How to cultivate outlier scientific talent

James W. Phillips and Laura Ryan are former neuroscientists who’ve written a proposal to save British science by basically blowing up the university system. Or at least building an alternative to it. Their diagnosis? The best scientists they know have all quit academia—not because they failed, but because they succeeded and realised the game is rigged. The incentive structure rewards safe, incremental research that gets published quickly rather than ambitious, years-long projects that might actually change the world. Frederick Sanger won two Nobel Prizes while publishing three papers in 20 years. Today he’d never get tenure.

Their solution is Lovelace Labs—a network of institutions modelled on Bell Labs, Xerox PARC, and the Cambridge LMB, where scientists would be core-funded for 15 years, assessed internally by colleagues who understand their work, and freed from the tyranny of grant applications and citation metrics. Where engineers work alongside theorists, where 30-year-olds run labs instead of spending a decade as research assistants, and where the founding director gets told by Number 10: “Here’s your money, we’re not going to mess around.”

Tom and Calum discuss with James and Laura:

• Why the smartest scientists quit: Laura’s smartest friend from her Cambridge PhD—someone who always wanted to be a scientist—left because the system is fundamentally unfair. James’s entire cohort of rising stars, the people doing work featured in the New York Times, have all left academic research except one,
• The replication crisis stems from broken incentives: Foundational Alzheimer’s research papers were fraudulent for 25 years because everyone benefits from piggybacking on existing results rather than exposing problems. Brain imaging studies lacked statistical power but it took 20 years for that to become common knowledge,
• Leo Szilard’s 1948 prophecy: He wrote a satirical story about a wealthy man who wanted to slow down science, so he invented peer review—pulling scientists out of labs into administration and forcing everyone to work on ideas that three peers would approve, killing all unusual fresh shoots,
• Peter Higgs couldn’t survive today: He published sparingly over 20 years, doing deep work that eventually won a Nobel Prize. Today’s system demands papers every six months with positive results—negative data is considered “time wasted” even if it’s exemplary science,
• China has overtaken us on neuroscience: Nine of the top 10 institutions in leading journals are now Chinese (it was two five years ago). Their packages to recruit talent: “Come over, we’ll give you your own lab, strong core facilities, hire whoever you want.” The UK’s pitch: “But we have Oxford!”,
• The Number 10 science establishment blocked honors: During the pandemic, two researchers (Bonner and Kataraman) created the rapid testing program with modeling that proved crucial. The science establishment blocked their honors and gave them instead to senior people who’d been blocking the rapid testing program,
• Alan Kay was 30 at Xerox PARC: When James asked him about top-down direction, Kay revealed he was the oldest person there at 30. In the UK, these people would still be postdocs working as research assistants. Demis, Dario, Sam Altman—all in their 30s when founding DeepMind, Anthropic, OpenAI,
• Max Perutz’s recipe for great science: “No politics, no committees, no reports, no referees, no interviews—just gifted, highly motivated people picked by a few people of good judgment.” The Cambridge LMB followed this and produced Nobel Prize after Nobel Prize,
• The UK over-indexes on universities: We rely more heavily on the university department model than almost any other advanced science nation. Germany has Max Planck and Fraunhofer. America has DOE labs and tech company research. We have... more universities in Midlands towns acting as jobs programs,
• Westminster ejects the misfits: James was part of the Cummings misfits experiment. As soon as key supporters left Number 10, the team began leaving. The Vaccines Task Force was crushed, the data science unit repeatedly attacked. Two of Labour’s three great appointments—Matt Clifford and Poppy Gustafsson—have already left.