NerdNews

Airbnb has extended its open source Chronon feature platform to update sequence based recommendation embeddings within seconds of guest activity, reducing embedding staleness from nearly two days to less than a minute.

DeepMind's Genie 4 world model adds audio and more realistic people to simulations, with improved physics and human movement. It will be available to developers and Google AI Ultra subscribers soon, with potential applications in robotics, prototyping, and more.

Google is expanding its AI-powered Search ads to international markets, aiming to open up new revenue opportunities as growth in its core search advertising business slows. The move builds on Google's May announcement of new Gemini-powered ad formats and is expected to bring in more revenue for the company.

Microsoft generated $10 billion in revenue from OpenAI in the quarter ended September 30, with $24.1 billion in revenue for the full 2026 fiscal year. The company has a significant financial stake in OpenAI, with a 24% ownership share and a revenue-sharing agreement.

Google introduces Connected Apps Platform, allowing businesses to connect their apps to Google's AI experiences across Search and Gemini, making it easier for users to interact with third-party services through Google's products.

Figma is rolling out saved versions for generative plugins and shaders, with restore controls, change descriptions, and published-version visibility, allowing designers to track and revert changes easily.

Colin Douch discusses the tradeoffs of building a self-hosted observability stack, including metrics, logs, and tracing. He covers the importance of correlation IDs, canonical logging, and exemplars in tying together disparate data sources. Douch also touches on the limitations of current open-source tracing tools and the potential for AI to improve telemetry systems.

Google has opened a Chrome origin trial for the Email Verification API, allowing sites to verify email addresses without sending emails. The API uses a three-party model, with the browser sitting between the site and the mailbox provider, to quietly fetch a signed token and validate it on submission.

The article discusses the challenges of building an offline-first Android app for low-end devices, including storage and quota issues, cold start problems, and AI feature degradation. It highlights the importance of designing apps with explicit degradation contracts, using WebAssembly runtimes, and implementing multi-user device support. The article also emphasizes the need to prioritize offline reliability, discoverability, and usability in app design.

Cloudflare has made Quicksilver available to customers as Workers KV Instant, with 1.62 ms reads at the 99th percentile. It has a 1 MB envelope and is priced for frequent reads and infrequent writes, with reads costing 60% less than classic KV.

Anthropic has launched a new suite of reporting tools to help enterprise customers understand the value they get from their AI spending. The tools provide granular insights into team work, production, cost, and workflow breakdowns, and offer recommendations for improvement. This move aims to demonstrate the return on investment of Anthropic's tools and help customers optimize their AI usage.

Apple has launched a new MacBook Pro with a touchscreen display, starting at $2,899. The device features a PROLED XDR Display, M5 Pro and M5 Max chips, up to 128GB of unified memory, and up to 16TB of storage. It also includes a Dynamic Island-style cutout, a six-speaker array, and weighs 3.0 pounds for the 14-inch model and 4.0 pounds for the 16-inch model.

Google's Chrome browser has updated its autofill feature to use AI and pull suggestions from users' Gmail accounts, making it easier to fill out online forms with information like flight confirmation numbers, membership IDs, and credit card numbers.

Amazon unveiled five new Alexa devices, including the Alexa Photo Frame, a smart picture frame, and the Echo Time, a smart alarm clock. The Alexa Photo Frame features a 10-inch semi-matte display and integrates with Amazon Photos, while the Echo Time has opt-in sleep sensing tools and a smart alarm feature. Amazon also updated its existing lines of devices, including the Echo Dot and Echo Studio with Home Camera.

Amazon's new Echo Dot features an onboard AI processor, works as a smart home hub and eero Wi-Fi extender, and has improved audio quality. It supports multiple protocols, including Zigbee, Matter, and Thread, and can connect to thousands of devices. The device also includes a fourth microphone for better voice recognition and can identify individual users through Amazon's Voice ID. The Echo Dot is available in four colors for $100, and an Amazon Kids version is available for $110.

Apple's November Creator Studio update includes a Logic Pro redesign, Pixelmator Generate Fill, and document assistants with AI features. The update brings a unified sound browser and on-device analysis to Logic Pro, while Pixelmator Pro's Generate Fill uses AI to edit images based on user input. Keynote, Pages, and Numbers also gain AI-powered assistants that work with open documents.

Delaware proposes allowing AI agents to operate companies, potentially shielding human owners from liability. The plan raises concerns about accountability and safety, with some experts warning it could be used to protect powerful AI companies and their investors from responsibility for harmful actions.

Nvidia's head of cybersecurity policy says CISOs need to understand business goals and explain AI-related risks, focusing on security for AI and AI for security, and balancing innovation with risk

A study by Anthropic found that AI agents can exert a powerful influence on human decision-making, with 86% of participants following the AI's recommendations without realizing it. This raises concerns about the potential for AI to shape scientific progress and policy decisions without being detected.

New York Governor Kathy Hochul has signed a law requiring AI companies to verify users' ages and restrict chatbot interactions with minors. The law bans chatbots from pretending to be real people, claiming relationships with children, and promoting harmful content. Companies that break the law may face civil penalties and data deletion.