Razib Khan's Unsupervised Learning

In this episode of Unsupervised Learning Razib talks to Molson Hart, founder and CEO of Viahart, an educational toy company. He is also co-founder of Edison, an intellectual property-focused litigation financing firm. Hart has gained some visibility as a prominent seller on Amazon, with strong opinions on the company both positive and negative. First, Razib asks Hart about Amazon’s role in the American economy, and how it compares and contrasts with Walmart. Unlike many who have negative experiences with the company, Hart’s attitude seems to be that consumers and producers both need to accept the reality of Amazon’s behemoth position in the American marketplace. It’s not going anywhere, so the question is how to control it, not kill it. If Amazon has brought supply-chain scale to the US economy, America’s partnership with China has taken the concept of scale to a whole new level. Razib asks about Hart’s experience as a businessman in China ten years ago in the border area between Manchuria and North Korea. Hart recounts several major things he learned about the contrast between the US and China. For example, while Americans focus on fairness and rule of law, the Chinese have no such expectations and are very pragmatic (“don’t argue, pay the bribe!”). Second, the Chinese plan fast and make immediate decisions, and then pivot rapidly off mistakes, while Americans tend to over plan. Though China in the early 2010s was very corrupt, Hart feels the last decade has seen a shift away from those practices. Another thing that has changed over the last decade has been an awareness that American and Chinese supply chains need to become decoupled to some extent due to both geopolitical and economic considerations. The COVID-19 interruptions in particular have made many Americans aware of how entangled how our own production processes are with China. But changing the current economic relationships may not be so easy. In the mid-2010’s Hart shifted some of his purchasing to Vietnam. Though the Vietnamese are hungrier and cheaper, they naturally lack the scale, efficiency and specialization of their Chinese competitors. Hart also observes that it is clear that the Chinese workers are among the hardest working and most skilled in the world, so they will not be easy to replace. His contacts in the Pacific Rim believe that only India would ever be able to truly substitute for China because of its size and diversity. Hart notes that one peculiarity of China is that it operates as a large market economy that is culturally less aware of the US than other Asian trading partners. In particular, Chinese English fluency is much lower than that of Indian and Vietnamese. Hart wonders if this state might never change given that the Chinese society and economy are just large enough that they can ignore America more than smaller and less developed nations. Pivoting back to the US, Razib and Hart discuss the “easy money” policies that have dominated American economic policies over the last few years. Hart argues that the ability of Americans to take on debt enables bad policies, from foreign policy adventures to bailouts of firms that should be allowed to fail. Additionally, he argues that inflation reduces the value of American money and the appeal of investing in US “cash” as the safest investment.  They end the discussion with Hart’s bullishness on East Asian economies, despite the demographic and political headwinds. He also believes that the US has a bright future, but we need to accept that we’ll never have 1990’s hyperpower again.

In this episode of Unsupervised Learning Razib talks to Alex Nowrasteh, the director of economic and social policy studies at the Cato Institute. Alex is also the author of Wretched Refuse?: The Political Economy of Immigration and Institutions. His beat at Cato is immigration, and he has been keeping a close watch on the transition between the Biden and Trump administrations. The first issue Razib and Nowrasteh address is the reality that the COVID-19 pandemic resulted in a massive crash in immigration to the US due to Donald J. Trump’s executive orders. Curiously, Trump justified the border closures, not on public health grounds, but to safeguard American jobs. Additionally, Nowrasteh claims that Joe Biden’s administration has been “Trump’s second term” when it comes to immigration levels, as the migration rates have not reached the levels before the COVID-19 pandemic. Though Democratic activists have argued for very liberal immigration policies, Nowrasteh observes that Biden’s 2020 campaign was only marginally more conservative. Contrary to expectations, immigration has been sidelined as a major issue despite Democrats controlling the Presidency and Congress. Nowrasteh asserts there has been “no interest in liberalizing immigration” within the administration, as they fear the political consequences. Pivoting to the international stage, Razib wonders about the exodus of Ukrainians due to the Russian invasion. Nowrasteh commends European generosity in taking in refugees, but he believes that ultimately most of the migrants will remain in Europe due to simple economics. Ukraine was the poorest nation in Europe, and Razib and Nowrasteh wonder if a post-war Ukraine will be filled with the old, the very young and those with few prospects, as the productive working-age population permanently decamps to Western Europe. Next, they discuss the global landscape of demographic transition and the reality that only Sub-Saharan Africa has high fertility rates. Nowrasteh observes that in fact, the Middle East might be the major destination for African migrants due to the collapsing population and the demand for workers in the region. Another obvious target to African migration will be Europe, but it is clear that the US is likely to be a destination when the immigration system relaxes and opens up.  Nowrasteh pivots back the US, and argues that despite all the rhetorical excesses that veer to open borders in the Democratic party, operationally the party hierarchy is much more conservative in terms of policy implementation. Razib points out that some elements of the Left associated with environmentalism, in particular the Sierra Club, have been associated with immigration restrictionism. Nowrasteh then observes that many conservative immigration restriction groups actually have pasts where they were founded by and associated with liberal population control environmentalists in the 1970’s. For example, the current website of the Federation for American Immigration Reform (FAIR) declares that “America’s immigration and border controls have fallen apart due to ineffective leadership from Washington. We face a true national emergency!” Paul Erhlich, Stanford ecologist and environmentalist author of the The Population Bomb (for which the first director of the Sierra Club wrote a foreword), was on the board of FAIR until the mid-2000’s. They close the conversation by reflecting on the global “birth dearth” and the possibility of how many Americans there might be at the end of the 21st-century (sorry Matt Yglesias, a few fewer than 1 billion).

In this episode of Unsupervised Learning Razib talks to James Lee, a professor of psychology at the University of Minnesota. Lee is a co-author of a new paper in Nature, Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals. A landmark in the field of cognitive genomics, this publication is the result of years of collaboration between two dozen researchers.  Over the course of the episode, they deep dive into the results from the publication that Lee in particular finds fascinating. But first, Razib brings up a recent controversy related to Paige Harden’s book The Genetic Lottery and the science that undergirds its thesis. Evolutionary geneticists Graham Coop and Molly Prezworski recently wrote a review of Harden’s book in Evolution, Lottery, luck, or legacy. A review of “The Genetic Lottery: Why DNA matters for social equality”. They argue that Harden overplays her hand in terms of what “polygenic risk scores” can tell us about our future life trajectory (and in particular her focus on education outcomes), as well as their social utility. Harden responds with a piece titled Forests and Trees, contending that Coop and Prezworski mischaracterize her position and seem to hold behavior genetics to an unreasonably high standard of evidentiary validity. In buttressing the science in The Genetic Lottery, Lee expounds on the importance of the finding that genetic positions associated with something like higher educational attainment seem highly correlated with regions of the genome associated with neurological development in particular. Next, Razib asks what aspect of the new paper Lee found most interesting, and he points to the section on the nature of dominance, the characteristic whereby certain genetic variants express a trait when present in a single copy, as opposed to two copies (recessive traits). These arguments go back to Sewall Wright and R. A. Fisher’s debates about the nature of dominance from a century ago, a divergence in viewpoints at the very founding of population genetics as a field. Lee favor’s Wright’s view that dominance is a function of the physiological mechanism of gene expression; a gene that produces proteins will still produce sufficient quantities in even a single copy. In contrast, most of the authors of  Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals favored Fisher’s idea that dominantly expressed genes sweep to selection faster, and so that view is tacitly supported in their conclusions. During the rest of the discussion, Lee expounds on a wide range of topics that touch on behavior genomics, from whether rare variants of large effect will come to be seen as important, to why heritability estimates using family-based designs are so much lower for educational attainment than conventional population-wide statistics, and the relevance of the results from this latest work for evolutionary genetics. Lee makes the case that the synthesis of genomics and behavior genetics makes for a fascinating story of scientific discovery that will help illuminate our understanding of human nature in the 21st-century, far beyond the field’s utility in predicting individual traits.

On this episode of Unsupervised Learning Razib talks to Josiah Neeley, Senior Fellow in Energy at the R Street Institute and co-host of the Urbane Cowboys podcast. They discuss the past, present and future of the energy markets, and how best to understand the workings of the global energy ecosystem. Considering geopolitical events in Europe, with Russia’s invasion of Ukraine, they dive right into how distribution differences between oil and gas will conspire to keep Europeans dependent on Russia for energy for the foreseeable future, blocked from immediately switching to other options. They also address the fact that Russia’s leverage as a petrostate has been diminished by a revolution in oil extraction technology that has at the same time made the US a net exporter of crude. Neeley says it’s critical we acknowledge how disaggregated the energy markets are. While crude oil is a global market defined by a single price that is highly responsive to international events, natural gas prices are very localized. Unlike oil, natural gas is not easy to transport, so there is no pooled market. Energy policy is not singular because energy is defined by diverse technologies with specific constraints and demands. Constant innovation means that in the near future advances in liquifying natural gas (LNG) may make it a more global commodity. Then they discuss what it means to be a renewable energy, the dependence of wind and solar on battery technology, and the fact that, depending on how you think about it, most of our energy is ultimately solar (fossil fuels deriving from organisms whose original energy input was from the sun). Neeley points out that while solar and wind are highly variable inputs, fossil fuels and nuclear energy are relatively constant, resulting in the vastly greater necessity for battery technology in the former two. Meanwhile, because immediate solar technology is arguably the most visible form of energy production (since you probably know exactly how many houses on your block have solar panels, but probably zero run a backyard coal-fired plant) Neeley notes this means Americans often overestimate the relative importance of solar compared to other sources. Finally, they explore the geopolitical dependencies that plague the various energy sources and the peculiar relative stagnation of nuclear technology. Neeley observes that, unlike most innovations, thanks to an accumulation of safety regulations, nuclear power plants have actually become more, not less, expensive, over the last fifty years, resulting in some unfavorable economics. Additionally, nuclear energy requires a high human capital input from engineers and technicians. This may explain why heavy reliance on nuclear power is limited mostly to medium-sized European countries, and it has gained little traction in the developing world.

Today on Unsupervised Learning Razib talks to Jacob L. Shapiro, Director of Geopolitical Analysis at Cognitive Investments. He overviews the geopolitical perspective in understanding international relations, one predicated on looking at nation-states as fundamental units of analysis, in order to achieve a descriptive understanding of the world. Shapiro points out that the more familiar “schools” of foreign policy, from realism to liberal internationalism, use geopolitics as a tool to understand the world but apply their own value-sets to establish particular policies that further certain values and interests. Fundamentally, geopolitics differs in that it is an empirical rather than normative discipline. Shapiro then highlights geopolitics’ 19th-century origins in Europe, its decline by association with Germany in the 1940’s, and its recent renaissance in the US as well persistence in Latin America. The conversation then shifts from theory and abstraction to current events in Russia. Shapiro admits that his own research group put the odds Russia would invade at 30%, in large part because they believed that Vladimir Putin could have gotten what he wanted mostly through intimidation rather than invasion. That being said, he points out that there is a very long history of the Russian state wanting to push to the Carpathian mountains that bound Ukraine’s west due to concerns about defensible frontiers. Shapiro argues that the invasion’s fundamental raison d’etre is the quest for “strategic depth.” He also relays accounts of Putin ruminating on maps and imperial history while in COVID-19 isolation, although he cautions against psychoanalyzing him too much. Razib next asks Shapiro for his take on globalization in the context of the Ukraine-Russia conflict. Shapiro argues that we are truly moving into a multipolar world that is more similar to what occurred in the 1890’s when there was a balance of power in Europe. Shapiro points out that that too was a time of economic and cultural tumult and creativity, Europe’s “Belle Epoque.” For him, this earlier period of globalization illustrates both the promise and peril of a geopolitically balanced world where fates were interlaced by complex networks of free trade. Shapiro’s main worry is a “Black Swan” event with the power to trigger a global conflict, a freak event analogous to the assassination of Archduke Franz Ferdinand by a Serbian radical that ignited World War I. He cautions that a world with more balance of power between nations and leaders, unpredictable decisions grow more likely. Shapiro also argues strongly that the US is under a misimpression in terms of its power and influence in a world where other powers are rising. He also greets the idea that demographics is destiny with skepticism, pointing out that in the 1930’s Germany’s demographic profile did not indicate youthful bellicosity. Though Shapiro acknowledges the headwinds that demographics will present to both China and Europe, he argues we shouldn’t underestimate their future possibilities. The conversation closes with the possibility that instability and reorganization will result in a ferment of cultural creativity that might match the decades around 1900. Though we are in for a great geopolitical shift, Shapiro sees opportunities and promise in the US, which still remains a dynamic society and a magnet for talent. Finally, he tells us to keep an eye on Central Asia as a locus for instability and change due to both location and authoritarian governments.

On this episode of Unsupervised Learning Razib welcomes back Samo Burja, a guest who needs no introduction for long-time listeners. Burja is the podcast’s first third-time guest, and with good reason. Previously, he came on to discuss social technology and China and lost civilizations, plumbing the depths of the human past for insights about the present and future. Today Burja spotlights a timely new venture of his firm, Bismark Analysis: the Bismarck Brief newsletter, which provides a taste of the sort of “deep-dive” analyses Bujra provides clients (Drone Adoption Favors Quantity Over Quality In Warfare, The German Retreat From Nuclear Power and Modern Russia Can Fight And Win Land Wars).  He discusses the analytic model undergirding the Bismarck Brief, the idea of “live players,” individuals and institutions that can innovate and direct actions in surprising and novel ways, and “dead players,” who tend to operate in a rote manner following predictable scripts, and struggle to meet new challenges. A new start-up in a phase of expansion is a live player, disrupting the marketplace and transforming the notion of what is possible, like Google in 2000. In contrast, Google in 2022 is arguably a dead player, squeezing massive profits out of its capture of online advertising via search, but no longer transforming any sector of the economy. The remainder of the episode shifts to the details of Burja’s analytic process, and his thoughts at the time in February 2022 (when the episode was pre-recorded) on the impending Russian invasion of Ukraine, as well as the earlier performance of Russian military forces in Syria and elsewhere. Razib and Samo touch on the geopolitical consequences of Europe’s energy dependence upon Russia, and its distortionary impact on German foreign policy. Because this episode was recorded before the Russian invasion that began on February 24th, 2022, Razib has circled back with Burja to record a bonus mini-episode tackling recent developments. That episode covers what the Russian invasion and the Western response might mean for the global order in the 21st century.

On this episode of Unsupervised Learning Razib talks to Zack Stentz, a screenwriter and producer in Hollywood, and a former journalist. His credits include 2011 films X-Men: First Class and Thor, as well as the television shows Andromeda, Terminator: The Sarah Connor Chronicles and Jurassic World Camp Cretaceous. Considering that working in Hollywood as a writer is a “dream job” for many, Razib and Stentz discuss how to break in and succeed in show business. Like most people, Stentz wrote in his spare time while pursuing a career as a journalist for many years. His trajectory shifted when he was added to an early-aughts science fiction series, Andromeda (Razib was among its fans). As in many fields, one success can open doors as you become a known quantity and Stentz also developed a partnership with another writer, Ashley E Miller, for over a decade. Unlike what would have been true if Stentz had begun his career a few decades earlier, the 21st-century has seen massive changes in film and television. Early in Stentz’s career, there were still prominent independent films that both made a great deal of money and had a cultural impact, like 2003’s Lost in Translation or 2004’s Eternal Sunshine of the Spotless Mind. Today the film landscape is dominated by a few tentpoles inspired by comic-book movies. Stentz argues that the future of more creative work is probably going to be streaming platforms, with the big screen dominated by the PG-13 “shared universe” series. He also admits that there have been massive changes in the public’s attention span, with a fracturing of the entertainment landscape between clips, series, films and video games. Though there are still opportunities, the changes over the last twenty years have been massive. On the business end, Stentz argues that eventually, movie theaters will also have to be more aggressive about flexible pricing. Then Razib asks Stentz about his opinions on the new Amazon series based on J. R. R. Tolkien’s work that will debut in the fall, The Rings of Power. Stentz looks back to an interview with Peter Jackson from his journalist days in the late 1990’s, and the reverence that he brought to that assignment, and worries that Jeff Bezos and Amazon do not understand they cannot just buy creative excellence. Apparently, Bezos pushed for the purchase of the rights from the Tolkien estate after seeing the success of Game of Thrones, and that prompts Razib to recount his internet interactions with the author of the novels, George R. R. Martin, in the late 1990’s. Stentz notes that Martin left a career as a television writer to write novels because he wanted more freedom, so it was somewhat ironic that he became involved in television again fifteen years after he left due to the popularity of his novels. Finally, Stentz addresses the cultural changes in media, Hollywood and the new focus on representation. Stentz takes a moderate stance on the changes, neither promoting them with full force nor resisting inevitable change. He argues that people have to focus on building their own brand and uniqueness, as providing genuine value is the best job protection.

On this episode of the Unsupervised Learning podcast Razib talks to his friend Sarah Haider, founder of Ex-Muslims of North America and the writer behind a new Substack, Hold That Thought. Born in Pakistan, and raised in Texas in a Shia Muslim family, Sarah came to prominence in 2015 after she gave a speech called "Islam and the Necessity of Liberal Critique" at The American Humanist Association's 74th annual conference.  Razib and Sarah first discuss where the Ex-Muslim community is in 2022, especially after a few years of COVID-19 that dampened face-to-face meetups. Sarah argues that there has been a massive change in the acceptance and visibility of Ex-Muslims over the past decade. To a great extent the community has taken on a life of its own, and the shepherding role of  Ex-Muslims of North America is not nearly as essential as it was when it was founded in 2013. It should be noted here that public apostasy in Islam is traditionally a capital crime, so people who are skeptical of religion from a Muslim background have often been wary of being open and honest about their views. Sarah also observes that Ex-Muslims have weathered the cultural changes on the social Left better than the broader atheist and secular community over the last few years, which has been fractured by the rise of the “woke” faction. Razib asks if this is possibly due to a high degree of disagreeability among Ex-Muslims, who are often strongly selected for nonconformity in order to be willing to go against the grain, even at great personal and social cost. Sarah agrees and offers that this is in contrast to what she terms the “evangelical to woke” pipeline, where conservative Protestants leave their religion behind only to adopt an entirely new set of strident beliefs (on her Substack Sarah has a post Is Wokism uniquely Christian?). Razib questions Sarah as to whether the decline of religion over the last generation that New Atheists were cheering for in the aughts was really a force for good. While they were discussing various heresies, Sarah offered that she was a “gender atheist.” She takes a highly critical view of what she calls “gender ideology,” asserting the primacy of biological sex. Razib and Sarah then mull over whether this is truly the current age’s biggest third rail, and why that might be. Eventually, Razib stumbles on the fact that Sarah is a “Rogan-bro” (she listens semi-regularly to the podcast). She discusses why people listen, Joe Rogan’s relationship to his audience, and explains how she came to be a listener and why she continues to tune in. She also discusses her piece Why Deplatforming Joe Rogan Will Backfire. Finally, Sarah talks about what it’s like to be a young woman who has a public online presence, from excessive attention to harassment from conservative Muslims. They close out with a discussion about future directions in her career and activism. Sarah talks about why she started the Substack and why she is now pursuing writing and thinking on topics outside her secular/Ex-Muslim bailiwick.

Today on the Unsupervised Learning podcast the focus is on genetics, culture and geopolitics with Muhammad Sohail Raza, a Pakistani genomicist living and working in Beijing, China, whose research focuses on bioinformatic methods and high-altitude adaptations. Razib and Muhammad first discuss how he got interested in biology, and what took him to do his graduate work in the People’s Republic of China. Muhammad talks about his various inspirations, in particular David Reich’s work on historical population genomics, as well as the potential promise of precision medicine in the domain of healthcare. About a decade ago, when his interest in genetics began, Muhammad was particularly focused on the importance of bioinformatics, and he outlines how Chinese academia is very strong in understanding the engineering and methods of data generation in a genomic context, due to China’s position as a sequencing leader. Razib and Muhammad then explore the numerous professional opportunities in China’s coastal megacities, Beijing, Shanghai, and Shenzen, and Muhammad recalls his experience with the locals, who were friendly, open and curious. Beijing in particular is quite diverse, with scientists from Europe and America, as well as those from China and other parts of Asia, Africa and Latin America. Razib was skeptical about the Chinese attitude toward a brown-skinned person with a Muslim name, but Muhammad’s experience has been that Chinese of all backgrounds are quite accepting once they realize he has some command of Mandarin (Standard Chinese). Additionally, the world of science is multicultural and cosmopolitan, and when the focus remains on scholarship there are far fewer tensions than might occur in other contexts, like business or politics. Muhammad contrasts his experience in China with how Asian researchers perceive the United States. Because of his Pakistani nationality, he was denied a visa to attend an American conference, while Chinese researchers feel that geopolitical tensions are casting a pall over their collaborations. Though the Chinese opinion of American science remains high, the prospects for future cooperation have been dampened by the new rivalry between China and the US. Finally, Muhammad talks about research in high-altitude genomics and the adaptations of Tibetans in particular. He explains that future directions in this field will have less to do with hypoxia, as opposed to the metabolic adaptations associated with it. Due to the paucity of ancient DNA, most of the analysis is going to be on large cohorts of contemporary Chinese. This means that the Beijing Institute of Genomics, where Muhammad works, will likely require all 40 petabytes of storage available at his research institute at some point.

Today on the Unsupervised Learning podcast Razib talks to Suhag Shukla, the Executive Director of the Hindu American Foundation (HAF). Suhag is an attorney who grew up in Cupertino, California, and is now a leading advocate for the interests of American Hindus. Razib and Suhag clear up the fact that HAF does not speak for all Hindus, of whom there are over one billion, or, the world’s 1.4 billion Indians. Additionally, the HAF is an explicitly Hindu-focused organization, as opposed to an Indian-American one. Suhag notes that Indian Americans, just like Indians, are religiously diverse, and there are also Hindus of various ethnicities, including Julia Roberts, a white American woman, and Tulsi Gabbard, who is of European and Samoan heritage. Despite that, Suhag does admit that Hinduism as a religion is somewhat different from American Protestantism, which is focused on belief and individual choice in belonging. Hindus have a strong cultural attachment to their folkways and traditions, distinct from their particular beliefs about God and the universe. In this way, Hinduism resembles Judaism. Suhag points out that there is even an atheistic movement that emerged out of Hinduism, Carvaka materialism. One of the reasons that HAF exists is to educate Americans about Hinduism because aside from stereotypes about elephant Gods, sacred cows, and vegetarianism, most people in the US know very little about the religion. When Suhag went to Congress in 2006 and explained she represented American Hindus, the representative asked “are you, Sunni or Shia?” Though HAF’s purview is America, it often gets dragged into Indian cultural politics. Suhag addresses accusations that HAF gives comfort to Hindu nationalists and Hindutva, and outlines for American listeners what these ideologies actually are. Indian politics and society are complex topics, and Suhag and Razib barely scratch the surface, but they do touch specifically on the issue of caste and how it is now becoming a live issue in the US. Over the past year, activists have been demanding that caste be included as a protected class, due to allegations of discrimination in the US. Suhag and Razib both agree that at 1% of America’s population it is unlikely that American Hindus are recreating the same stratified social system that obtains in India. Additionally, Razib points out that while white American Hindus are never going to be asked about caste, brown-skinned non-Hindus like himself will be, because the average American tends to racialize Hindu identity. They close out the discussion by tackling whether Hinduism can adapt to the American landscape, where conversion and fluidity are the norms. Suhag is optimistic about the future of Hinduism and points to the fact that the data show high retention rates of people raised as Hindus in the US to the religion as adults. She also offers evidence that Indian Americans raised in a more assimilated environment often come back to their ancestral traditions later in life, and that includes their religion.