Episode Summary
Executive Summary: The episode argues that science, like finance, is a consensus system with broken incentives. Brian Armstrong and Patrick Joyce explain how ResearchHub uses crypto tools—tokens, open governance, bounties, and reputation—to reward high-quality scientific work, improve discovery and peer review, and enable more open, commercializable, and decentralized science.
Main Topics: Science as a consensus technology (Priority: 5/5): The hosts frame science as a repeatable, testable method for building shared truth, comparable to a blockchain-style consensus mechanism for reality. They emphasize empiricism, falsifiability, and replication as the core of scientific progress. Why the current scientific system is broken (Priority: 5/5): The guests describe modern science as distorted by citation incentives, publication pressure, risk aversion, institutional gatekeeping, and poor capital allocation. This leads to fewer negative results, more gaming of metrics, and weak translation from research to real-world products. How crypto and DeSci can help (Priority: 5/5): ResearchHub is presented as a DeSci platform that uses tokens and community governance to financially reward valuable scientific contributions, such as paper summaries, peer review, bounties, and funding decisions. ResearchHub product model (Priority: 4/5): The conversation details ResearchHub’s Reddit-like feed, ERC-20 token incentives, bounties for scientific tasks, reputation systems, and experiments in funding research and simplifying commercialization through standardized licensing. Open science, openness, and decentralization (Priority: 4/5): The guests connect DeSci to the open science movement, arguing that open code, open data, and community-controlled rules can make science more auditable, forkable, and accessible to more participants worldwide. Future of science and commercialization (Priority: 4/5): The discussion envisions a future where scientific claims are auditable with public data, anyone can contribute from anywhere, and breakthroughs can be licensed and commercialized with far less friction and legal overhead.
Key Arguments: Science is a truth-finding protocol built on observation, hypothesis, and falsification, not blind trust in authorities. Modern science is distorted by Goodhart’s Law: citations and publications became targets, so they are increasingly gamed. Negative results are underreported because they are less rewarding for careers and funding, creating a skewed scientific record. Scientists are incentivized to be risk-averse and optimize for grants instead of pursuing bold or potentially transformative research. Crypto can align incentives in science by using tokens and governance to reward useful contributions directly rather than relying on proxy metrics. Open source and blockchain-style transparency can make science more collaborative, auditable, and community-driven. Commercialization of breakthroughs is inefficient today because universities and tech transfer offices impose high-friction, one-off negotiations and often capture most of the upside. A better system would let researchers retain more value and allow outsiders to license innovations easily through standard terms. DeSci can broaden participation so valuable ideas can come from anywhere, not just elite universities or a few geographic hubs. The long-term goal is not just better publishing but a better economic system for producing, validating, funding, and translating science into products.
Data Points: Scientific papers published per year: 2 to 3 million - Brian Armstrong cites the scale of annual scientific publishing as evidence that most papers never get read or commercialized. University overhead in research funding: About 50% - Brian says roughly half of research funding can go to university overhead instead of the researcher or the discovery. Research papers with negative results in the 1990s: About 40% - Patrick says negative-result papers were roughly 40% of the literature in the 1990s. Research papers with negative results today: About 20% - Patrick says the share of negative-result papers has fallen to around 20% now. Probability of first-year PhD students becoming research professors: Less than 2% - Patrick uses this to explain the intense competition and incentive to optimize for publishable output. Theoretical gravity acceleration: 9.81 meters per second squared - Patrick uses this as an example of a highly reproducible scientific claim. ResearchCoin supply distributed annually: About 5% - Brian says roughly 5% of RSC supply is being distributed per year, with further decentralization expected over time. Commercialization rate of papers: 99.9% do not get commercialized - Brian argues that almost all papers never become products, despite the potential of the few breakthroughs that do. Stanford tech transfer revenue concentration: Mostly from Google and Genentech - Brian says Stanford’s historical tech transfer revenue has been dominated by a tiny number of major breakthroughs.
Pivotal Quotes: "the model of the first scientific journal was actually Nelson Verba, which means take no one's word for it." — Patrick Joyce: Used to contrast real science with the slogan 'trust the science' and emphasize skepticism and testing. "Goodhart's Law. So, the idea that a measure ceases to be a good measure when it becomes a target for behavior." — Patrick Joyce: Explains why citation counts and publication volume can be gamed once they become the main success metric. "science is at its best when there can be an accepted theory that has a lot of evidence, and somebody comes up with a new theory, and it's testable, it's falsifiable" — Brian Armstrong: Describes the ideal scientific process as open to challenge and revision through evidence.
Implications: If DeSci succeeds, science could become more open, auditable, and economically aligned: better peer review, more funding for valuable work, easier commercialization, and broader participation beyond elite institutions.