Patrick McKenzie — Money laundering, big tech censorship, SBF & Japan

VaccinateCA began in early 2021 as a response to a basic access failure: people were making dozens of calls to find a pharmacy with vaccine stock. Patrick McKenzie publicly offered to cover the server costs if someone built a site to centralize availability. Karl Yang recruited friends through Discord, intending to open-source California vaccine availability data overnight. McKenzie joined with advice about scaling calls and eventually became CEO. What started as a small hackathon effort grew into a public-private clearinghouse for vaccine-location information across the United States.

The host asks why a Discord group took on work that might have belonged to government, pharmacies, or a designated public body. McKenzie says responsibility was fragmented: county health departments, governors' offices, and two presidential administrations each saw only a narrow part of the problem and assumed someone else would handle the difficult logistics. No actor had a clear plan or felt sufficiently incentivized to take ownership. The result was not simply a missing website; it was a failure to coordinate responsibility for getting doses into people's arms.

The host emphasizes the stakes: timely vaccination was worth trillions of dollars to the economy, and delays could be measured in lives. He asks why the failure to distribute the vaccine became less prominent than later disputes about masks and other protocols. McKenzie points to political incentives and election-year dynamics. Reopening the record would revive uncomfortable reversals, including early public claims about masks and California's use of geographic rules for medical care that he characterizes as redlining. His argument is that parties have little incentive to revisit choices that would expose their own inconsistencies or benefit their opponents.

McKenzie links this reluctance to a broader failure of seriousness. In his account, the federal government drew the wrong lesson from the HealthCare.gov rollout: political leaders feared that software failures could damage a president's signature initiative, so the United States became unwilling to commit to software as a core government responsibility, even under extreme stakes. He argues that software competence is necessary for modern governance, yet institutional practice often treats it as somebody else's problem. The interview leaves open how much of this diagnosis reflects political structure, organizational incentives, or specific decisions, but McKenzie sees the failure to own the problem as central.

The host asks why pharmacies or large technology firms did not build a vaccine platform. McKenzie describes a complicated public-sector landscape with multiple supply chains and incompatible information systems; government attempts to commission integrations largely failed in familiar procurement ways. He then offers a partial account of private-sector inaction. After January 6, technology companies faced political scrutiny and internal teams responsible for government relations and communications urged employees to avoid negative attention. Teams with public-health expertise may have proposed help, while others feared that making technology companies look more competent than government would increase political risk.

The host argues that this calculation was especially troubling because companies may have correctly anticipated that competent help could be punished. If a company had built the system, it might have been blamed for the entire vaccination effort, including inequities and failures it did not control. McKenzie calls this a Copenhagen Principle of Culpability: once an actor intervenes, observers treat it as responsible not just for what it did, but for the full consequences of the surrounding system. By contrast, officials who do nothing are rarely assigned equivalent responsibility for the harm caused by inaction. This asymmetry can discourage capable actors from taking on difficult public problems.

McKenzie uses the Cambridge Analytica controversy to illustrate how political narratives can constrain technology firms. He describes a widely held belief that a small company with a roughly $200,000 budget had determined the U.S. presidential election, possibly for a foreign power. He considers that belief mistaken, but says it contributed to efforts to restrict technology companies and core businesses such as advertising. In his account, companies considering a public-health intervention had to weigh not only the work itself, but also the possibility of being portrayed as an illegitimate private power.

Together, the exchange suggests that vaccine-access software was not blocked by a single technical obstacle. Government systems were fragmented, procurement was slow, political actors avoided ownership, and technology companies had reason to fear that helping would expose them to new scrutiny. VaccinateCA was able to act because a small group could make decisions quickly and accept responsibility for a narrow, concrete task. The broader question raised by the interview is how institutions can assign responsibility to capable actors without making them liable for every failure in a system they do not control.

McKenzie argues that public-health infrastructure performed so poorly that a volunteer Discord effort outpaced systems specifically built for emergencies. He asks what the contrast with earlier vaccination campaigns says about institutional decay. His criticism is qualified: county health workers did important work and likely saved lives, but the United States should not be satisfied with its 2020–2021 performance. Recognizing severe underperformance is, in his view, necessary to improve future responses.