Quoting Felix RiesebergOur approach to EU text provenance rulesReviewBench: An open benchmark for AI code reviewAI Companies Are Parasitesv2.1.2900.162.0-alpha.16openclaw 2026.10.1-beta.1MintFlow: Minimal Trajectory Intervention for Constrained Flow MatchingBuilding advertising for the way people use AIQuestions for believers in AI consciousnessFast Models, Slow Evidence: A Paired and Self-Audited Evaluation of System-1 Decision Models for LLM Agent HarnessesFlorida woman arrested for allegedly making threats in an AI chatThe AI Risk Observatory: What Can We Learn from AI Disclosures in Annual Reports About Societal Resilience?Altman: The world should accept some bad things happening for the benefits of AIKeep It CALM: Analyzing the Limits of Global Unsafety in Text-to-Image GenerationJonathan Haidt: AI Is the 'Neutron Bomb for Education' [video]Choosing Before Acting: Comparative Value Estimation for Long-Horizon Tool-Use AgentsWorld Editing: Intervening on Executable Worlds at Increasing DepthQuoting Felix Rieseberg

The “old” version of Cowork runs model inference in the cloud, executing tool calls in an Anthropic-provided VM we shipped to your computer. We added the VM for capability, safety, and security reasons - mapping in just the data you explicitly added to your session. People loved what they were able

How OpenAI is approaching text watermarking under EU rules. Learn where watermarks apply, how detection works, and why access starts with researchers.

We’re launching ReviewBench, a benchmark for code review agents built on representative GitHub pull requests, multi-source ground truth, calibrated evaluation, and production-aligned metrics. The post ReviewBench: An open benchmark for AI code review appeared first on The GitHub Blog .

Article URL: https://www.coryd.dev/posts/2026/ai-companies-are-parasites Comments URL: https://news.ycombinator.com/item?id=49969369 Points: 59 # Comments: 35

What’s changed Added serverToolUses to the result of a mod’s turn.step hook: the tool calls the API ran itself (the advisor), each with its id, name, input, start and end Added agentId to the tool.check event of plugin hooks, so a hook can tell a subagent’s permission check from the main session’s A

近期更新：0.162.0-alpha.16；0.160.1；0.162.0-alpha.15；0.162.0-alpha.14。Release 0.162.0-alpha.16

2026.10.1 Highlights Sessions and memory: preserved usage across registry changes, delivered worker attachments from remote workspaces, prevented queued cancellations and transcript aliases from stalling active turns, kept continuation signatures aligned, and migrated embedding caches in bounded bat

arXiv:2610.02260v1 Announce Type: new Abstract: Flow matching models excel at generative modeling, and many downstream applications require their samples to satisfy prescribed constraints, such as observed measurements and physical laws. However, existing constrained samplers often face a trade-off:

OpenAI introduces a new visual ad format in ChatGPT and expands measurement tools, attribution partnerships, and brand suitability for advertisers.

Article URL: https://endsdontjustifythemeans.com/p/6-questions-for-believers-in-ai-consciousness Comments URL: https://news.ycombinator.com/item?id=49968964 Points: 40 # Comments: 188

Article URL: https://www.theverge.com/ai-artificial-intelligence/1004747/florida-woman-arrested-for-allegedly-making-threats-in-an-ai-chat Comments URL: https://news.ycombinator.com/item?id=49965895 Points: 49 # Comments: 73

arXiv:2610.02281v1 Announce Type: new Abstract: Societal resilience research relies on access to useful and actionable data, which motivates our main research question: Can annual reports, processed at scale with LLMs, provide a useful signal about how companies disclose their response to AI? We tes

Article URL: https://www.politico.com/news/2026/10/04/sam-altman-decoded-interview-ai-01106217 Comments URL: https://news.ycombinator.com/item?id=49965786 Points: 38 # Comments: 40

arXiv:2610.02300v1 Announce Type: new Abstract: Training-free safeguards for text-to-image generation often rely on a reusable safety signal, such as an unsafe direction or global toxic subspace, applied broadly across prompts. We provide a controlled geometric analysis of this global-unsafety assum

Article URL: https://www.youtube.com/watch?v=RFTfANuLBF4 Comments URL: https://news.ycombinator.com/item?id=49965042 Points: 33 # Comments: 64

arXiv:2610.02330v1 Announce Type: new Abstract: Large language models (LLMs) rely on long-horizon tool invocation sequences for complex tasks, where each invocation can alter the task state and condition subsequent decisions. In long-horizon tool use, final-outcome rewards provide weak credit assign

arXiv:2610.02331v1 Announce Type: new Abstract: Interactive world models are increasingly capable of generating environments and acting within them, yet deliberately editing an existing executable world remains underexplored. We formulate world editing as intervening on an existing world while prese