AI News Today

arXiv:2608.14644v1 Announce Type: new Abstract: Real-world LLM deployments increasingly rely on runtime-injected prohibitions--enterprise policies, PII redlines, tool boundaries--that vary per request and per tenant. Conventional post-training is structurally ill-suited: SFT hides the violation signal in compliant labels, and DPO's sequence-level preferences mismatch token-localized violations. We propose DUET, a token-selective on-policy distillation method for prohibition compliance. DUET pair

arXiv:2608.14645v1 Announce Type: new Abstract: Studying climate change requires reducing uncertainties in CO2 and CH4 emission estimates to better distinguish anthropogenic from natural sources, which motivates spaceborne measurements with improved revisit frequency and spatial coverage. In this context, the Horizon Europe SCARBOn project assesses a low-cost satellite constellation featuring the NanoCarb imaging interferometer as its core sensor for monitoring CO2 and CH4 emissions in the atmos

arXiv:2608.14646v1 Announce Type: new Abstract: Interpretable representation learning remains a key challenge in modern neural computation, particularly when models are expected not only to perform but also to explain their reasoning. This paper introduces iFuzz-Meta, an interpretable fuzzy rule-based learning framework that preserves human-understandable reasoning structures within modern neural architectures. Each fuzzy rule corresponds to a semantic and spatial prototype defined in the origin

arXiv:2608.14649v1 Announce Type: new Abstract: We present dLLM-SetScore, a training-free method that uses discrete masked-diffusion language models for multi-label text classification. For each candidate label, it asks a short yes/no question and compares the probabilities of the two answer tokens at one masked position. The method uses no task-specific fine-tuning or training on textual-entailment datasets; a 200-example labelled validation slice selects thresholds, temperature, and prompt wor

arXiv:2608.14650v1 Announce Type: new Abstract: Existing adaptive-inference and world-action-model systems use cheap-stage outputs or predicted futures to allocate additional computation. We study a narrower question: under paired exact-reset physical outcomes, can a Medium-derived interface predict when switching to a separately frozen Full predictor improves task-specific decision loss enough to justify sequential overhead? Our contribution is a paired evaluation and audit protocol, not a new

arXiv:2608.14652v1 Announce Type: new Abstract: The development of 0.1$^{\circ}$ global weather forecasting models based on machine learning (ML) is constrained by the limited availability of high-resolution data, as decades of reanalysis are only available at 0.25$^{\circ}$ resolution. While existing approaches fine-tune 0.25$^{\circ}$ forecast models on limited 0.1$^{\circ}$ samples, we show that this transfer is hindered by the irreversible information loss inherent in coarse-resolution forec

arXiv:2608.14653v1 Announce Type: new Abstract: Prediction uncertainty is a widely adopted metric for quantifying model confidence, with downstream applications spanning model explanation, data selection, and prediction rollback. Despite its demonstrated utility, the potential of uncertainty quantification to enhance code generation in large language models (LLMs) remains largely underexplored, raising a critical question: to what extent can uncertainty serve as an effective signal for improving

You can now provision and manage Algolia directly from the Vercel Marketplace. Algolia is a hosted search platform. You send it your content, it builds an index, and your app queries that index instead of your database. Results come back in milliseconds, with typo tolerance, synonyms, and facet counts built in. Key capabilities: Index product catalogs, documentation, or any other content Add recommendations and AI ranking to surface the most relevant results Point the Algolia Crawler at a Vercel

Article URL: https://serplens.com Comments URL: https://news.ycombinator.com/item?id=49356542 Points: 1 # Comments: 0

Article URL: https://www.jitbit.com/alexblog/ai-agents-and-3d-printers/ Comments URL: https://news.ycombinator.com/item?id=49356461 Points: 2 # Comments: 0

I built Relay around a simple idea: many of us have an unused PC server at home, or a VPS dedicated to AI-assisted coding, but the coding agents running there are still tied to that machine’s terminal, I just don't want to ssh/rdp into it every single time. Relay brings Claude Code, Codex, OpenCode, and Hermes into one interface that you can access from your phone, browser, or another computer. Sessions stay alive, so you can start work on one device and continue from another. You can upload/dow

Article URL: https://recursive-economy.pages.dev/ Comments URL: https://news.ycombinator.com/item?id=49356213 Points: 2 # Comments: 0

Article URL: https://aaif.io/blog/a2a-joins-aaif Comments URL: https://news.ycombinator.com/item?id=49356130 Points: 1 # Comments: 0

Article URL: https://jamanetwork.com/journals/jama/fullarticle/2852952?guestAccessKey=562af865-a038-4425-993f-ff25e5e3a130 Comments URL: https://news.ycombinator.com/item?id=49356100 Points: 1 # Comments: 0

Article URL: https://warpdd.com Comments URL: https://news.ycombinator.com/item?id=49355834 Points: 3 # Comments: 0

Article URL: https://endstatefallacy.com/ Comments URL: https://news.ycombinator.com/item?id=49355716 Points: 1 # Comments: 0

Article URL: https://www.youtube.com/watch?v=xH7U7w9Qzlo Comments URL: https://news.ycombinator.com/item?id=49355702 Points: 2 # Comments: 0

Article URL: https://www.npr.org/2026/08/18/nx-s1-5929575/ai-suicide-risks-mental-health Comments URL: https://news.ycombinator.com/item?id=49355684 Points: 10 # Comments: 2

After a stunning debut last week with cyber capabilities so advanced they reportedly found a previously undetected vulnerability in Cursor, GLM-5.3, the new frontier open source language model from Chinese startup z.ai, has now hit the application programming interface (API) — allowing developers the ability to build atop it and plug it into their agents and applications. Developers who previously subscribed to a GLM Coding Plan are currently limited to the OpenAI Chat Completions-compatible pro

And everything else I learned about productivity this year

Article URL: https://www.cerebras.ai/cs4 Comments URL: https://news.ycombinator.com/item?id=49354949 Points: 272 # Comments: 192

Large language models (LLMs) exhibit a wide range of human-like behaviors, from expressing thoughts and emotions, to engaging in relationship-building with users, to refusing requests and maintaining boundaries. Despite their prevalence, researchers and practitioners lack methods and empirical insights to make informed decisions about when and what types of human-like behaviors LLMs should exhibit. To fill this gap, we present a multi-dimensional analysis of the prevalence, potential effects, an

Modern AI agents increasingly rely on search infrastructure to execute complex, neuro-symbolic reasoning workflows. These workflows often compile into deeply nested, non-monotonic Boolean queries over text fields. However, standard query evaluation strategies over inverted indices face severe theoretical limits when handling these structures. Stateful iterator models (Document-at-a-Time) are structurally bounded by NC^1 formula evaluation, suffering a worst-case O(2^|Q|) exponential blowup in qu

Vercel Agent now works in Slack. Mention it the way you'd pull a teammate into a thread, and it reads the discussion, answers with the context of the platform running your app, and turns the team's decisions into changes you approve. Vercel for Slack is available today in Public Beta for Pro and Enterprise teams. Work starts in conversation . An alert gets noticed in a channel, a fix gets agreed on in a thread, and then someone leaves Slack to do the work. With Vercel Agent, you can tag it in fo

Fish Audio 's audio models are now available on AI Gateway. To celebrate the launch, every Fish Audio model is free on AI Gateway for the next 30 days, through September 18. Capability Regular Through September 18 Text-to-speech $15.00 per million characters Free Speech-to-text $0.36 per hour of audio Free Four models from Fish Audio are available, including their latest text-to-speech model: fish-audio/s2.1-pro (text-to-speech): Built for low-latency streaming; clones a voice from a reference r

Vercel Agent is now available in Slack. Mention @Vercel in any channel or thread, and Agent joins the conversation with relevant context from your Vercel projects, including deployments, builds, logs, metrics, configuration, and connected repos. For example, use it to: Investigate incidents: Ask why a deployment failed or what’s causing a runtime error. It investigates and explains what it found. Write and review code: Work with your team and Agent to fix build and CI failures, review pull reque

Vercel Connect now includes a managed connector for Microsoft , so your apps and agents can access all Microsoft products such as Teams, OneDrive and more. As a Vercel Managed Connector , Vercel registers the Entra application and configures federated credentials for it, so you don't have to create an app registration or manage client secrets yourself. Authorize the app for your Microsoft tenant once, then attach the connector to the projects and environments it needs. Create a connector from th

Article URL: https://modelmap.cc Comments URL: https://news.ycombinator.com/item?id=49354664 Points: 79 # Comments: 7

<p> Create design systems, build with agents </p> <p> <a href="https://www.producthunt.com/products/balsa-ui?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/1226289?app_id=339">Link</a> </p>

<p> AI agents that run in a loop </p> <p> <a href="https://www.producthunt.com/products/cronloop-ai?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/1226287?app_id=339">Link</a> </p>

<p> Weird, playful desktop app for building with AI agents </p> <p> <a href="https://www.producthunt.com/products/berd?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/1226285?app_id=339">Link</a> </p>

<p> Manage your CMS content from Claude, Cursor, and Codex </p> <p> <a href="https://www.producthunt.com/products/marble-7?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/1226267?app_id=339">Link</a> </p>

Cursor, known for its AI Code Editor, is launching a new code-hosting platform to rival developers' long preferred favorite, GitHub.

Article URL: https://linear.app/data Comments URL: https://news.ycombinator.com/item?id=49353432 Points: 137 # Comments: 70

ChatGPT Ads is expanding to 31 European markets. Learn how advertisers can reach people as they explore, compare options, and make decisions.

Learn how to use Claude and ChatGPT for project planning, scheduling, risk management and other project management tasks for $19.99.

Kollmorgen discusses the importance of fitting voltage systems closely to the application while still making the best use of infrastructure. The post How to achieve high voltage in industrial systems without high complexity appeared first on The Robot Report .

<p>Amid race with Anthropic, firm plans to overhaul research and training and require more safety parameters after hack</p><p><a href="https://www.theguardian.com/technology/openai">OpenAI</a> on ⁠Tuesday said it had slowed down the ⁠pace of ⁠its ​AI development while it overhauled its ⁠research and training systems.</p><p>The company’s researchers were ⁠caught unaware last month ​when an ‌AI agent ‌under testing <a href="https://www.theguardian.com/technology/2026/jul/22/openai-says-its-models-

I am grateful that Anthropic is producing periodic Risk Reports.