🔍 Android 17 adds virtual gamepad for foldables

Android 17 introduces a gaming mode for foldable phones that turns half the unfolded screen into a virtual gamepad with touch buttons. It arrives in the coming months.

The virtual controller works at the system level, supporting any game with physical gamepad compatibility. It offers a full set of controls: D-pad, sticks, A, B, X, Y buttons, bumpers, triggers L1-L3 and R1-R3, plus start.

Users can customize the layout by moving sticks, resizing buttons, and adjusting vibration. The mode activates when unfolding the phone and hides or disables automatically if a real controller connects.

📊@tech

🤖 OpenAI’s GPT-5.6 Sol Beats Claude Mythos

OpenAI’s GPT-5.6 Sol edges out Claude Mythos 5 in agent coding with 88.8% vs. 88% on Terminal-Bench 2.1. The Sol Ultra variant scores 91.9%.

Sol matches Anthropic’s model in cybersecurity using three times fewer tokens. The US government restricts its release, which OpenAI opposes.

OpenAI plans wider API and Cerebras access in July. Officials previously blocked Anthropic’s Fable 5, showing ongoing regulatory hurdles.

📊@tech

💻 China’s LineShine Tops Supercomputer List Without GPUs

China’s LineShine from Shenzhen leads the TOP500 supercomputer ranking with 2.19 exaflops, 22% faster than the US’s El Capitan.

LineShine uses nearly 14 million computing cores and a custom design without GPUs, avoiding US export limits.

It combines scientific computing and AI, running a decade of weather forecasts over East Asia in 15 hours.

Though top-ranked, LineShine isn’t fastest for all tasks, some large clusters excel at less precise calculations but aren’t listed.

📊@tech

🤖 Claude controls robot dog 20 times faster

Anthropic tested Claude on a four-legged robot with tasks that human teams completed less than a year ago. Claude connected to the robot’s camera and sensors, wrote control code, and set up object detection in under 10 minutes.

Claude] was 37 times faster than humans without AI and nearly 19 times faster than humans with AI. It wrote ten times less code than the previous human+Claude team, with most working on the first try.

📊@tech

📊 Sunlight on Your Face Took 100,000 Years

A photon warming your face now began its journey from the Sun's core about 100,000 years ago, when early humans were just leaving Africa and Neanderthals still lived. Inside the Sun, photons collide constantly with electrons, bouncing randomly through dense plasma, making their path trillions of times longer than a straight line.

This slow random walk happens in the Sun's radiative zone, covering 70% of its radius. Photons lose energy step by step, shifting from high-energy gamma rays to visible light by the time they reach the surface. The final stretch through the convective zone takes only months, but the bulk of the journey is a century-long crawl.

📊@tech

🔍 Gemini now searches and buys apps in Play Store

Google integrated Play Store into Gemini, letting users find apps by simply telling the AI what they need. For example, saying "map for travel abroad" prompts Gemini to locate and open the app page directly.

Gemini also enables buying Play gift cards and in-game items right in chat. This feature is limited to personal Google accounts, users 18+, and is rolling out gradually on Android.

📊@tech

🤖 US Allows Anthropic Access to Mythos 5

The US government cleared Anthropic to expand access to Mythos 5 after a security review. The model was previously restricted due to concerns under the Trump administration.

Anthropic is now the first major AI lab to pass this government filter publicly. The exact security issues were not disclosed but have been resolved.

📊@tech

🖥 Microsoft updates Excel Copilot for finance teams

Microsoft rolled out new features for Copilot in Excel aimed at financial teams. It now includes ready-made skills for common tasks, connections to six financial data sources, and full transparency of its actions.

Copilot uses templates to perform tasks like company valuation models or quarterly closes step-by-step instead of improvising. It pulls data directly from sources like CB Insights, PitchBook, FactSet, and S&P Global without manual input.

Before making changes, Copilot shows a detailed plan of affected cells and formulas. All edits are logged and marked as done by Copilot. The update is available for Microsoft 365 Copilot users on Windows, Mac, and browser.

📊@tech

🍏 Apple Plans Touchscreen MacBook on M5 Chips

Apple is preparing a touchscreen MacBook using current top-tier M5 Pro and M5 Max chips, skipping the M6 Pro and Max generations. The new 14- and 16-inch models are expected between late 2026 and early 2027.

The laptop will feature an OLED display, Dynamic Island interface from iPhone, and a refreshed chassis design. The M7 chip series is still in testing and likely to arrive by the end of 2027, with M7 Ultra possibly in 2028.

📊@tech

🤖 Anthropic launches Claude Sonnet 5

Anthropic released Claude Sonnet 5, the most agentic Sonnet model. It plans actions, uses a browser and terminal, and handles tasks once needing larger models.

Sonnet 5 narrows the gap with Opus 4.8 at a lower cost. It completes complex tasks fully, unlike earlier Sonnet versions that stalled.

Testers say it writes tests to reproduce bugs, fixes code, and reverts changes to confirm issues without guidance. It also cuts false info, resists manipulation, and declines harmful requests better.

Sonnet 5 is now default on free and Pro plans and available on Max, Team, Enterprise, Claude Code, and API. Intro pricing lasts until August 31.

📊@tech

Ⓜ️ Meta’s Brain2Qwerty v2 reads silent typing

Meta’s Brain2Qwerty v2 reconstructs text typed silently from continuous brain signals without exact keypress timing, reaching 61% word accuracy on average and about 70% with top users.

It processes raw MEG data through an encoder, an Aligner detecting word boundaries and embeddings, then a large language model refines the output. The system costs millions and isn’t consumer-ready.

Accuracy remains low for practical use, but a near-perfect scaling law shows quality improves almost logarithmically with more data, with no saturation yet. Noninvasive thought reading may just need bigger datasets.

📊@tech

🙀 OpenClaw Launches Native iOS App

OpenClaw, the open-source AI agent, has released a native app for iPhone and iPad. Previously accessed via Telegram on mobile, it now offers a dedicated client available on the App Store.

OpenClaw runs independently on Mac or PC, connects to APIs of any model like Claude, OpenAI, or Gemini, and executes tasks directly within the system.

📊@tech

🤖 Anthropic restores Fable 5 and Mythos 5 models

The US Commerce Department lifted export restrictions on Anthropic's Fable 5 and Mythos 5, allowing access to resume this Wednesday. The rollout starts on Claude platforms and will expand to AWS, Google Cloud, and Microsoft Foundry.

Anthropic had to disable these models after Amazon researchers flagged a hacking risk. To fix this, they trained a new protective classifier that blocks the exploit over 99% of the time, redirecting blocked requests to Opus 4.8.

📊@tech

🍏 Cursor launches iOS app for AI coding agents

Cursor released a public beta of its iOS app, allowing users to run AI coding agents directly from their iPhones. You can pick a repository, start an agent, and even dictate tasks by voice.

The app sends notifications when the agent finishes, letting you review and merge code changes immediately. It supports all paid subscriptions, with a 75% discount on the Composer 2.5 model until July 5.

📊@tech

🤖 Anthropic launches Claude Science for research

Anthropic introduced Claude Science, an AI agent designed as an end-to-end research infrastructure. It handles literature search, data analysis, computation, visualization, and article preparation in one interface.

Claude Science comes with a built-in skill base and offers various visualizations and artifact displays. It is currently in beta and available as an app for MacOS and Linux.

📊@tech

🔍 Google Builds Data Center from Old Smartphones

Google and UC San Diego researchers are creating a data center using 2,000 recycled Pixel smartphones, launching this fall. They remove screens and batteries, using only the motherboards clustered into a computing platform.

The project runs a Linux server distro on Android hardware, turning each phone into a small server. Some smartphone cores outperform traditional server cores in benchmarks, but limited memory and I/O restrict tasks.

A 20-phone cluster handled 75+ students’ workloads faster than a standard AWS backend. The full cluster aims to deliver about 50 servers at a fraction of typical costs. Reliability remains a key concern since phones aren’t built for continuous heavy use.

📊@tech

🌐 China hits 51.3 Tbps internet speed

Chinese researchers achieved 51.3 Tbps internet speed using hollow-core fiber optics, surpassing the default 100 Mbps by 513,000 times.

The hollow-core fiber transmits signals through an air channel instead of glass, reducing signal loss and boosting speed.

📊@tech

😐 xAI launches Voice Agent Builder

xAI introduced Voice Agent Builder, a platform to create voice agents that handle calls and tasks over the phone. Users describe agent behavior in plain language, upload documents, connect tools, and the agent manages calls with background noise, accents, interruptions, and complex queries in 25 languages.

The agents answer questions from uploaded documents, schedule meetings in Google Calendar or Outlook, check order status via APIs, search the internet for current info, and transfer calls to humans if needed. Each call is recorded with a transcript and action log.

📊@tech