Click any tag below to further narrow down your results
Links
Sigil Wen, a Thiel Fellow who grew up in Silicon Valley's AI circles, launched Underdog, an AI assistant that runs entirely on your computer rather than in the cloud. The product makes money from transaction fees on payments the AI helps you make, not from selling your data or showing ads.
- Underdog runs a 27-billion parameter model locally on Mac and Windows (with Linux, iPhone, Android coming soon), keeping all user data on-device and encrypted, with no cloud servers involved.
- Wen claims the smaller on-device model performs comparably to Claude Opus from six months ago on benchmarks and handles everyday tasks like research and homework questions without needing state-of-the-art capabilities.
- The business model takes a small percentage of payment transactions processed through Stripe instead of charging subscriptions or mining user data, aligning the company's incentives with user privacy rather than data extraction.
This issue covers how to make design systems AI-ready with structured specs and audit scripts, and argues for global preload-based loading states instead of scattered spinners. It also highlights Homebrew 6.0’s security and sandbox upgrades, an AMD auto-update RCE fix, and new on-device AI features from WWDC.
- Design systems become AI-ready by storing specs as structured Markdown, locking tokens into closed named variables, and running audit/sync scripts to catch hard-coded overrides and stale docs.
- Loading UX improves by preloading and caching data at the router/app level with one global fallback, instead of scattering per-component spinners.
- Homebrew 6.0.0 adds mandatory "tap trust" approval for third-party repos, defaults to a faster JSON API, and brings sandboxing to Linux.
- AMD's AutoUpdate tool fetched metadata over HTTPS but downloaded executables over unsigned HTTP, a flaw AMD initially resisted fixing before patching it.
Apple celebrated its 50th anniversary while facing significant challenges in the AI space. The company is shifting its strategy by partnering with Google for AI enhancements to Siri, amid concerns about user data and competition from rivals. Industry analysts suggest Apple must adapt quickly to maintain its relevance in a rapidly evolving technological landscape.
- Apple is licensing Google's Gemini AI to power upgraded Siri features, a reversal of its privacy-first stance that risks feeding user data into Google's ad business
- Siri has stagnated since its 2011 launch, with insiders like co-founder Dag Kittlaus arguing it never reached its potential without Steve Jobs' vision
- Apple is betting on on-device AI chips to process queries locally rather than in the cloud, aiming to preserve privacy while catching up on AI capability
- New AI features are slated by year-end, but early consumer reaction has been mixed, reflecting doubts about Apple's overall AI strategy