Weekly reads: Jun 15, 2026

Apple’s Siri AI revamp drew most of the attention at WWDC, but iOS 27 will also add useful AI capabilities across other parts of the iPhone experience. The article highlights that Apple’s broader AI features may matter as much as, or more than, the Siri update for everyday users.

A Wall Street Journal investigation found that Polymarket allegedly paid creators to film deceptive clips showing themselves placing bets and celebrating wins without disclosing sponsorships. The report identified more than 1,100 such videos, with signs that some used lookalike domains rather than real betting activity.

Dany Bittel describes working with crysta.ai to convert a Gaussian splat of an insect into a physical transparent print. The process voxelizes the splat, assigns ink and transparency values to each voxel, and prints it layer by layer, producing a result Bittel compares to modern amber.

The article traces an educational reverse proxy project, TinyGate, from a worker-based design to an epoll implementation and finally a full rewrite around io_uring. It explains that epoll improved performance but still requires extra read/write syscalls after readiness notifications, creating overhead at scale.

Finland is investing in libraries like Helsinki’s Oodi as community hubs that offer far more than books, from workspaces and language classes to music pods, sports gear, and craft activities. The model contrasts with library closures in the US and UK, emphasizing libraries as shared public spaces that support learning, inclusion, and everyday participation.

The author says they subconsciously give more weight to books published before 2022, especially from unfamiliar writers, because those works likely predate widespread LLM use. They acknowledge that AI tools can produce strong work, but still feel that fully human writing and editing carry a sense of effort that affects how they read.

Researchers found serious vulnerabilities across LangGraph, Langflow, and LangChain-core, including SQL injection and path traversal bugs that can lead to remote code execution or secret exposure. Langflow attacks are already active in the wild, while LangGraph and LangChain issues show how agent frameworks can expose OpenAI keys, database credentials, and internal tokens.