Razib Khan's Unsupervised Learning

  On this episode of Unsupervised Learning Razib talks to the pseudonymous commentator “Peachy Keenan.” A native of Los Angeles with an Ivy League education, Keenan worked in entertainment before detouring into punditry, writing for the Claremont Institute’s The American Mind, appearing on Fox News and penning Domestic Extremist: A Practical Guide to Winning the Culture War. Razib and Keenan discuss her peripatetic and unique journey from a relatively apolitical member of America’s liberal professional managerial class to a conservative Catholic housewife with a large family. Keenan talks about her ability to connect with audiences of all stripes despite her partisan leanings as the product of her cosmopolitan upbringing among coastal elites. Though in her values and practices she lives the life of the “domestic extremist,” she still retains an aesthetic appreciation of the broader culture in which she grew up. Domestic Extremist is to a great extent a roadmap from where she was, to where she is. Keenan offers a sort of  primer on how to change the “factory settings” for the American professional class, proposing traditional family life as an exit out of the endless rat race. They also discuss the reality that the modern conservative culture falls short of produce any art for its own sake, at most putting out fare that ranges from  overly didactic films produced by the Daily Wire to the cringe-inducing Christian film industry. Keenan emphasizes that good art must be good art, first and foremost, and whatever ideological valence should be layered in with subtlety and taste. She also discusses the problems with raising consciousness among conservative philanthropists about the problem of right-wing philistinism, and why aesthetic excellence would be a boon in any attempt to recapture the cultural high ground.

  On this episode Razib talks to Jesse Singal, a journalist who has covered the social science beat for the last decade. Singal has an undergraduate degree in philosophy from University of Michigan and a master’s in public affairs from Princeton. Currently a freelance journalist who writes his own Substack, Singal-Minded, and contributes to Blocked and Reported with Katie Herzog, Singal is formerly an editor at New York Magazine. His first book The Quick Fix: Why Fad Psychology Can't Cure Our Social Ills, covered the replication crisis. Razib and Singal first talk about what he learned, and unlearned, during his time as a reporter at New York Magazine, especially social psychology results that were long on glamor but short on robustness. They discuss how long we’ve known that social psychology had a problem, and whether it still hasn’t reformed itself. Singal also reflects on his role in publicizing sexy findings, and how journalism has taken steps to be more careful lately. They also address some of the specific findings that came out of early 2010’s social science, from implicit bias to power posing. Next, Razib asks Singal about youth gender medicine, and the major controversies over the last few years. Singal discusses the differences between female to male transitions as opposed to male to female, and relates the whole domain back to the replication crisis and the lack of good research. They also discuss political and social aspects, and where Singal sees youth gender medicine going in the next few years.

On this episode of Unsupervised Learning Razib talks to economist Sam Hammond.  Canadian-born Hammond serves as the Senior Economist at the Foundation for American Innovation. His work primarily focuses on innovation and science policy, with particular attention to the societal and institutional impacts of disruptive technologies such as artificial intelligence. Before his role at FAI, Hammond was Director of Poverty and Welfare Policy at the Niskanen Center. Hammond also held a research fellowship at the Mercatus Center at George Mason University, focusing on policy issues related to technology and regulation. He holds a BA in Economics from Saint Mary’s University and MA's in Economics from George Mason University and Carleton University. After a quick discussion about Canadian housing, Razib and Hammond consider his piece 95 theses about AI. Hammond’s contention is that AI might prove as impactful as the printing press, or, at the outer edge equivalent to photosynthesis. Nearly two years into the current “AI hype cycle” we still haven’t found the “killer app” of AI, but thinkers like Hammond are getting ahead of the likely inevitable societal changes. He believes that change is inevitable, and the details that need to be worked out are how we as a species adapt and evolve in response to our technology. Hammond contends that the AI-revolution is likely to produce changes in the next generation analogous to industrial transformations of the late 19th centuries and early 20th centuries, when cars, electrification and airplanes transformed civilization.    For early access, feel free to explore it there. https://www.razibkhan.com/p/sam-hammond-i-for-one-welcome-our   

    The full episode is available on: https://www.razibkhan.com/p/14000-years-of-natural-selection On this episode of Unsupervised Learning Razib talks about what we have learned from a blockbuster new preprint, Pervasive findings of directional selection realize the promise of ancient DNA to elucidate human adaptation. Synchronously released was the Ancient Genome Selection browser, which allows you to trace the allele frequency of variants of interest over the last 10,000 years. Razib covers: The relationship of selection to adaptation and the Darwinian understanding of evolution Non-genetic selection Types of biological selection like positive, negative, background and balancing selection Hard vs. soft sweeps and their relevance to detecting selection in the genome Older forms of natural selection detection between species (dN/dS, Tajima’s D) Newer forms of selection detection within species with haplotype structure, outlier SNP analysis and site frequency spectra The Generalized Linear Mixed Model used to model allele frequency change over time, and estimates of selection in cases where population structure and drift are not sufficient Specific examples of SNPs whose variation can be examined in the browser and are clearly cases of selection Survey of traits that were revealed under selection, including blood groups, pigmentation and intelligence Critiques of the methods due to not accounting for drift or population structure, and its limitations in relation to the ability to port across populations due to LD structure