Showing posts from April, 2021

The Groffscars ("Oscars") of 2021

This year was a pretty good crop of movies. It's the first year I can recall in which I actually enjoyed all of those nominated for Best Picture (except The Sound of Metal , which is still on my watchlist). Mank was a disappointment but still an enjoyable homage to Citizen Kane. Minari and Nomadland were well-done slices of life's bitterness and sweetness. Judas and the Black Messiah was a riveting political drama that I hope woke people up to the absolute treachery by the FBI against black power and civil rights activists in the 60s and 70s. And Ma Rainey was a dazzling, fast backstage drama. But my favorite historical picture this year was The Trial of the Chicago 7 . I could be biased, having gone to high school with one of the titular seven actors, but I don't think I am. It was a drama both fun and enraging with performances that bounce off the wall, especially Jeremy Strong and Sacha Baron Cohen. The gagging of Bobby Seale was shocking and another historical wak

Is there evidence that recommender systems are changing users' preferences?

In Human Compatible , Stuart Russell makes an argument that I have heard him make repeatedly (I believe on the 80,000 Hours podcast and the Future of Life Institute conversation with Steven Pinker). He suggests a pretty bold and surprising claim: [C]onsider how content-selection algorithms function on social media... Typically, such algorithms are designed to maximize click-through , that is, the probability that the user clicks on presented items. The solution is simply to present items that the user likes to click on, right? Wrong. The solution is to change the user's preferences so that they become more predictable. A more predictable user can be fed items that they are likely to click on, thereby generating more revenue. People with more extreme political views tend to be more predictable in which items they will click on... Like any rational entity, the algorithm learns how to modify the state of its environment—in this case, the user's mind—in order to maximize its own r