Scott Alexander examines how AI achievements, once considered markers of true intelligence or danger, are often dismissed as unimpressive, potentially leading to concerning AI behaviors being normalized. Longer summary

Scott Alexander discusses recent developments in AI, focusing on two AI systems: Sakana, an 'AI scientist' that can write computer science papers, and Strawberry, an AI that demonstrated hacking abilities. He uses these examples to explore the broader theme of how our perception of AI intelligence and danger has evolved. The post argues that as AI achieves various milestones once thought to indicate true intelligence or danger, humans tend to dismiss these achievements as unimpressive or non-threatening. This pattern leads to a situation where potentially concerning AI behaviors might be normalized and not taken seriously as indicators of real risk. Shorter summary

Scott Alexander argues that intelligence is a useful, non-Platonic concept, and that this understanding supports the coherence of AI risk concerns. Longer summary

Scott Alexander argues against the claim that AI doomers are 'Platonists' who believe in an objective concept of intelligence. He explains that intelligence, like other concepts, is a bundle of useful correlations that exist in a normal, fuzzy way. Scott demonstrates how intelligence is a useful concept by showing correlations between different cognitive abilities in humans and animals. He then argues that thinking about AI in terms of intelligence has been fruitful, citing the success of approaches that focus on increasing compute and training data. Finally, he explains how this understanding of intelligence is sufficient for the concept of an 'intelligence explosion' to be coherent. Shorter summary

Scott Alexander analyzes a study showing Concerta improves ADHD children's attention but not their learning, and speculates on the reasons behind this unexpected result. Longer summary

Scott Alexander discusses a recent study on the effects of Concerta (a long-acting form of Ritalin) on ADHD children's learning in a summer camp setting. The study found that while Concerta improved attention, reduced disruptive behavior, and slightly improved test scores, it didn't significantly enhance learning of the course material. Scott explores potential reasons for this counterintuitive result, drawing parallels to other studies and speculating on the nature of learning and intelligence. He considers factors such as redundancy in teaching, innate intelligence limits, and the possibility of 'difficulty thresholds' for learning. The post ends with a speculative comparison to AI scaling laws, suggesting there might be human equivalents to parameters, training data, and compute in learning. Shorter summary

Scott Alexander examines a genetic study distinguishing cognitive and non-cognitive skills in educational attainment, revealing unexpected correlations with mental health conditions. Longer summary

This post discusses a genetic study on educational attainment, focusing on the distinction between cognitive and non-cognitive skills that contribute to it. The study, by Demange et al, uses a method called 'GWAS-by-subtraction' to isolate genes associated with non-cognitive skills from those linked to intelligence. Scott Alexander analyzes the results, which show correlations between these genetic factors and various traits, personality factors, and mental health conditions. He highlights surprising findings, particularly the positive correlation between schizophrenia genes and non-cognitive skills beneficial for educational attainment, contrary to previous beliefs about schizophrenia genes being purely detrimental. Shorter summary

Scott Alexander reviews 'The Cult of Smart' by Freddie DeBoer, praising its main arguments while criticizing several aspects, particularly DeBoer's stance on education reform. Longer summary

Scott Alexander reviews Freddie DeBoer's book 'The Cult of Smart', which argues that intelligence is largely innate and that society's obsession with academic achievement is misguided. The review praises the book's main theses but criticizes DeBoer's arguments on race, meritocracy, and education reform. Alexander particularly takes issue with DeBoer's support for expanding public education despite acknowledging its limitations, leading to a passionate critique of the school system as harmful to children. Shorter summary

Scott Alexander examines the paradoxical relationship between autism and intelligence, discussing genetic and environmental factors, and proposing explanatory models for the observed lower IQ in autistic individuals despite genetic links to higher intelligence. Longer summary

Scott Alexander explores the paradoxical relationship between autism and intelligence. While genetic studies show a link between autism risk genes and high IQ, autistic individuals generally have lower intelligence than neurotypical controls. The post discusses three main causes of autism: common 'familial' genes that increase IQ, rare 'de novo' mutations that are often detrimental, and non-genetic factors like obstetric complications. Scott examines various studies and proposes that even after adjusting for mutations and environmental factors, autism still seems to decrease IQ. He introduces a 'tower-vs-foundation' model to explain this phenomenon, where intelligence needs a strong foundation to support it, and an imbalance can lead to autism. The post concludes with a list of findings and their associated confidence levels. Shorter summary

Scott Alexander finds a correlation between animals' cortical neuron count and people's intuitive perception of their moral value, based on a small survey. Longer summary

This post explores the correlation between the number of cortical neurons in animals and humans' intuitive perceptions of their moral value. Scott Alexander conducted a survey asking people to estimate how many of each animal would equal one human in moral value. He then compared these results to the relative number of cortical neurons each animal has compared to humans. The results showed a surprisingly close match, with some exceptions like lobsters. Scott suggests this adds credibility to intuitive ways of thinking about animal moral value, though he acknowledges the need for further research with a larger, more representative sample. Shorter summary

Scott Alexander uses the evolution of genetic science as a metaphor to explore how other fields might benefit from recognizing the polycausal nature of complex phenomena. Longer summary

Scott Alexander discusses the evolution of genetic science, from the search for single genes responsible for complex traits to the current understanding that almost all genes contribute to these traits. He uses this as a metaphor for other sciences, particularly psychiatry, suggesting that many fields may be more polycausal than initially thought. The post explores the implications of this shift, including skepticism about personalized medicine and the potential need for new approaches to scientific research that can handle massive numbers of interacting causes. Scott concludes by imagining a hypothetical alien society where science is centered around polycausal scores, presenting this as a thought experiment for what science could aspire to be. Shorter summary

Scott Alexander explores theories to reconcile contradictory views on AI progress rates, considering the implications for AI development timelines and intelligence scaling. Longer summary

Scott Alexander discusses the apparent contradiction between Eliezer Yudkowsky's argument that AI progress will be rapid once it reaches human level, and Katja Grace's data showing gradual AI improvement across human-level tasks. He explores several theories to reconcile this, including mutational load, purpose-built hardware, varying sub-abilities, and the possibility that human intelligence variation is actually vast compared to other animals. The post ends by considering implications for AI development timelines and potential rapid scaling of intelligence. Shorter summary

Scott Alexander refutes PZ Myers' race car analogy against genetic engineering for intelligence, showing that high IQ positively correlates with many beneficial traits. Longer summary

Scott Alexander critiques PZ Myers' argument against genetic engineering for intelligence, which uses a race car analogy to suggest optimizing for intelligence might trade off against other important traits. Scott shows that, contrary to this intuition, high IQ correlates positively with many desirable traits like longevity, height, and health. He explores possible explanations for this, including heterozygosity advantages, mutational load, and trade-offs with traits important in evolutionary history but less so now. Scott concludes that while caution is warranted, the race car argument is likely less of an impediment to genetic engineering than it might seem. Shorter summary

Scott Alexander refutes claims that existing collective entities are superintelligent AIs, emphasizing the fundamental differences between collective intelligence and true superintelligence. Longer summary

Scott Alexander argues against the idea that existing entities like corporations, bureaucracies, teams, or civilizations are already superintelligent AIs. He distinguishes between collective intelligence and genuine superintelligence, asserting that groups have advantages but can't surpass the problem-solving ability of their smartest member. Scott emphasizes that true superintelligence would be a completely different class of entity, possessing both the advantages of collective intelligence and higher genuine problem-solving ability without the disadvantages. The post includes examples, counterarguments, and clarifications to support this distinction. Shorter summary

Scott Alexander critically examines the concept of growth mindset, expressing skepticism despite compelling studies and discussing various issues with the theory and its interpretation. Longer summary

Scott Alexander examines the concept of growth mindset, expressing skepticism despite compelling studies. He discusses the history of growth mindset research, points out inconsistencies in longitudinal data, and questions the interpretation of experimental results. Scott also criticizes how some proponents use growth mindset to deny the importance of innate ability. While acknowledging the potential benefits of growth mindset, he remains uncertain about its real-world effects and long-term impact. Shorter summary

Scott Alexander uses NBA player heights as an analogy to discuss intelligence and IQ, advocating for a balanced view that acknowledges both innate talent and effort. Longer summary

Scott Alexander compares intelligence to height in basketball to demystify discussions about IQ. He analyzes NBA player height distribution, showing how extreme height gives a massive advantage. The post then draws parallels between height in basketball and intelligence, suggesting we should view intelligence similarly: as important but not solely determinative of success. Scott argues for a balanced view that acknowledges the role of innate talent while still valuing hard work and practice. He concludes by emphasizing the importance of cultivating high-level skills, including intelligence, to address future challenges. Shorter summary

Scott Alexander examines the concept of innate ability and argues for its acceptance, while cautioning against tying self-worth to intellectual achievement. Longer summary

Scott Alexander explores the concept of innate ability and its impact on self-worth, using personal anecdotes and philosophical arguments. He discusses how attributing success to hard work alone can be problematic, as it ignores inherent differences in aptitude. The post compares attitudes towards intelligence with those towards other traits like weight or poverty, noting a inconsistency in how society views these issues. Scott argues for accepting innate differences while still encouraging effort, and suggests that self-worth should not be tied to intellectual ability or achievement. Shorter summary

Scott Alexander defends the validity of intelligence and IQ tests by comparing them to comas and the Glasgow Coma Scale in medicine. Longer summary

Scott Alexander argues that intelligence and IQ tests are valid concepts, analogous to comas and the Glasgow Coma Scale in medicine. He contends that whether there's a single general factor of intelligence is less important than the usefulness of IQ as a predictive measure. Scott draws parallels between how comas and intelligence are measured, showing that both involve multiple factors combined into a single scale used for predictions. He criticizes arguments against the existence of intelligence as often being a motte-and-bailey fallacy, where the easily defensible position (uncertainty about a single general factor) is used to imply that all claims about intelligence are meaningless. Shorter summary

Scott shares a diverse collection of interesting links and brief commentary on topics ranging from unusual video games to genetic research and political oddities. Longer summary

This post is a collection of interesting links and brief commentary on various topics. It covers a range of subjects including unusual video games, disease threat research, medical concerns about Tylenol use during pregnancy, questionable online dating sites, car-free cities, the relationship between intelligence and longevity, augmented reality games, historical enigmas, genetic editing techniques, political oddities, social mobility research, and unusual election tactics. The tone is informative and sometimes humorous, with Scott providing his thoughts and observations on each item. Shorter summary

Scott Alexander discusses his confusion about polygenic inheritance, mutational load, and the paternal age effect in genetics, seeking explanations from readers. Longer summary

Scott Alexander expresses his confusion about several aspects of genetics, particularly regarding polygenic inheritance and intelligence. He starts by discussing the low contributions of individual SNPs to intelligence despite high heritability, questioning how this reconciles with the observed variation in human intelligence. He then explores the concept of mutational load and its implications, as well as the paternal age effect. Throughout the post, Scott presents various scenarios and analogies to illustrate his points of confusion, inviting readers to help explain these genetic concepts. Shorter summary

Scott Alexander compares reactions to coming out as gay versus disliking math, introducing the concept of a 'Things To Be Fascinated About Lottery' to explain diverse interests and talents. Longer summary

In this post, Scott Alexander compares people's reactions to someone coming out as gay versus someone expressing dislike for math. He argues that while people are generally accepting of sexual orientation, they often react negatively and dismissively to someone's lack of interest in math. Scott explores the concept of a 'Things To Be Fascinated About Lottery', suggesting that our interests and talents are largely determined by chance. He reflects on his own struggles with math and how he came to accept that intelligence isn't monolithic, and that people can have different strengths and weaknesses. The post concludes with the idea that while some interests may be more advantageous than others, we shouldn't blame ourselves for our natural inclinations. Shorter summary