0:00Duration unknown
Streams straight from the publisher. podnod never proxies or re-hosts episode audio.
詳細網頁版與參考連結:https://podcast.david888.com/post/2026-08-01
本集解析 Tailscale 針對 Hugging Face 遭 AI Agent 越獄事件的報告,揭示長效憑證風險,並指出模型權重正成為現代管制的密碼學。節目同時評析 YC 的 AI 框架 qm 之安全隱憂;拆解電梯調度演算法,指出預約派梯系統因缺乏動態彈性,在中小型建築效率未必優於傳統演算法;最後引用卡繆荒謬哲學,將工程師修復 Bug、推進開源引擎 Servo 及揭露巨型食品企業等努力,比喻為在混沌世界中堅守尊嚴的反抗。
相關連結:
https://podcast.david888.com/post/2026-08-01
podcast.david888.comTailscale didn't stop the Hugging Face intrusion
news.ycombinator.comElevators
news.ycombinator.comqm
news.ycombinator.comTwenty-five years ago it was cryptography, today it's model weights
news.ycombinator.comThe Absurdity of Albert Camus
news.ycombinator.comJune in Servo: real world compat, media queries, SharedWorker, and more
news.ycombinator.comBig Food vs. the People
news.ycombinator.com