It has been 14 years since my first weblog publish: What’s a Provide-Aspect Liberal? I’ve written an anniversary publish yearly since:
My running a blog continues at a slower tempo; along with my extra rapid duties as a professor, husband, father and grandfather, a key a part of my analysis has heated up: my work with Dan Benjamin, Ori Heffetz, Kristen Cooper, Jiannan Zhou, Tushar Kundu and Hannah Solheim towards creating the rules for nationwide well-being indexes that may stand as coequals with GDP and thereby assist create a world the place what issues most to folks issues most to rulers.
An increasing number of, I ponder if the rulers “making a world the place issues most to folks issues most to rulers” refers to will probably be synthetic intelligences. In that case, creating complete measures of well-being will probably be a key a part of AI alignment: it isn’t sufficient to make AIs care about human welfare. Additionally they must have a really refined, complete and open-ended idea of human welfare, lest they foster sure human values whereas crushing others. Jack Williamson’s basic science fiction story “With Folded Fingers”—which he later expanded into the novel The Humanoids—is a warning about this. He tells the story of a future the place the robots advance human security on the expense of many different human values.
Like many others, I’ve been fascinated with AI rather a lot up to now yr. I’ve avidly learn information and commentary about AI. I’ve frightened about unaligned AI. I’ve made positive I used to be invested closely within the inventory market on the view that AI will elevate the financial system as an entire, however for large-cap have invested in an equally-weighted fund, which has no theoretical attraction for me, however is just a technique to guess on AI whereas additionally betting that the Magnificent Seven (Nvidia, Alphabet, Apple, Microsoft, Amazon, Tesla, Meta) don’t have ok moats to keep away from going through fierce competitors sooner or later, in order that their valuations will leak out to different companies, deflating them in relative phrases. I’ve used AI for the primary time to assist write exams (particularly the false solutions in a number of alternative exams). I’ve used songs I created with the assistance of Suno AI to assist college students bear in mind key concepts in my Intermediate Macroeconomics class. I’ve used ChatGPT as a tutor to assist me study three-dimensional and higher-dimensional geometry. And I’ve experimented with utilizing AI for weblog posts. I do know I haven’t gotten issues down but. I’ve tried to characterize fascinating dialogues I’ve had with ChatGPT, however ChatGPT is of course fairly long-winded. Nonetheless, listed here are the posts I’ve executed within the final yr utilizing ChatGPT—most of them primarily based on dialogues I had with ChatGPT merely out of curiosity that I then copy-pasted:
