What this covers
Philip Ball, a prolific science writer, joins Tyler Cowen to contest the reductionist explanations that dominate how we think about biology, medicine, and society. The conversation ranges across genetics, technology, social change, and the institutional structures of science itself, with Ball pushing back repeatedly against single-cause narratives—the selfish gene as a complete account of life, AI apocalypse myths, technological solutions to social problems—in favor of understanding how phenomena emerge at higher levels of causation.
Ball argues that the selfish gene metaphor works precisely for evolutionary genetics but fails to explain what life is, because genes do not act as independent agents; they are replicated by cells working cooperatively. CRISPR gene editing succeeds for rare monogenic diseases like sickle cell anemia but cannot address major health conditions and traits—heart disease, obesity, height, intelligence—because these involve hundreds or thousands of genes interacting at higher organizational levels where intervention should actually be aimed. He extends this critique to social problems: declining fertility is not an infertility crisis solvable by artificial wombs but a matter of social and economic choice. The conversation also touches on how science itself operates within political and social contexts that scientists cannot escape, drawing on historical examples like Heisenberg under the Nazis, and how institutional pressures to publish drive scientific misconduct and confirmation bias. Ball discusses the AI apocalypse myth as a convenient distraction from mundane, present-day harms, the history of science-religion relations, and the architectural meaning embedded in Gothic cathedrals, among other turns.
Ball argues that most of the phenomena we care about in biology, medicine, and society arise at higher levels of organization than single causes (genes, technologies, individual scientists), so reductionist framings—the selfish gene, technological fixes for social problems, AI apocalypse myths—mislead us about how things actually work and where to intervene.
- The gene-centric view is valid for evolutionary genetics but fails to explain how life works because most traits and diseases are polygenic and arise at higher organizational levels
- Social problems like falling fertility require social solutions, not technological ones like artificial wombs
- Science operates within social and political contexts and cannot pretend to a 'purer sphere' above them
Understanding biological complexity requires recognizing multi-level coordination and communication that current models and frameworks inadequately capture.
- All cell types in the human body—not just neurons—can communicate with one another through chemical signals (as in synapses), mechanical signals (tugging on a cell membrane can induce it to grow or develop differently), and electrical signals; we know these languages but not the kinds of conversations cells have to coordinate something like regrowing a fully formed limb.
“if you tug on the membrane of a cell, that can induce the cell to do something or to grow or develop in a certain way that it wouldn’t have done otherwise.”
- CRISPR works predictably for rare monogenic diseases like sickle cell anemia because a single gene causes the condition, but most major health conditions (heart disease, obesity, diabetes) and physical traits (height, intelligence) correlate with differences across hundreds or thousands of genes that each perform many tasks; so gene editing is unlikely to address them, because the real causation arises at a higher organizational level where genes interact, and that is where intervention should be aimed.
“It’s been used, for example, to cure sickle cell anemia, which comes about because of a gene, a mutation to one particular gene... we can have a predictable response... because the condition is being caused by a single gene.”
- While neurons exchanging electrical pulses resemble transistors exchanging pulses on silicon chips, it is unclear whether there is any straightforward—or even any—translation between silicon computation and the cognitive processing of neurons, and more importantly unclear whether treating the brain as a computer is a useful way to help neuroscientists formulate theories of how it works.
“It’s really not clear at the moment whether there is any straightforward or even any translation at all between the kind of computation that is done by silicon circuits and the kind of cognitive processing that is being done by neurons.”
- Creatures that regenerate limbs (salamanders, axolotls) seem to maintain a kind of stem cell that can regrow into all the tissue types a limb requires, unlike human adult stem cells (e.g., in bone marrow) which produce only a limited number of cell types; how this is coordinated in an already-grown body rather than an embryo is a profound and intensely researched open question with implications for regenerative medicine.
“in creatures that can regenerate their limbs, they seem to have a kind of stem cell that can regrow into all the tissue types that that limb requires.”
Religion provided the intellectual backdrop for most of science's history rather than opposing it fundamentally.
- For most of the history of science, religion (Islam, Christianity, Judaism) was the backdrop within which science developed rather than its antagonist; the idea of a fundamental tension between science and religion is largely a modern construct that only began to be emphasized in the late 19th century for polemical reasons.
“I think this idea that there was somehow a big tension between science and religion, in many ways, it’s a modern construct. It’s something that really only started to be emphasized in the late 19th century for various polemical reasons.”
- The 17th-century Scientific Revolution is better understood as the age when curiosity was liberated: where medieval theologians often saw curiosity as suspicious prying into matters beyond humankind's preserve, the Renaissance and humanism made universal curiosity acceptable, and for scientists like Robert Boyle, exploring God's creation was framed as a Christian duty.
“It was really the first golden age of curiosity, that time. To my mind, that’s a better way of looking at it than calling it the Scientific Revolution. It was the age when curiosity was liberated.”
- The 11th-12th century 'Medieval Renaissance' of learning in France was largely imported from Islamic countries: during the Islamic golden age, scholars translated ancient Greek works (Aristotle, Plato) into Arabic, and European scholars then traveled to Moorish Spain to learn Arabic and translate those works from Arabic into Latin, creating the influx of learning into Western Europe.
“The Islamic scholars translated a lot of the works of the ancient Greeks — Aristotle and Plato and many others — into Arabic.”
- Gothic cathedrals like Chartres were models of a medieval cosmology built on Platonic principles of cosmic harmony, planned according to ideal simple geometric ratios (1:2, 2:3) of proportion; the real meaning is in plain sight rather than hidden codes, because the architects explicitly drew on Platonic ideas of harmony revived in the Cathedral School of Chartres.
“The Gothic cathedrals themselves are planned according to ideal ratios, simple ratios of height and width and so on, of proportion, ratios of 1:2 and 2:3.”
Population decline and falling fertility reflect social choices about women's autonomy and career options, not infertility or need for technological intervention.
- Demographers have long predicted population will peak around nine billion then decline to perhaps six billion, which would be a fantastic thing because population growth is one of the big challenges driving demands for food, energy, and economic growth; there is no reason to believe such a decline would be inherently bad or that population will shrink to nothing.
“it’s going to peak at maybe something like nine billion or so. Then it’s predicted to decrease to something more like perhaps six billion, which I think would be a fantastic thing”
- Falling birth rates in some Asian and African countries appear connected to the increasing empowerment and education of women, who increasingly question the traditional route of marrying and raising a family young versus pursuing a career and independence—a positive development whose underlying cause (why women face an either/or choice) we need to understand.
“there seems to be some connection to the increasing empowerment of women, which has to be a good and important thing. Particularly, in some Asian countries, women are increasingly thinking, “Why should I follow the traditional route of getting married young and raising a family rather than having a career?””
- Declining fertility is a social and economic problem that will not be solved by technology like artificial wombs; the decline is not caused by infertility but by the choices people make, so the answer lies in understanding and addressing those social changes rather than manufacturing children by artificial means.
“There is a social and economic problem that is not going to be solved by technology. The idea that if... there is the case of declining fertility, it’s not at all obvious that what we therefore need is more children made by artificial means.”
- Artificial womb technology is advancing slowly from two directions—technologies keeping alive extremely premature babies who would previously have died, and the ability to grow embryos for longer outside the womb (forbidden past 14 days in humans, but animal embryos can be grown almost to half-gestation).
“There are technologies now that are being used to keep alive babies who are born very, very prematurely that would previously have died. Then, from the other direction, it’s becoming possible to grow embryos for longer and longer periods outside the womb.”
Scientists have historically acquiesced to political ideology rather than mounting effective opposition to it.
- History gives little encouraging evidence that scientists have ever mounted effective opposition to imposed ideology or political power; German physicists under the Nazi regime largely acquiesced, mounting no real effective opposition despite few being actively sympathetic.
“On the whole, the scientists acquiesced to what the regime wanted them to do. Very few of them were actively sympathetic to the Nazi party, but they mounted no real effective opposition whatsoever.”
- Scientists need to abandon the attitude—exemplified by Heisenberg under the Nazis—that science operates in a purer sphere removed from political nastiness; they must recognize they work within a social and political context they sometimes need to take control of, which requires institutional structures rather than leaving individuals to their own moral devices.
“I think scientists need to get real, really, about the fact that they are working within a social and political context that they have to be able to work with”
Institutional pressures to publish constantly drive misconduct, bias, and wasted researcher time despite being entirely fixable.
- Unrealistic institutional pressures on scientists—especially young ones—to publish as much as possible leave them no time to think, drive high-profile cases of scientific misconduct including fabricated data, and create strong pressures for cognitive biases to creep in as researchers seek the results they hope to find; this is eminently fixable because there is no reason science must work this way.
“There are these tremendous pressures, absolutely unrealistic pressures on scientists — particularly young scientists starting their careers — to publish as much as they can.”
- The expansion and increasing institutionalization of science may make it a victim of its own success, giving rise to a more conservative, cautious approach—a regression to the mean toward mediocre and safe work—as high-risk frontier and blue-skies research becomes harder to fund and publish amid demand for short-term commercial return, which could partly explain a slowdown in transformational discoveries.
“it gives rise to a more conservative approach to things, a more cautious approach to things, so that it becomes harder for scientists who are working at what you might think of as a more high-risk frontier of research.”
Metaphors and myths help societies think through anxieties and shape how people understand complex phenomena.
- Modern narratives acquire mythic status (taking at least 50 years) because they satisfy a need to explore social anxieties in story form; myths are tools for thinking with—they have no moral or clear conclusion but let us examine anxieties, as with the emerging zombie myth which is about humanity overrun by a mindless horde rather than individual zombies.
“these narratives acquire mythic status because they satisfy some need in us to explore social anxieties in stories in a way that allows us to look at them and explore them and think about them. Myths are tools for thinking with.”
- Invisibility, traced from the myth of Gyges in Plato's Republic through H.G. Wells's The Invisible Man, functions as a metaphor for lack of responsibility: when people can evade responsibility for their actions, that capability corrupts almost inevitably—a dynamic visible today in online anonymity, sometimes called the Gyges effect, which encourages people to do and say things they never would in person.
“Invisibility there is a metaphor for lack of responsibility. When you can evade responsibility for your actions... that is a corrupting capability to have.”
The selfish-gene metaphor usefully explains allele spread but fails to capture what life actually is or how it works.
- The 'selfish' metaphor refers only to an allele spreading at the expense of other variants of the same gene; it says nothing about what genes do biochemically, because a set of genes in a genome must work together cooperatively to make an organism—the gene producing one enzyme is not competing with the gene producing another enzyme.
“That’s very different from the notion of a set of genes in a genome working together to make an organism. They have to work together to make an organism, so they’re not competing with each other.”
- The selfish-gene metaphor is a productive model specifically for evolutionary genetics—explaining how alleles spread when a mutation confers an adaptive advantage—but it fails as a basis for understanding what life is, because there is no life in that picture; genes are portrayed as agents that replicate, when in fact genes do not replicate but are replicated by cells, an important distinction for understanding how life works.
“The problem I have is when that comes to be seen as the basis for understanding everything that life is. Because the problem with that is that there is no life in that picture.”
Commercial interests systematically distort scientific reporting; vaccines were properly tested through financial acceleration while pharmaceutical underreporting remains a real unsolved problem.
- The Covid vaccines were properly tested while being accelerated to a needed degree, and that acceleration was primarily a matter of money: so much funding was thrown at the problem that steps normally done in succession could be done all at once, with drug companies able to take financial risks because they had the backing.
“There was so much money thrown at the problem that it was possible to do things all at once that normally would have been done in succession, that the drug companies could afford to take those risks because they had the financial backing.”
- Pharmaceutical companies tend to underreport work that conflicts with the message they want to put out when developing a drug, including underreporting of negative or null effects, so the conflict between scientific research and commercial interests is a real and unsolved problem.
“pharmaceutical companies themselves tend to underreport any work that seems to conflict with the message they want to put out if they’re developing a drug. There’s underreporting of negative effects or of null effects.”
- Climate scientists went through an incredibly careful, cautious process of not wanting to be alarmist or claim more than the data allowed; this caution may have delayed raising the alarm, but it is precisely why we can today say with such confidence that climate change is real and primarily anthropogenic.
“It felt to me as though that is why today we can say with such confidence that climate change is real, and that we know something about the causes, and that it seems clear that they’re primarily anthropogenic because there was that very cautious, very careful checking-out of the grounds”
Space's terrifying beauty should remind us how precious and fragile Earth is as humanity's only habitable place.
- Space appears beautiful from Earth precisely because Earth is a place of safety; in reality space is a deathly place constantly trying to kill us with radiation, cold, and vacuum, so the beauty we feel is a terrifying sublime that should remind us how precious our planet is, as illustrated by William Shatner reporting profound emptiness and a sense of looking at death on his Blue Origin flight.
“Space, in reality — when we’re out there, when people are out there, it’s constantly trying to kill us in all sorts of ways. With radiation, obviously; it’s incredibly cold; it’s a vacuum. So, it’s a deathly place.”
Ambitious theories-of-everything like the free energy principle remain too abstract to guide actual experimental work and their utility remains undecided.
- The free energy principle is an interesting, formalizable theory (a generalized theory of agency) that many neuroscientists are exploring, but like many theories-of-everything it tends to be so mathematically abstract that it is hard to relate to actual lab work or to use to pose new questions and design experiments; whether it will prove useful remains undecided.
“they become so abstract, so mathematically abstract that it becomes very hard to see how you relate the ideas... to the lab work that a biologist, that a neuroscientist is doing.”
AI apocalypse narratives distract from real present dangers like deepfakes and misinformation while conveniently deferring corporate responsibility.
- The AI apocalypse myth—super-intelligent AI wiping out humanity—is conveniently fed by tech companies because it sits in the indefinite future and lets them claim they are taking care of it, while distracting from the real, mundane here-and-now dangers of AI like deepfakes, misinformation, and misuse driven by attributing AI too much power.
“That, I think, is a very convenient myth for AI companies because it’s something that’s out there in the indefinite future, and they can say, “Oh, sure, we’re worrying about that, and we’re taking care of that.””
Microplastics accumulate everywhere without breaking down and some mimic hormones, making them intuitively undesirable despite unclear harm extent.
- Microplastics are a clear global pollution problem found everywhere from the Antarctic to marine organisms, and some plasticizer additives chemically resemble hormones and may mimic them to induce hormonal changes; while it is too early to know the extent of harm, any substance that does not break down and simply accumulates in the body is intuitively undesirable.
“any substance that doesn’t break down, that simply accumulates in the body in this way — that doesn’t sound like something that you want to have happening in your bodies and in your tissues.”
Cellular symmetries and asymmetries may simply be evolutionary accidents with no clear adaptive benefit.
- The body is largely bilaterally symmetric but has small internal asymmetries; while we now understand developmentally how symmetry gets broken from an initially symmetrical bundle of cells, whether these asymmetries have any adaptive or physiological benefit is unclear and they may in some cases simply be an evolutionary accident.
“whether there’s an adaptive benefit or a physiological benefit to these small asymmetries that appear internally in the body, I’m really not sure. It may just, in some cases, be an evolutionary accident.”
Science writing requires genuine insight and a distinctive voice developed through wide reading, not mere translation of technical material.
- Successful science book writing is not a way to become rich and requires loving the work; it is not enough to do a translation job of a difficult field—you must bring something of your own, some insight, and find your own voice, which is developed by reading very widely.
“It’s not enough to just do a translation job of a difficult field. You really need to have something to bring to it, some insight to bring to it.”
Ball's prolific output comes from refusing specialization and spending years exploring new topics until insight emerges.
- Ball attributes his prolific output to deliberately refusing to specialize: rather than becoming known for one area, he uses the privileged opportunity to spend years digging into entirely new topics he is curious about, waiting for the 'penny to drop' moment when he sees something new worth telling, as happened with his Gothic cathedrals book.
“for me, it’s just such a wonderful opportunity to express and give free rein to my curiosity, to find out about these new topics.”