- Introducing Claude for Small Business
anthropic-news· 13-may
May 13, 2026Announcements
- PwC is deploying Claude to build technology, execute deals, and reinvent enterprise functions for clients
anthropic-news
Announcements
- Anthropic forms $200 million partnership with the Gates Foundation
anthropic-news· 14-may
May 14, 2026Announcements
- Revealing Interpretable Failure Modes of VLMs
arxiv-ai· 15-may
arXiv:2605.12674v1 Announce Type: new Abstract: Vision-Language Models (VLMs) are increasingly used in safety-critical applications because of their broad reasoning capabilities and ability to generalize with minimal task-specific engineering. Despite these advantages, they can exhibit catastrophic failures in specific real-world situations, constituting failure modes. We introduce REVELIO, a framework for systematically uncovering interpretable failure modes in VLMs. We define a failure mode as a composition of interpretable, domain-relevant concepts-such as pedestrian proximity or adverse weather conditions-under which a target VLM consistently behaves incorrectly. Identifying such failures requires searching over an exponentially large discrete combinatorial space. To address this chal…
- Do Androids Dream of Breaking the Game? Systematically Auditing AI Agent Benchmarks with BenchJack
arxiv-ai· 15-may
arXiv:2605.12673v1 Announce Type: new Abstract: Agent benchmarks have become the de facto measure of frontier AI competence, guiding model selection, investment, and deployment. However, reward hacking, where agents maximize a score without performing the intended task, emerges spontaneously in frontier models without overfitting. We argue that benchmarks must be secure by design. From past incidents of reward hacks, we derive a taxonomy of eight recurring flaw patterns and compile them into the Agent-Eval Checklist for benchmark designers. We condense the insights into BenchJack, an automated red-teaming system that drives coding agents to audit benchmarks and identify possible reward-hacking exploits in a clairvoyant manner. Moreover, we extend BenchJack to an iterative generative-adver…
- Macro-Action Based Multi-Agent Instruction Following through Value Cancellation
arxiv-ai· 15-may
arXiv:2605.12655v1 Announce Type: new Abstract: Multi-agent reinforcement learning (MARL) in real-world use cases may need to adapt to external natural language instructions that interrupt ongoing behavior and conflict with long-horizon objectives. However, conditioning rewards on instructions introduces a fundamental failure mode as Bellman updates couple value estimates across instruction contexts, leading to inconsistent values when instructions interrupt macro-actions. We propose Macro-Action Value Correction for Instruction Compliance (MAVIC), which corrects Bellman backups at instruction boundaries by correcting the incoming instruction objective and restoring the continuation value under the current objective. Unlike reward shaping, MAVIC modifies the bootstrapping target itself, e…
- DisaBench: A Participatory Evaluation Framework for Disability Harms in Language Models
arxiv-ai· 15-may
arXiv:2605.12702v1 Announce Type: new Abstract: General-purpose safety benchmarks for large language models do not adequately evaluate disability-related harms. We introduce DisaBench: a taxonomy of twelve disability harm categories co-created with people with disabilities and red teaming experts, a taxonomy-driven evaluation methodology that pairs benign and adversarial prompts across seven life domains, and a dataset of 175 prompts with human-annotated labels on 525 prompt-response pairs. Annotation by four evaluators with lived disability experience reveals three findings: harm rates vary sharply by disability type and will compound in non-text modalities, terminology-driven harm is culturally and temporally bound rather than universally assessable, and standard safety evaluation catch…
- CHAL: Council of Hierarchical Agentic Language
arxiv-ai· 15-may
arXiv:2605.12718v1 Announce Type: new Abstract: Multi-agent debate has emerged as a promising approach for improving LLM reasoning on ground-truth tasks, yet current methodologies face certain structural limitations: debate tends to induce a martingale over belief trajectories, majority voting accounts for most observed gains, and LLMs exhibit confidence escalation rather than calibration across rounds. We argue that the genuine value of debate, and dialectic systems as a whole, lies not in ground-truth tasks but in defeasible domains, where every position can in principle be defeated by better reasoning. We present the Council of Hierarchical Agentic Language (CHAL), a multi-agent dialectic framework that treats defeasible argumentation as an engine for belief optimization. Each agent ma…
- BEHAVE: A Hybrid AI Framework for Real-Time Modeling of Collective Human Dynamics
arxiv-ai· 15-may
arXiv:2605.12730v1 Announce Type: new Abstract: Existing AI systems for modeling human behavior operate at the level of individuals or detect events after they occur. As a result, they systematically fail to capture the collective dynamics that determine whether a group remains stable or transitions into escalation or breakdown. We propose a different foundation: a group of interacting humans constitutes a complex dynamical system in the precise mathematical sense, exhibiting emergence, nonlinearity, feedback loops, sensitivity near critical points, and phase transitions between qualitatively distinct regimes. The state of such a system is not located within any single participant; it is distributed across mutual influence loops and observable through the micro-dynamics of the body. We in…
- State-Centric Decision Process
arxiv-ai· 15-may
arXiv:2605.12755v1 Announce Type: new Abstract: Language environments such as web browsers, code terminals, and interactive simulations emit raw text rather than states, and provide none of the runtime structure that MDP analysis requires. No explicit state space, no observation-to-state mapping, no certified transitions, and no termination criterion. We introduce the State-Centric Decision Process (SDP), a runtime framework that constructs these missing inputs by having the agent build them, predicate by predicate, as it acts. At each step the agent commits to a natural-language predicate describing how the world should look, takes an action to make it true, and checks the observation against it. Predicates that pass become certified states, and the resulting trajectory carries the four …
- Think Twice, Act Once: Verifier-Guided Action Selection For Embodied Agents
arxiv-ai· 15-may
arXiv:2605.12620v1 Announce Type: new Abstract: Building generalist embodied agents capable of solving complex real-world tasks remains a fundamental challenge in AI. Multimodal Large Language Models (MLLMs) have significantly advanced the reasoning capabilities of such agents through strong vision-language knowledge and chain-of-thought (CoT) reasoning, yet remain brittle when faced with challenging out-of-distribution scenarios. To address this, we propose Verifier-Guided Action Selection (VegAS), a test-time framework designed to improve the robustness of MLLM-based embodied agents through an explicit verification step. At inference time, rather than committing to a single decoded action, VeGAS samples an ensemble of candidate actions and uses a generative verifier to identify the most…
- On the Size Complexity and Decidability of First-Order Progression
arxiv-ai· 15-may
arXiv:2605.12691v1 Announce Type: new Abstract: Progression, the task of updating a knowledge base to reflect action effects, generally requires second-order logic. Identifying first-order special cases, by restricting either the knowledge base or action effects, has long been a central topic in reasoning about actions. It is known that local-effect, normal, and acyclic actions, three increasingly expressive classes, admit first-order progression. However, a systematic analysis of the size of such progressions, crucial for practical applications, has been missing. In this paper, using the framework of Situation Calculus, we show that under reasonable assumptions, first-order progression for these action classes grows only polynomially. Moreover, we show that when the KB belongs to decidab…
- Learning Transferable Latent User Preferences for Human-Aligned Decision Making
arxiv-ai· 15-may
arXiv:2605.12682v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as reasoning modules in many applications. While they are efficient in certain tasks, LLMs often struggle to produce human-aligned solutions. Human-aligned decision making requires accounting for both explicitly stated goals and latent user preferences that shape how ambiguous situations should be resolved. Existing approaches to incorporating such preferences either rely on extensive and repeated user interactions or fail to generalize latent preferences across tasks and contexts, limiting their practical applicability. We consider a setting in which an LLM is used for high-level reasoning and is responsible for inferring latent user preferences from limited interactions, which guides downs…
- How Pinterest Built a Production MCP Ecosystem
bytebytego· 11-may
In this article, we look at how Pinterest designed that ecosystem and what they had to get right beyond the protocol itself.May 11 • ByteByteGo330611
- A Guide To Event-Driven Architectural Patterns
bytebytego· 14-may
Distributed systems are built out of services that need to communicate, and the simplest way to do that is for one service to call another directly and…15 hrs ago • ByteByteGo944
- Become an AI Engineer | Enrollment Ends Soon
bytebytego· 08-may
Our 6th cohort of Becoming an AI Engineer starts in about a week.May 8 • ByteByteGo26344
- EP214: Claude Code vs. OpenClaw: 5 Design Dimensions
bytebytego· 09-may
Both are highly capable, but they have key architectural differences.May 9 • ByteByteGo29064
- High Performance Rate Limiting at Databricks
bytebytego· 13-may
In this article, we look at how Databricks implemented rate limiting at scale, how they shrank the critical path, and the accuracy tradeoff that…May 13 • ByteByteGo23821
- How Figma Upgraded Data Pipeline from Multi-Day Latency to Real-Time
bytebytego· 12-may
In this article, we will learn what happened as Figma grew and how its engineering team handled the growth in terms of the data pipeline issues.May 12 • ByteByteGo27834
- When "idle" isn't idle: how a Linux kernel optimization became a QUIC bug
cloudflare· 12-may
We investigated a bug where CUBIC's congestion window became pinned at its minimum floor, causing a performance to plummet. The fix involved correctly measuring idle periods to distinguish RTT wait times from actual application idleness.
- Our billing pipeline was suddenly slow. The culprit was a hidden bottleneck in ClickHouse
cloudflare· 14-may
When a partitioning change to our petabyte-scale ClickHouse cluster caused critical billing jobs to stall, standard metrics showed no obvious errors. This post explores how we identified severe lock contention in ClickHouse's query planner and built upstream patches to fix it.
- Browser Run: now running on Cloudflare Containers, it’s faster and more scalable
cloudflare· 13-may
We’ve enabled higher usage limits, faster performance, better reliability, and increased shipping velocity for our Browser Run product by rebuilding on top of Cloudflare’s Containers. Here’s how.
- Genymobile/scrcpy
github-trending
Display and control your Android device This GitHub repo (https://github.com/Genymobile/scrcpy) is the only official source for the project. Do not download releases from random websites, even if their name contains scrcpy. scrcpy (v4.0) pronounced "screen copy" This application mirrors Android devices (video and audio) connected via USB or TCP/IP and allows control using the computer's keyboard and mouse. It does not require root access or an app installed on the device. It works on Linux, Windows, and macOS. It focuses on: lightness: native, displays only the device screen performance: 30~120fps, depending on the device quality: 1920×1080 or above low latency: 35~70ms low startup time: ~1 second to display the first image non-intrusiveness: nothing is left installed on the Android devic…
- influxdata/telegraf
github-trending
Agent for collecting, processing, aggregating, and writing metrics, logs, and other arbitrary data. Telegraf Telegraf is an agent for collecting, processing, aggregating, and writing metrics, logs, and other arbitrary data. Offers a comprehensive suite of over 300 plugins, covering a wide range of functionalities including system monitoring, cloud services, and message passing Enables the integration of user-defined code to collect, transform, and transmit data efficiently Compiles into a standalone static binary without any external dependencies, ensuring a streamlined deployment process Utilizes TOML for configuration, providing a user-friendly and unambiguous setup experience Developed with contributions from a diverse community of over 1,200 contributors Users can choose plugins from …
- K-Dense-AI/scientific-agent-skills
github-trending
A set of ready to use Agent Skills for research, science, engineering, analysis, finance and writing. Scientific Agent Skills 🔔 Claude Scientific Skills is now Scientific Agent Skills. Same skills, broader compatibility — now works with any AI agent that supports the open Agent Skills standard, not just Claude. New: K-Dense BYOK — A free, open-source AI co-scientist that runs on your desktop, powered by Scientific Agent Skills. Bring your own API keys, pick from 40+ models, and get a full research workspace with web search, file handling, 100+ scientific databases, and access to all 135 skills in this repo. Your data stays on your computer, and you can optionally scale to cloud compute via Modal for heavy workloads. Get started here. A comprehensive collection of 135 ready-to-use scienti…
- obra/superpowers
github-trending
An agentic skills framework & software development methodology that works. Superpowers Superpowers is a complete software development methodology for your coding agents, built on top of a set of composable skills and some initial instructions that make sure your agent uses them. Quickstart Give your agent Superpowers: Claude Code, Codex CLI, Codex App, Factory Droid, Gemini CLI, OpenCode, Cursor, GitHub Copilot CLI. How it works It starts from the moment you fire up your coding agent. As soon as it sees that you're building something, it doesn't just jump into trying to write code. Instead, it steps back and asks you what you're really trying to do. Once it's teased a spec out of the conversation, it shows it to you in chunks short enough to actually read and digest. After you've signed o…
- roboflow/supervision
github-trending
We write your reusable computer vision tools. 💜 notebooks | inference | autodistill | maestro 👋 hello We are your essential toolkit for computer vision. From data loading to real-time zone counting, we provide the building blocks so you can focus on building applications around your models. 🤝 💻 install Pip install the supervision package in a Python>=3.9 environment. pip install supervision Read more about conda, mamba, and installing from source in our guide. 🔥 quickstart models Supervision was designed to be model agnostic. Just plug in any classification, detection, or segmentation model. For your convenience, we have created connectors for the most popular libraries like Ultralytics, Transformers, MMDetection, or Inference. Other integrations, like rfdetr, already return sv.Detec…
- ruvnet/RuView
github-trending
π RuView turns commodity WiFi signals into real-time spatial intelligence, vital sign monitoring, and presence detection — all without a single pixel of video. π RuView Beta Software — Under active development. APIs and firmware may change. Known limitations: ESP32-C3 and original ESP32 are not supported (single-core, insufficient for CSI DSP) Single ESP32 deployments have limited spatial resolution — use 2+ nodes or add a Cognitum Seed for best results Camera-free pose accuracy is limited (PCK@20 ≈ 2.5% with proxy labels) — camera ground-truth training targets 35%+ PCK@20; the pipeline is implemented, but the data-collection and evaluation phases (ADR-079 P7–P9) are still pending, so no measured camera-supervised PCK@20 has been published yet Contributions and bug reports welcome at Issu…
- shiyu-coder/Kronos
github-trending
Kronos: A Foundation Model for the Language of Financial Markets Kronos: A Foundation Model for the Language of Financial Markets Deutsch | Español | Français | 日本語 | 한국어 | Português | Русский | 中文 Kronos is the first open-source foundation model for financial candlesticks (K-lines), trained on data from over 45 global exchanges. 📰 News 🚩 [2025.11.10] Kronos has been accpeted by AAAI 2026. 🚩 [2025.08.17] We have released the scripts for fine-tuning! Check them out to adapt Kronos to your own tasks. 🚩 [2025.08.02] Our paper is now available on arXiv! 📜 Introduction Kronos is a family of decoder-only foundation models, pre-trained specifically for the "language" of financial markets—K-line sequences. Unlike general-purpose TSFMs, Kronos is designed to handle the unique, high-noise char…
- supertone-inc/supertonic
github-trending
Lightning-Fast, On-Device, Multilingual TTS — running natively via ONNX. Supertonic — Lightning Fast, On-Device, Accurate TTS Supertonic is a lightning-fast, on-device text-to-speech system designed for local inference with minimal overhead. Powered by ONNX Runtime, it runs entirely on your device—no cloud, no API calls, no privacy concerns. 📰 Update News 2026.04.29 - 🎉 Supertonic 3 released with 31-language support, improved reading accuracy, fewer repeat/skip failures, and v2-compatible public ONNX assets. Demo | Models 2026.01.22 - Voice Builder is now live! Turn your voice into a deployable, edge-native TTS with permanent ownership. 2026.01.06 - 🎉 Supertonic 2 released with 5-language support. The v2 code path is preserved on the release/supertonic-2 branch. 2025.12.10 - Added supe…
- Random AI Explained Fast
hn-ai· 15-may
Article URL: https://www.youtube.com/watch?v=XURpiqSelBw Comments URL: https://news.ycombinator.com/item?id=48144690 Points: 2 # Comments: 0
- The rise and fall of an AI-driven 'local news outlet' in South Florida
hn-ai· 15-may
Article URL: https://floridatrib.org/2026/05/14/the-rise-and-fall-of-an-ai-driven-local-news-outlet-in-south-florida/ Comments URL: https://news.ycombinator.com/item?id=48145022 Points: 2 # Comments: 0
- Multi-LLM trading harness with live leaderboard on Alpaca paper trades
hn-ai· 15-may
Article URL: https://github.com/achaljhawar/1rok Comments URL: https://news.ycombinator.com/item?id=48144911 Points: 1 # Comments: 0
- FilePilot AI – local-first desktop file manager with optional AI summaries
hn-ai· 15-may
Article URL: https://github.com/cuiheng511/filepilot-ai Comments URL: https://news.ycombinator.com/item?id=48144650 Points: 2 # Comments: 0
- What Are AI Ethics
hn-ai· 15-may
Article URL: https://krellixlabs.com/en/blog/what-are-ai-ethics Comments URL: https://news.ycombinator.com/item?id=48145017 Points: 1 # Comments: 0
- Ask HN: How do you catch regressions when you change your AI agent's prompt?
hn-ai· 15-may
i've been building agents for a while and kept running into the same problem: change the sys prompt, swap models, or tweak for the agent calls a tool — and something subtly breaks. The only way I found out was manually running it or a user reporting it. Comments URL: https://news.ycombinator.com/item?id=48144593 Points: 1 # Comments: 0
- What Are the Different Types of AI Testing Tools?
hn-ai· 15-may
Comments URL: https://news.ycombinator.com/item?id=48144848 Points: 2 # Comments: 0
- Agentic evals or LLM as a judge? considering cost, time and quality
hn-ai· 15-may
Comments URL: https://news.ycombinator.com/item?id=48144995 Points: 1 # Comments: 0
- OpenAI just lost its enterprise AI crown to Anthropic
hn-ai· 15-may
Article URL: https://www.businessinsider.com/anthropic-tops-openai-business-ai-adoption-ramp-index-2026-5 Comments URL: https://news.ycombinator.com/item?id=48144749 Points: 4 # Comments: 0
- 'There are no rules':spotlight on Gossip Goblin as AI film-making enters new era
hn-ai· 15-may
Article URL: https://www.theguardian.com/technology/2026/may/14/gossip-goblin-ai-film-making-new-era-hollywood Comments URL: https://news.ycombinator.com/item?id=48145080 Points: 2 # Comments: 0
- [AINews] Codex Rises, Claude Meters Programmatic Usage
latentspace· 14-may
a quiet day lets us report on a long trend of the major coding agents
- [AINews] Everything is Conductor
latentspace· 15-may
an ultra quiet day lets us highlight a smaller trend.
- AI-Native Healthcare: 100M Doctor Visits, 10–20 Hours Saved, Prior Auth in Minutes — Janie Lee & Chai Asawa, Abridge
latentspace· 14-may
How Abridge is quietly turning the patient and clinician conversation into the operating system of healthcare
- Fragments: May 14
martin-fowler· 14-may
Last week I spent a day at a retreat that brought together several people working in software development to talk about the profession’s future with the rise of agentic programming. The event was help under the Chatham House Rule, so I can’t attribute the comments and stories I heard. (If anyone recognizes themselves, and would like attribution, let me know.) Here are a few tidbits that caught my notebook. ❄ ❄ One group developed a behavioral clone of GNU Cobol compiler in Rust. The result is 70K lines of Rust and was built in 3 days. This is yet another sign of the ability of LLMs to do a good job of porting existing code to a new platform. Good regression tests are extremely valuable here (and I don’t know how good GNU Cobol’s are). There’s also the possibility of building a test suite …
- Bliki: Interrogatory LLM
martin-fowler· 14-may
When we need an LLM to perform a complex task, we often need to feed it a lot of context. Coming up with a design for a new feature requires descriptions of how we want the feature to appear to the user, guidelines on how it should be implemented, information on external systems to consult, and so on. All this can be several pages of markdown. The obvious way to do this is for a human to write this context, but an alternative is to use an LLM to write this context after interviewing a human. The way I can do this is to prompt the LLM to interrogate me. It should ask me all the questions it needs to create this appropriate context. I can feed much of the information it needs, and tell it other sources it needs to consult if it can't figure those out itself. Once it's done, it can then crea…
- What is Code
martin-fowler· 12-may
Increasingly humans delegate writing code to agents. Will there even be source code in the future? To wrestle with this question, we have to understand what code is. Unmesh Joshi sees code as having two distinct but intertwined purposes: instructions to a machine and a conceptual model of the problem domain. He explores why it's vital to build a vocabulary to talk to the machine, how programming languages are thinking tools, and how this affects our future as we work with LLMs. more…
- Scaling ArchUnit with Nebula ArchRules
netflix-tech· 08-may
- Welcome to the Datasette blog
simonw· 13-may
<p><strong><a href="https://datasette.io/blog/2026/new-blog/">Welcome to the Datasette blog</a></strong></p> We have a bunch of neat Datasette announcements in the pipeline so we decided it was time the project grew an official blog.</p> <p>I built this using OpenAI Codex desktop, which turns out to have the Markdown session transcript export feature I've always wanted. Here's <a href="https://gist.github.com/simonw/885b11eee46822622b8031a1f4e5f3a3">the session that built the blog</a>. See also <a href="https://github.com/simonw/datasette.io/issues/179">issue 179</a>. <p>Tags: <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/datasette">datasette</a>, <a href="https://simonwillison.net/tags/generative-ai">generative-ai</a>, <a href="https://simonw…
- Not so locked in any more
simonw· 14-may
<p>This <a href="https://simonwillison.net/2026/May/14/mitchell-hashimoto/">Mitchell Hashimoto quote</a> about Bun migrating from Zig to Rust reminded me of a similar conversation I had at a conference last week.</p> <p>I was talking to someone who worked for a medium sized technology company with a pair of legacy/<a href="https://simonwillison.net/2018/Jul/17/mark-norman-francis/">legendary</a> iPhone and Android apps.</p> <p>They told me they had just completed a coding-agent driven rewrite of both apps to React Native.</p> <p>I asked why they chose that, given that coding agents presumably drive down the cost of maintaining separate iPhone and Android apps.</p> <p>They said that React Native has improved a lot over the past few years and covered everything their apps needed to do.</p> …
- Quoting Boris Mann
simonw· 13-may
<blockquote cite="https://bsky.app/profile/bmann.ca/post/3mlp2ipupv22z"><p>“11 AI agents” is meaningless as a phrase. </p> <p>If I said “I have 11 spreadsheets” or “I have 11 browser tabs” to do my work, it means about the same thing.</p></blockquote> <p class="cite">— <a href="https://bsky.app/profile/bmann.ca/post/3mlp2ipupv22z">Boris Mann</a></p> <p>Tags: <a href="https://simonwillison.net/tags/ai-agents">ai-agents</a>, <a href="https://simonwillison.net/tags/ai">ai</a>, <a href="https://simonwillison.net/tags/agent-definitions">agent-definitions</a></p>
- datasette-ip-rate-limit 0.1a0
simonw· 14-may
<p><strong>Release:</strong> <a href="https://github.com/datasette/datasette-ip-rate-limit/releases/tag/0.1a0">datasette-ip-rate-limit 0.1a0</a></p> <p>The <a href="https://datasette.io/">datasette.io</a> site was being hammered by poorly-behaved crawlers, so I had Codex (GPT-5.5 xhigh) build a configurable rate limiting plugin to block IPs that were hammering specific areas of the site too quickly.</p> <p>Here's <a href="https://github.com/simonw/datasette.io/blob/b6022bf9987661b94a26d3143028193a6cabfdcf/datasette.yml#L103-L116">the production configuration</a> I'm using on that site for the new plugin:</p> <pre> <span class="pl-ent">datasette-ip-rate-limit</span>: <span class="pl-ent">header</span>: <span class="pl-s">Fly-Client-IP</span> <span class="pl-ent">max_keys</span>: <span clas…