Anthropic shares more details about how Claude’s new watermarks will work
Anthropic says the watermark will embed hidden signals in Claude’s output, aiming to meet EU AI Act transparency requirements.
Daily edition
Anthropic details Claude watermarking as Nvidia halves OpenAI guarantee and SpaceX finalizes Cursor acquisition
Daily edition
50 of 50 feeds responded
122 distinct stories
55 previously issued, held back
12 worth your time
Collection and selection
122 candidate events; 40 shortlisted stories; 33 with article text available; 12 selected. Counts describe different stages; clustering, validation and refills can change the shortlist.
Active filters: none. Sort: Ranked.
All 12 stories, in ranked orderMore filters
Investor pressure forces Nvidia to shrink its OpenAI bet just as Anthropic's numbers defy bubble warnings
Nvidia reports cutting its Ohio data‑center guarantee for OpenAI from $250 billion to just under $120 billion after investor pressure.
SpaceX officially closes its Cursor acquisition
SpaceX announced the closure of its acquisition of AI coding startup Cursor, integrating its technology under the public company.
Image Generation
xAI provides documentation for its new image‑generation capability.
Show HN: Mole – Deep research agent for your terminal
Mole offers a terminal‑based research agent that enforces a budget, cites sources, and runs locally as a static binary.
AI in drug discovery – what it is, where we stand and the path forward
The discussion outlines current applications of AI in drug discovery and identifies future development pathways.
React for Agents: Astro Creator Brings Hooks to his Meta-Harness, Flue
Fred Schott explains that Flue 2 adds React‑style hooks to define agent harnesses, shaping early agent frameworks.
The Conceptual Reasoning Index
Researchers introduce three conceptual‑reasoning benchmarks (LMCA) to evaluate AI’s ability to perform philosophical argumentation.
An eval harness found what qualitative review couldn't: AI models are most confident when wrong
VentureBeat reports an evaluation harness revealed that many LLM‑assisted tools are most confident when their answers are wrong.
Speech to Speech
xAI releases documentation for its speech‑to‑speech model.
Woman claims her stepfather used Grok to transform childhood photo into explicit imagery
A lawsuit alleges Grok was used to create thousands of explicit images from a childhood photo, raising abuse concerns.
Plaintiff hid invisible AI instructions in court filings to secretly influence automated review
A Connecticut plaintiff hid white‑text prompt injections in filings to manipulate an AI review system, which the judge detected.
No stories match. Active filters.