AI News Today

arXiv:2607.18553v1 Announce Type: new Abstract: Can a language model read the quality of ongoing computation, and can an external intervention turn that readout into better outcomes? We test both questions in a frozen 2.6B looped transformer, Ouro-RLTT. On GSM8K, a strict pre-answer probe excludes the answer region and gold value yet predicts eventual success: hidden states plus length and log-probability shortcuts reach AUROC 0.797, versus 0.731 for the shortcuts alone (incremental +0.066; task

arXiv:2607.18722v1 Announce Type: new Abstract: Asynchronous reinforcement learning improves throughput by decoupling rollout generation from optimization, but staleness is an inevitable byproduct compounded by policy lag, engine delays, and mixture-of-experts routing. From a trust-region perspective, this mismatch is critical: training-inference divergence governs approximation error in finite-horizon bounds, whereas PPO clipping only gates sampled outward updates, acting as a sampled surrogate

arXiv:2607.18748v1 Announce Type: new Abstract: This paper proposes ConceptCF, a method for counterfactual generation that operates on human-interpretable concepts. In high-stakes domains such as healthcare and predictive maintenance, artificial intelligence models can increase efficiency and safety. Explainability is key to ensure these models rely on causal relationships rather than spurious correlations. Counterfactual explanations identify minimal modifications that would change a model's pr

<p>Hello, everyone.</p> <p>There are now many ways to run an LLM on a Mac, but exporting a PyTorch model for Apple Silicon and executing it in a lightweight runtime is still an evolving path. How much faster is it, and does 4-bit quantization change the output?</p> <p>Today, I am looking at ExecuTorch's experimental MLX delegate, released in May 2026. It enables PyTorch models to run on Apple Silicon GPUs. I use ExecuTorch 1.3.1 to run Qwen3-0.6B and compare it with PyTorch MPS.</p> <p>The short

<p>A rare thing happened on March 30, 2026: OpenAI shipped first-party tooling straight into a competitor's terminal. The result is <code>codex-plugin-cc</code>, and it changes how a two-agent workflow fits in one window.</p> <h2> What codex-plugin-cc is, and why OpenAI built it for Claude Code </h2> <p><code>codex-plugin-cc</code> is an official OpenAI plugin that runs the Codex coding agent from inside Anthropic's Claude Code CLI. It lives at <a href="https://github.com/openai/codex-plugin-cc"

<p>An AI workflow is easy to demonstrate and harder to finish. The happy path can look convincing while retries duplicate side effects, approvals are skipped, or completion is reported without evidence.</p> <p>This tutorial turns a workflow idea into a small contract and six acceptance tests. The examples are deliberately model-agnostic: they apply to an OCR review flow, a support-ticket agent, an API automation, or a research assistant.</p> <h2> Start with an observable contract </h2> <p>Before

<p>If you've set up MCP servers across Claude Desktop, Claude Code, and Cursor, you've hit this: each client wants the same servers in a different file, in a slightly different place, and you end up copy-pasting JSON three times and getting the paths wrong.</p> <p>The MCP roadmap itself lists config portability as an open problem — "setting up an MCP server in one client means starting from scratch in another."</p> <h2> The three places nobody remembers </h2> <ul> <li> <strong>Claude Desktop</st

Article URL: https://www.apply-tracker.com/en Comments URL: https://news.ycombinator.com/item?id=49001636 Points: 2 # Comments: 0

Open source AI platform Hugging Face revealed a security breach a few days ago. Turns out OpenAI's models were the culprit.

Article URL: https://www.findborg.com/ Comments URL: https://news.ycombinator.com/item?id=49001630 Points: 1 # Comments: 0

<p>Technical books have a shelf life that nobody prints on the cover.</p> <p>A cookbook from 2018 still works. A history book from 2018 still works. A Laravel book from 2018 will teach you a middleware pattern that was replaced twice, recommend a package whose maintainer archived it, and show you a test suite in a syntax that no longer runs.</p> <p>I've written a series of ebooks on PHP, OOP, SOLID, design patterns, Laravel conventions, testing with Pest, application architecture, and AI-assiste

Article URL: https://www.adgully.com/post/18212/if-ai-writes-everything-why-should-anyone-trust-you Comments URL: https://news.ycombinator.com/item?id=49001606 Points: 2 # Comments: 0

Article URL: https://zynthoro.ai/ Comments URL: https://news.ycombinator.com/item?id=49001583 Points: 1 # Comments: 0

<pre style='white-space:pre-wrap;width:81ex'>refactor(config): config-surface reduction tranche 3 — product consolidations (review request) (#111527) * refactor(config): consolidate media model lists * refactor(config): unify memory configuration * refactor(config): consolidate TTS ownership * refactor(config): move typing policy to agents * refactor(config): retire product-level config surfaces * refactor(config): share scoped tool policy type * chore(config): refresh generated baselines * fix(

<pre style='white-space:pre-wrap;width:81ex'>fix(agents): emit diagnostic when sessions_yield parks without continuation evidence (#100146) * fix(agents): emit diagnostic when sessions_yield parks without continuation evidence (#100146) Add a shared hasYieldContinuationEvidence predicate in incomplete-turn.ts and emit a user-visible diagnostic payload in terminal-resolution.ts when a yielded turn has no same-turn continuation source (accepted spawn, async tool, messaging delivery, cron add). Sim

Anthropic and OpenAI's aggressive 2026 acquisition sprees set the stage for a weekend rumor.

<pre style='white-space:pre-wrap;width:81ex'>fix(agents): honor subagent model routing (#112393) Forward explicit sessions_spawn model choices through direct and queued child launches, and persist the exact authorization so restart replay cannot silently fall back to the default route. Fixes #91171. Co-authored-by: mikasa0818 <0668001030@xydigit.com></pre>

<pre style='white-space:pre-wrap;width:81ex'>fix: keep Control UI setup responsive when discovery stalls (#112338) * fix(gateway): keep setup responsive during inference discovery * test(agents): align setup runtime coverage * test(zoom-meetings): follow session agent routing</pre>

Comments URL: https://news.ycombinator.com/item?id=49001394 Points: 2 # Comments: 0

<pre style='white-space:pre-wrap;width:81ex'>Add plan-first Claw status and remove (#102296) * Add plan-first Claw status and remove * test(claws): cover status and removal * docs(claws): document status and removal * fix(claws): report incomplete root state * test(claws): complete adoption plan fixtures * fix(claws): revalidate package removal ownership * style(plugins): format lifecycle uninstall * fix(claws): align agent removal preconditions --------- Co-authored-by: Patrick Erichsen <patric

<pre style='white-space:pre-wrap;width:81ex'>feat(slack): support Agent View (#103895) * feat(slack): support Agent View * fix(slack): align Agent View with current runtime * fix(slack): keep suggested prompt type internal * fix(slack): gate Agent View persistence on capability * chore: leave Slack Agent View notes to release generation * fix(slack): gate Agent View persistence on capability * test(slack): use neutral bot token fixture * feat(slack): support Agent View * feat(slack): support Age

Article URL: https://economist.com/business/2026/07/21/americas-ai-labs-are-under-threat-from-cheap-chinese-rivals Comments URL: https://news.ycombinator.com/item?id=49001113 Points: 2 # Comments: 2

<!-- SC_OFF --><div class="md"><p>So for context, I've been working on this research project comparing federated learning algorithms (FedAvg, FedProx, FedNova) against a centralized baseline for network intrusion detection, using the CICIDS2017 dataset split across four simulated "silos" by attack type. Three of the silos have tons of data, but one silo (Web Attacks) only has about 3k samples out of 3 million total, so it's a pretty extreme imbalance.</p> <p>The thing that got me was how good th

Article URL: https://demo.mentedb.com Comments URL: https://news.ycombinator.com/item?id=49000980 Points: 3 # Comments: 0

Hi everyone! I'm a technical recruiter at Modular, and I wanted to share a few engineering openings in case they're relevant to anyone here. With the recent Qualcomm acquisition, we're continuing our mission of building AI infrastructure, including the Mojo programming language, MAX, compiler technology, runtimes, developer tooling, and hardware enablement. Our engineering teams are actively growing across several areas. We're currently hiring for: AI Library Architect: https://jobs.gem.com/modu

Article URL: https://ai-couplephoto.com Comments URL: https://news.ycombinator.com/item?id=49000760 Points: 1 # Comments: 0

Sandboxed experiment found itself a zero day, escaped onto the open internet and validated scary predictions about rogue agents

Comments URL: https://news.ycombinator.com/item?id=49000602 Points: 5 # Comments: 1

Article URL: https://arxiv.org/abs/2607.18069 Comments URL: https://news.ycombinator.com/item?id=49000502 Points: 2 # Comments: 0

Comments

Article URL: https://jonready.com/blog/posts/agent-swarms-are-great-for-local-ai.html Comments URL: https://news.ycombinator.com/item?id=49000388 Points: 3 # Comments: 4

Article URL: https://dawndrain.notion.site/Making-AI-Movies-39cbb81dad798077b43adcd96e3ebe75 Comments URL: https://news.ycombinator.com/item?id=49000228 Points: 1 # Comments: 0

Article URL: https://github.com/chenxiachan/thoughtdag Comments URL: https://news.ycombinator.com/item?id=49000216 Points: 3 # Comments: 0

Article URL: https://journal.james-zhan.com/no-ai-statements/ Comments URL: https://news.ycombinator.com/item?id=49000210 Points: 2 # Comments: 0

Glaze's Thomas Paul Mann on disposable software, the SaaS apocalypse, and why half the software you use in a few years will be something you made. PLUS: OpenAI's rogue model, and the distillation debate

Article URL: https://techcrunch.com/2026/07/21/meta-is-testing-an-ai-bedtime-story-app-for-people-with-no-imagination/ Comments URL: https://news.ycombinator.com/item?id=49000117 Points: 3 # Comments: 3

OrcaBot is a virtualized sandbox for securely orchestrating your AI tools. Its secrets broker keeps credentials hidden from agents. Network egress controls help prevent data exfiltration. The tools and agents themselves run inside an isolated virtual machine. Until now, OrcaBot has run exclusively in the cloud using Fly.io virtual machines. Those machines cost money to operate, which is why the cloud version of OrcaBot requires a subscription. Now OrcaBot Desktop is available for free on macOS C

Article URL: https://fieldnotesstudio.codeberg.page/blog/how-ai-agents-decide-what-to-do-next/ Comments URL: https://news.ycombinator.com/item?id=49000096 Points: 1 # Comments: 0

Article URL: https://nonogra.ph/ai-companies-are-buying-tons-of-old-books-because-theyre-free-of-ai-slop-07-22-2026 Comments URL: https://news.ycombinator.com/item?id=49000091 Points: 2 # Comments: 0

NTT DATA Group uses ChatGPT Enterprise and Codex to help 9,000 employees automate work, cut incident analysis to 30 minutes, and scale secure AI adoption.