Interview
Razib Khan - Genomics, Intelligence, and The Church of Science
- Long-term equilibrium for intelligence is expected to shift due to negative selection on educational attainment genes, a process driven by societal incentives reducing reproductive output among highly educated individuals that may take centuries to complete.
- Brain size reduction in humans is projected to correlate with climatic warming and the nutritional demands of agriculture, favoring smaller brains for survival despite associated physical trade-offs like weaker bones.
- Mendelian diseases such as cystic fibrosis and sickle cell anemia are anticipated to be cured by gene editing within the next 10 years, with potential polygenic editing for intelligence following within 20 years, focusing initially on correcting de novo mutations and copy errors across thousands of SNPs.
- Future human evolution and population genetics are characterized by specific trends, including 99.5% endogamy per generation among Indian jatis, distinct lineage traits in the Nair and Haka groups, and the potential emergence of Fujianese Chinese cultural output contingent on government policy during industrialization.
- Regional differences in elite behavior are noted, with South Indian Brahmins expected to remain overrepresented in big tech leadership due to historical migration and colonial pipelines, while elites in Uttar Pradesh and Bihar are predicted to favor local, zero-sum power structures over global risk-taking.
- The biological variance in intelligence between sexes is hypothesized to stem from the single X chromosome in males exposing deleterious alleles and developmental instability during masculinization.
- Academic science is forecast to become an increasingly exclusive "winnowing profession" with extreme Pareto distribution (5-95), where a shift from truth-seeking to comfort risks institutional dissolution and where publication counts continue to distort actual scientific quality.
- Civilization is described as metastable with risks of collapse leading to knowledge loss in critical infrastructure, such as the inability to maintain legacy software, necessitating technical revival to prevent a long twilight of decline.
- Historical progress is generally viewed as a gradual exponential curve, though the current steep trajectory in information and social technology could constitute a step function leading to radical transformation or regression within the next century.
- The self-domestication hypothesis linking genetic changes to reduced aggression and intelligence is expected to remain unsupported by genomic evidence despite its popularity.
- Technical writing strategies are projected to succeed by applying machine learning to domains of broad interest like personality or genetics to ensure engagement.