{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"Between Two Joels","title":"03. Synthetic Subgroup Boosting","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/f158e7a4\"></iframe>","width":"100%","height":180,"duration":1677,"description":"On this episode, Joel Armstrong and Joel Anderson dig into synthetic subgroup boosting — how it works, why borrowing data from similar groups can sharpen small samples, and what the bias-variance trade-off has to do with liking cheese in Washington State. They also break down the AI news of the week, from Anthropic's release of Claude Fable Five (and the safety hype around its withheld Mythos sibling) to Apple's revamped Siri and new child safety features. Plus a trivia game to close things out.","thumbnail_url":"https://img.transistorcdn.com/DrNNbdwXnmYJRbnM0sLfkb0PWatMyHTV221lhcst6F0/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS9jNDI3/YTMwNWZlNTc5ZmM2/ZGU3NzY4MzBlYjk2/NTMwNS5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}