombinator.com/item?id=48575575 Points: 2 # Comments: 0

Article URL: https://retool.com/blog/retool-launches-react-ai-app-builder Comments URL: https://news.ycombinator.com/item?id=48575541 Points: 3 # Comments: 0

<!-- SC_OFF --><div class="md"><p>I'm not that well versed in ML yet. I know that "Attention is all you need" was based on work that was done with a couple of high end gaming GPUs at the time. I can afford that.</p> <p>Suppose for arguments sake that I have caught up on ML such that I have the competence to recreate state of the art results should I have access to the required hardware, do I still need access to huge amounts of hardware infrastructure to be able to contribute to the field at a f

NVIDIA’s self-improvement program for robots enlists teams of AI coding agents.

Article URL: https://www.browserless.io/blog/writing-an-ai-agent Comments URL: https://news.ycombinator.com/item?id=48575480 Points: 6 # Comments: 0

Researchers show that for certain kinds of games, an overlooked class of algorithms performs much better than expected.

An open specification for finding and verifying tools, skills, and agents across the web.Agents are ...

<p>Exposing a tool to an agent over MCP takes ten minutes. Building an MCP server that survives a model you don't control, on a tight token budget with limited thinking time, is the part nobody warns you about.</p> <p>We learned the difference shipping our own, consumed by third-party agents whose models we don't pick. Three principles came out of it, each one we only fully believed after it broke in production:</p> <p><strong>TL;DR — three MCP server best practices from our trenches:</strong></

Article URL: https://antonmks.substack.com/p/fast-food-and-ai Comments URL: https://news.ycombinator.com/item?id=48575284 Points: 2 # Comments: 0

<p>When we wrote code before large language models could write it for us, code duplication was rarely our problem, at least not if we worked alone or in a small team.</p> <p>That has changed.</p> <h2> A principle born in a quieter time </h2> <p>"Don't Repeat Yourself" was coined by Andy Hunt and Dave Thomas in The Pragmatic Programmer back in 1999, a year when our biggest worry was whether the clocks would survive January. The idea is reassuringly simple. Every piece of knowledge in your system

French President Macron and Indian PM Modi raised alarms at the G7 summit that the U.S. could cut off access to American AI overnight — a fear the Anthropic blackout just made real.

Anthropic's tools are getting chummy with each other.

When Anthropic quietly released Claude Design in April as a " research preview ," it generated the kind of instant traction most product teams dream about: more than one million users in its first week. It also generated a problem. The tool consumed tokens so voraciously that a PCWorld reviewer burned through 80 percent of his weekly Claude Pro allowance in roughly 25 minutes, producing just three variations of a single webpage prototype. "We're talking another token-hungry Claude product here,"

<p>Retrieval-Augmented Generation (RAG) is a powerful pattern to build applications that can query, understand, and extract insights from your custom documents (like PDFs, resumes, and reports) by feeding them as context to Large Language Models (LLMs).</p> <p>This guide walks you through building a complete RAG API step-by-step, explaining the architecture, code, and debugging learnings along the way.</p> <h2> 1. Architecture Overview </h2> <p>A typical RAG pipeline is divided into two parts:</

<p><em>Hello, I'm Maneshwar. I'm building git-lrc, a Micro AI code reviewer that runs on every commit. It is free and source-available on Github. <a href="https://github.com/HexmosTech/git-lrc" rel="noopener noreferrer">Star git-lrc</a> to help devs discover the project. Do give it a try and share your feedback.</em></p> <p>For most of the internet's life, searching for something felt a bit like talking to a very literal-minded robot. </p> <p>You'd type "comfy shoes for standing all day," and it

<p>I recently used Cursor Agent Mode with Auto Mode enabled to do something simple: recommend a font pairing and update two files in my project. An <code>index.html</code> and an <code>index.css</code>. That's it! </p> <p>The agent added a Google Fonts <code><link></code> tag, a <code><link></code> Gstatic with a crossorigin within it, and shuffled a few CSS variables as well as styling rules for the body tag and H tags. </p> <blockquote> <p>It cost 1% of my monthly usage budget.<br> <a href="ht

Article URL: https://fideai.org/research/fmg-bench/ Comments URL: https://news.ycombinator.com/item?id=48574941 Points: 2 # Comments: 2

Article URL: https://www.cnbc.com/2026/06/17/anthropic-amodei-google-hassabis-us-ai-coalition-g7.html Comments URL: https://news.ycombinator.com/item?id=48574918 Points: 2 # Comments: 0

With everything going on lately regarding AI policy, and the cat already being out of the bag with AI systems that can run on feasibly obtainable personal compute, at what point do things tip over governments confiscating said compute in order to enforce a certain regulatory goal? Today's regulation is mostly targeted at frontier labs, hyperscalers, chip makers, etc. But if compute is the choke point, and things are getting more efficient, where is the limiting principle "on the way down"? At wh

Article URL: https://rohangandhi.com/posts/how-to-stop-babysitting-ai-code/ Comments URL: https://news.ycombinator.com/item?id=48574893 Points: 1 # Comments: 0

A few useful links: - https://mhclgdigital.blog.gov.uk/2025/06/12/extract-using-ai-to-unlock-historic-planning-data/ - https://github.com/digital-land/digital-land/issues/360 - https://ai.gov.uk/evaluations/extract-evaluation/ Comments URL: https://news.ycombinator.com/item?id=48574862 Points: 1 # Comments: 0

A Microsoft researcher built a working neural network out of goats, bridges, and ice ramps in the Age of Empires II map editor. What looks like a joke is a pointed critique of AI research methods. His analysis of 315 papers found that more than half already assume language models have human-like traits before the experiment even starts. Replace the chat interface with wandering goats, and the math doesn't change, but the feeling that you're talking to someone does. The article Microsoft research

Article URL: https://www.flourish.org/2026/06/agree-everyone-ai/ Comments URL: https://news.ycombinator.com/item?id=48574778 Points: 1 # Comments: 0

I have running into the same problems recently instrumenting my LangChain and CrewAI agent pipelines. If something goes wrong with a tool call, there was no way to way to prove what the agent did, in what order, and whether the logs have been modified. Observability platforms like LangSmith and Langfuse are great at optimizing for agent behavior, tokens and costs but they do not produce legally defensible and auditable artifacts. That's the reason why I decided to build RootSign, an SDK that ins

Article URL: https://ir.smartbird.ai/news-releases/news-release-details/smartbird-appoints-new-ceo-advance-ai-infrastructure-strategy Comments URL: https://news.ycombinator.com/item?id=48574698 Points: 2 # Comments: 0

A prototype wiring system for dilution refrigerators could advance the realization of practical quantum computers.

We built Else, a tool for PMs to prototype on top of their actual product. I started using Else to build Else, and I now inadvertently spend all my time turning it into my ideal dev environment. Else works similar to most cloud agent systems with some key differences that I personally think have a huge impact, but I’m worried I have tunnel vision. WHY DO I THINK IT'S SO GOOD? It’s equally good at the first 80% (the agent spends an hour building the feature) and the last 20% (you need to test, it

<!-- SC_OFF --><div class="md"><p>Hi All, I've been running experiments on targeted SFT for specific capability dimensions on a 31B model. After running small training run to prime the model slightly in the direction I want, then ran a judge across 40 domains scoring six independent quality dimensions. One dimension consistently scored weakest across five runs.</p> <p>I am now training contrastive variants from the same checkpoint - examples with that dimension deep vs examples with it deliberat

Anthropic has joined the Frontier coalition, which received another $915M in pledges to fund carbon removal projects.