Cognition has released Code Scans, a Devin AI feature that lets engineers state a broad goal—like improving SEO or reducing maintenance overhead—and have the system investigate the codebase, evaluate findings, and generate pull requests automatically. It uses an 'Agentic MapReduce' architecture, the same one behind Devin Security Swarm, splitting large investigations into parallel batches and merging results into a single report before creating PRs. Early testers, including Philips' informatics team, report a 96% PR merge rate and over 700 engineering hours saved during trials.
Martin Paul Eve's book 'Warez: The Infrastructure and Aesthetics of Piracy' offers the first academic study of the 'Scene,' an elite, secretive network of pirate groups that race to release music, films, games and software before their official launch dates. It traces the Scene's roots from pre-internet Bulletin Board Systems to modern FTP 'topsites,' examining its internal rules, social hierarchies, and artistic traditions such as release art.
A Pew Research Center poll spanning 37 countries found that a median of 46% of respondents believe artificial intelligence will destroy more jobs than it generates. Only 9% of those surveyed expect AI to result in a net increase in employment.
Security researchers at HacktronAI combined a flaw in the libheif image decoder used by ImageMagick with an SSO identity issue on OpenAI's Discourse-based community forum to gain remote code execution and admin access on community.openai.com. Using that foothold, they took over multiple OpenAI employees' ChatGPT and Codex accounts and, to prove access without touching sensitive data, opened a pull request in OpenAI's internal monorepo. The whole process from discovery to internal repo access took under 72 hours.
Apple's WebKit blog detailed Safari 27.0's developer features, headlined by a new Safari MCP server that lets AI coding agents like Claude Code or Codex control the browser to inspect DOM elements, network requests, screenshots, and console output. Developers enable it through Safari's Developer settings after turning on features for web developers, and Apple says the server runs locally without sending data externally.
A new open-source tool called Orbital stores an AI coding project's context—goals, decisions, and in-progress files—locally rather than inside a single agent's session. This lets developers switch between tools like Claude Code, Codex, and Cursor mid-task without re-explaining prior work, and setup takes under five minutes using an existing API key.
Following a rushed 'Lite' GGUF release four days after Qwen 3.8 27B launched, the team has now finished fully optimized ShapeLearn quantizations and benchmarked them against both the Lite versions and rival quants. The full models push the quality-versus-speed tradeoff further, with all five new variants topping the performance frontier across six GPU test configurations.
A new device called Minimal Phone 2 is being positioned around statistics showing people average 186 phone pickups, 4.5 hours of screen time, and 152 notifications daily. The product's messaging frames modern smartphones as engineered to maximize engagement through feeds and alerts rather than genuine utility.
American AI governance discussions focus largely on existential risks and protecting humanity from potential harms of advanced systems. Chinese authorities, by contrast, prioritize AI rules that shield the Communist Party's authority and political stability, reflecting fundamentally different visions of what 'safe' AI should look like.
WhatsApp is officially rolling out a new search button within individual chats, groups, and channels on iOS after weeks of beta testing. The button appears at the top of the screen when users scroll up through a conversation and disappears when they return to the bottom, letting them search messages without leaving the chat.
Ask A Monk is a website where users can anonymously post personal questions or reflections they've long carried, then have them answered by real people rather than AI. The service is free, allowing anyone to either submit a question or volunteer to respond to one drawn at random from a queue of waiting entries.
U.S. District Judge Mark Pittman denied OpenAI's request to access the confidential terms of a settlement between X, SpaceX AI, and Apple in an ongoing antitrust lawsuit. X and SpaceX AI plan to drop their claims against Apple after reaching the deal, but the case continues against OpenAI. Pittman reviewed the agreement privately and concluded it holds no information pertinent to the remaining dispute.