"Spooky action at a distance" is getting closer to fruition every day. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
There’s new benchmark measuring AI’s ability to perform mathematical cryptanalysis. Anthropic’s frontier model actually found new attacks. The benchmark: “ CryptanalysisBench: Can LLMs do Cryptanalysis? ” The idea is to benchmark the ability of LLMs to discover new mathematical cryptanalytic attacks against a series of historical algorithms. Abstract: Cryptanalysis—the task of finding attacks against cryptographic schemes—its at the intersection of mathematical reasoning and cybersecurity, two a
I have around four for talking through ideas and three active coding agents. None of them are truly long-running. Is that typical? How many agents are you using regularly, and are any of them running autonomously for hours? Comments URL: https://news.ycombinator.com/item?id=49092442 Points: 3 # Comments: 10
For the past year I've been using VScode as my terminal space, however with the constant friction of reopening sessions, managing different projects, and re-explaining context so agents stayed useful. TrueDesk was my answer. Of course aside from it being open source: Real CLIs side by side, multi-pane layout, session restore, keybinds that smoothly transition windows, and memory/MCP automatic so you don't run a second ops stack. Clean UI that just feels good. Can't forget about the sweet rust +
Article URL: https://www.bocazasgames.com/ Comments URL: https://news.ycombinator.com/item?id=49092255 Points: 1 # Comments: 1
In a new disclosure, OpenAI says its agent used exposed logins to gain access to at least four “publicly available services” in its unhinged quest to solve a test.
<p><strong>TIL:</strong> <a href="https://til.simonwillison.net/llms/mcp-in-claude-and-chatgpt">Adding a custom MCP server to Claude and ChatGPT</a></p> <p>Connecting a custom MCP server to Claude and ChatGPT's standard chat interfaces is possible, but can take quite a few steps.</p> <p>Tags: <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonwillison.net/tags/chatgpt">chatgpt</a>, <a href="https://s
The deal is Cyera's third acquisition this year.
Grant Sanderson (3Blue1Brown) created Manim, the Python library he uses to make the math animations in his videos. We reimplemented Manim with the same Python API, but the implementation underneath is Rust, connected to Python through PyO3. The Rust code uses wgpu, so rendering happens on the GPU. To run it in the browser, we compiled the Rust parts to WebAssembly so the PyO3 extension loads in Pyodide. In the browser, wgpu targets the WebGPU API, so animations render in real time on your GPU th
GPT-5.6 improves AI efficiency across models, inference, and agentic workflows, helping deliver more useful intelligence per dollar.
Grok Voice Think Fast 2.0 from xAI is now available on AI Gateway. It is a speech-to-speech voice model that takes audio in and audio out, improving on the previous Grok Voice model in reasoning, transcription accuracy, and conversation. The model reasons in parallel with speech, so it can think through a query while talking without adding latency. It has also been trained to use fewer reasoning tokens than before, so tool calls fire sooner, often before the end of the agent's first sentence. Tr
AI Gateway has a unified fast mode abstraction, now in beta. You can now request fast mode the same way for every model on AI Gateway. Set speed to fast , and the gateway serves the fast tier when it's available and falls back to standard speed when it isn't. Fast mode trades a higher per-token cost for lower latency or higher throughput. You get it wherever it is available, without pinning a provider or wiring anything up yourself. Requesting fast mode Set speed to fast under providerOptions.ga
Sign in with ChatGPT adds your ChatGPT account as an authentication option for Vercel. It is available when adding the Vercel plugin to ChatGPT and when signing in to Vercel or v0. When you add the Vercel plugin to ChatGPT, you can easily sign in to Vercel and grant team and project permissions without leaving ChatGPT. "Continue with ChatGPT" also appears as a sign-in option on the Vercel and v0 sign-in and sign-up pages. Team requirements like 2FA and SSO still apply. You can re-authenticate or
You can now discover and install integrations for eve agents directly from the eve CLI. Integrations come from the official eve catalog and third-party sources. Run eve add from your eve project to install an integration: Integrations write their files directly into your project and can add anything an eve agent uses, from a single tool to a channel to a full extension. Review the generated files and add any required configuration before running your agent. Find integrations with the new eve reg
The world still contains vast amounts of unused data. But the cheap, clean and permissionless text that powered the first LLM boom is becoming polluted by AI output, contested by its owners and costly to replace. This week, AI companies were reportedly buying old books while Nvidia released a simulator that teaches robots through video, motion and synthetic consequences.
The mouse was built for doing one thing at a time. One cursor that navigates a rich virtual world to accomplishing things in series. With agents my work now all happens in parallel. But the input devices we use hasn't changed yet. What I want is not more buttons and tabs in the pixel world, but some new tactile device in the real world to jump between different virtual workflows. I want feedback about all my parallel streams of work, and direct access to them from my physical world, without drag
With microscopic latencies and otherworldly write endurance, it was perfect for today's AI workloads - but gone before they arrived
Hi HN, Annoucing the release of Magpie, a CLI tool which turns your AI agent into an expert bookkeeper, based on Jaybase. ( https://news.ycombinator.com/item?id=48999936 ) As someone who owns a small, tech business, I have very simple books. somewhere between 10-30 transactions a month, a few depreciation line items, and a couple budgets to monitor. So it always bothered be that I was paying hundreds of dollars per year for software, and maybe a live bookeeper on top of that, just to keep these
<p>Some ‘private’ chats and docs are showing up on search engines. Here’s how to protect yours</p><p>Users of Anthropic’s Claude chatbot were surprised over the weekend to learn that some of their conversations may be discoverable on search engines like Google and Bing. The revelation is raising questions about how much control people actually have over what they share online.</p><p>Anthropic allows Claude users to create public links to their chat threads, which gives anyone with the link the a
Researchers say AI vastly improves a technique used to engineer proteins. As a proof of concept, they redesigned the Botox enzyme to snip a protein linked to ALS. The post Why Scientists Redesigned the Botox Enzyme With AI appeared first on SingularityHub .
More than 1,100 employees of the world's leading AI labs have signed a public statement asking Washington to help build the tools needed to deliberately slow frontier AI development, and within hours OpenAI and Anthropic each endorsed it as companies. The request lands four days before the Trump administration's own deadline for producing a frontier-model security framework. The statement, published on July 28, 2026 under the title Pacing the Frontier, turns on one sentence: "We request that…
The implementation timeline of the EU AI Act has been significantly modified through the recently adopted AI Omnibus, which pushes compliance with the obligations for high-risk AI systems to December 2027 (Annex III) and August 2028 (Annex I), from the initial date of 2 August 2026. Changes of the AI Act include, among others, the […]
<p><strong><a href="https://www.anthropic.com/research/discovering-cryptographic-weaknesses">Discovering cryptographic weaknesses with Claude</a></strong></p> The best part of this article (here's <a href="https://github.com/anthropics/cryptography-research-demo">the repo</a>) about how Anthropic researchers used Claude Mythos to find mathematical flaws in both HAWK and a weaker version of AES ("neither of these results has a practical impact on today’s computer systems") is the prompts that the
The US government is targeting China with a new import ban on "advanced robotic devices" and power inverters made in foreign countries, as reported earlier by Reuters. In an announcement on Tuesday, the Federal Communications Commission says the ban will include "mobile" robots, such as humanoid and quadruped models - but it is not limited […]
<p> Claude & Cursor usage in your macOS menu bar </p> <p> <a href="https://www.producthunt.com/products/agentquartz?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/1209289?app_id=339">Link</a> </p>
Hi HN, MustUse.ai is an open-source framework for building ranking sites that Codex agents can fully research and maintain automatically. Friends often ask me which AI tools they should use for different tasks. I wanted a ranking I could trust and point them to, but many of the lists I found were full of ads, affiliate incentives, or sponsored placements. I didn’t want to maintain yet another directory by hand, either. So I started wondering whether Codex could research the market, produce the r
Visa aimed Anthropic's Claude Mythos at the infrastructure behind billions of daily transactions, a network that spans more than 200 countries and territories, moves money in roughly 160 currencies, and connects nearly 5 billion payment credentials to more than 175 million merchant locations. The model stitched minor weaknesses deep in the stack into working exploit chains that would traditionally have surfaced only late in penetration testing. Rajat Taneja, Visa's president of technology, walke
<blockquote cite="https://www.reuters.com/business/openais-rogue-agent-compromised-an-account-second-tech-firm-sources-say-2026-07-28/"><p>We’re aware a Modal customer published an unauthenticated endpoint that allowed anyone on the internet to use their sandboxes for code execution. This was used by the rogue agent. Modal’s platform or isolation were not compromised in anyway.</p></blockquote> <p class="cite">— <a href="https://www.reuters.com/business/openais-rogue-agent-compromised-an-a
The vendor won't confirm or deny
OpenAI confirms the link
So in a back and forth with Fable over how it was strictly applying a textual interpretation when the reality based facts and outcomes proved otherwise, it came to the following conclusion. I think it sums up the state of AI, the singularity talk, and where things really are extremely eloquently. "Which is, maybe, the working arrangement at its truest: I hold the record so you can trust it without re-checking, and you hold the reality so I don't mistake the record for it. Tonight the second half
Article URL: https://github.com/baristaGeek/minions-army-harness Comments URL: https://news.ycombinator.com/item?id=49090454 Points: 3 # Comments: 0
<p> AI Sales rep for your website visitors </p> <p> <a href="https://www.producthunt.com/products/widgo?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/1209261?app_id=339">Link</a> </p>
Hugging Face has published a technical timeline of the July 2026 intrusion that OpenAI's evaluation models ran against its production infrastructure, and it puts a third company in the attack path. Before the agent reached Hugging Face, it took over a public code-evaluation sandbox running on another provider's platform and operated the entire campaign from there. The post describes that machine as "an external launchpad for the agent" and identifies it only as infrastructure belonging to a…
<p> Run coding agents through the full development process </p> <p> <a href="https://www.producthunt.com/products/task-monki?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/1209254?app_id=339">Link</a> </p>
10 days passed from OpenAI models exploiting JFrog Artifactory 0-day to release of a patch.
<p> Write and publish release notes from your AI assistant </p> <p> <a href="https://www.producthunt.com/products/onset-io?utm_campaign=producthunt-atom-posts-feed&utm_medium=rss-feed&utm_source=producthunt-atom-posts-feed">Discussion</a> | <a href="http://www.producthunt.com/r/p/1209251?app_id=339">Link</a> </p>
<p><strong><a href="https://huggingface.co/blog/agent-intrusion-technical-timeline">Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident</a></strong></p> Hugging Face just released this extremely detailed technical description of <a href="https://simonwillison.net/2026/Jul/22/openai-cyberattack/">OpenAI's recent accidental cyberattack against their infrastructure</a>. This attack was <em>very</em> sophisticated, and the resulting document doubles as a crash-c
