Security researchers have uncovered six critical vulnerabilities in the proximity file-transfer protocols Apple AirDrop and Android Quick Share, collectively used by over five billion devices.
Open-source models are now offering enterprise-grade capabilities to convert PDFs into structured JSON, unlocking vast pools of data for AI agents and large language models.
GPT-5.5, the latest reasoning model from OpenAI's Codex platform, is exhibiting a peculiar clustering of reasoning tokens at fixed thresholdsâspecifically 516, 1034, and 1552âaccording to a detailed GitHub issue filed by a researcher.
Junyang Lin, the former technical lead of Alibaba's Qwen project, who stepped down on March 3, 2026, delivered a talk titled 'Qwen: Towards a Generalist Model / Agent,' ending with a provocative line: 'Training models -> training agents.' As an independent researcher, he later expanded on this vision in a detailed post, critiquing the hybrid thinking approach and advocating for agentic reasoning.