I don’t always check to be sure my podcasts have downloaded 3x before my flight but I just did for @posit.co / @mchow.com ‘s latest Test Set episode. Loving this series full of hype free data talk that makes every other Monday a touch better!
I love the causal ML literature but my most lukewarm take is that we gotta rebrand “nuisance models” as “nuance models”. No reason to insult intricate models that are dealing with quite a lot of complexity for you
Wait wait -- is "get hurt, get PT, briefly flirt with becoming a PT, read PT reddit, retreat" a 30s development milestone and not just me? This was my 2025 😅