Mark Zuckerberg Interviews Patrick Collison and Tyler Cowen on the Nature and Causes of Progress (Bonus)

Episode Summary

Executive Summary: Mark Zuckerberg, Tyler Cowen, and Patrick Collison argue that modern progress is real but slowing in many domains, making a new “science of progress” urgent. They focus on how institutions, incentives, and funding structures shape innovation, and highlight under-studied bottlenecks in science, housing, health care, and mobility. The conversation blends historical examples, empirical data, and philanthropy strategy to ask how society can generate more useful knowledge and faster improvement.

Main Topics: Why progress should be studied as a distinct subject (Priority: 5/5): The speakers argue that progress is not constant over history and that understanding its causes could help society solve problems faster. They frame progress studies as a broad, interdisciplinary agenda rather than a single new field. Signs that progress is slowing (Priority: 5/5): They point to falling productivity growth, rising costs in health care and education, and infrastructure delays as evidence that many sectors now produce less progress per dollar than in the past. Science funding and research productivity (Priority: 5/5): The discussion explores whether scientific institutions, especially NIH-style grant systems, are optimally structured. They argue that too much bureaucracy and too little tool-building may be slowing discovery. Housing, mobility, and cost disease (Priority: 4/5): Housing restrictions and urban bottlenecks are presented as major brakes on opportunity and economic growth. They also connect housing, rent, education, and health care as domains where costs rise faster than value. Philanthropy, experimentation, and long-term bets (Priority: 4/5): Zuckerberg and Collison discuss how foundations and initiatives like CZI and Emerging Ventures can support long-horizon, risky, nontraditional projects that conventional grant systems avoid. Historical clusters of creativity (Priority: 3/5): Examples such as Vienna, Edinburgh, and late-19th-century American universities are used to suggest that progress often emerges from dense, collaborative, and reform-minded environments.

Key Arguments: Progress is historically non-linear; the post-Industrial Revolution acceleration is unusual and worth explaining. A small decline in growth or productivity compounds dramatically over time, making slow progress an “invisible crisis.” Economic growth is not always benign, but more resources and knowledge make it easier to solve climate change, poverty, and debt. Current scientific systems may be producing large inputs but only constant or diminishing outputs, implying low efficiency. Management practices, urban housing supply, and science funding all appear to have large potential returns if studied and reformed. Many institutions have become risk-averse and bureaucratic, reducing experimentation and making it harder to create new universities, build infrastructure, or reorganize science. The most valuable philanthropy is long-term, willing to take risks, and focused on enabling new capabilities rather than only funding existing structures. Collaborative, high-trust cultures and immigration flows have often coincided with major bursts of creativity and institutional innovation. Academics are incentivized to produce publishable “bricks,” not to assemble larger, practical “buildings” that solve real-world problems. Zuckerberg sees a role for technology and tool-building in compounding scientific progress, while Cowen stresses that understanding historical and institutional patterns is necessary but insufficient without action.

Data Points: U.S. productivity growth (1920-1970): about 1.9% per year - Tyler contrasts mid-century productivity growth with today’s much slower pace. U.S. productivity growth today: around 0.4% per year - Tyler cites current estimates to illustrate slower progress. Extreme poverty reduction since Zuckerberg was born: more than 1 billion people - Used to show that progress has delivered enormous global benefits. Global life expectancy increase since Zuckerberg was born: about 6 years - Evidence of progress in health outcomes. Infant mortality decline since Zuckerberg was born: more than 50% - Illustrates measurable social progress. Return to foreign aid plus coaching: 100% to 400% - Tyler cites a Duflo and Banerjee paper as an example of high-value evidence worth debating intensely. People moving when taking a new job in the U.S. (1980): about 40% - Used to show higher geographic mobility in the past. People moving when taking a new job in the U.S. (last year): about 1 in 10 - Shows declining mobility due to high housing costs. Bay Area/New York housing counterfactual GDP effect: $3,700 more per person - Collison cites Shea and Moretti’s model of more elastic housing supply. NIH funding by age (1980): 12x more dollars to researchers under 40 than over 50 - Shows younger scientists were once more strongly supported. NIH funding by age (today): 5x more dollars to researchers over 50 than under 40 - Indicates a reversal toward older scientists. NYC subway construction (1900 project): 23 stations in 4.7 years for just over $1 billion (2019 dollars) - Example of past infrastructure speed and efficiency. NYC Second Avenue subway: 3 stations in 17 years for $4.5 billion - Used to illustrate modern infrastructure productivity decline. Golden Gate and Bay Bridge construction: Both completed within 4 years - Example of rapid large-scale public works in the 1930s. Man-made island in San Francisco: Completed in 18 months - Used to underscore how quickly major projects once moved. California high-speed rail timeline: Started 11 years ago; projected completion in 2033 - Example of slow contemporary infrastructure delivery. Top 25 universities: 7 American universities founded in a 30-year span (1861-1891) - Used to show historically concentrated institutional innovation. Science inputs since the first half of the 20th century: about 50x to 100x larger - Active PhDs, papers, and funding all rose sharply, but outputs did not rise proportionally. Health care spending share of GDP: about 18% - Tyler notes rising spending without matching improvements in life expectancy. SG healthcare spending share of GDP: about 4% - Used as a comparison for more efficient health care delivery. Singapore life expectancy: slightly higher than the U.S. - Supports the claim that lower spending can still yield strong outcomes. Top 1% paper likelihood at HHMI: about twice as likely - Pierre Azoulay study comparing similarly situated scientists with and without HHMI grants.

Pivotal Quotes: "We sort of transitioned from this regime where we weren't making much progress to one where we have been making much more. Is this the best we can do?" — Patrick Collison: Opening framing of progress studies and the possibility of much faster future improvement. "I would rather address those problems with more resources rather than fewer, whether it is paying off our debts, addressing climate change, fixing global poverty." — Tyler Cowen: Why economic growth matters even if progress can create harms. "The incentive is not to build a building in most cases. ... The bricks are there in the millions, but the bricks in the buildings are a different thing." — Tyler Cowen: Why academia underproduces large, integrated solutions despite abundant papers.

Implications: Listeners should see progress as fragile, uneven, and institution-dependent. The conversation suggests that better incentives, more experimentation, and bolder long-term funding could unlock faster gains in science, housing, and public welfare.