Why Nutrition Science Got It WRONG & the Case for the Carb-Insulin Model | Gary Taubes & Mike Haney
What this video covers
A Levels podcast interview from January 2026 in which host Mike Haney presses Taubes on two threads: the carbohydrate-insulin model of obesity, and a more philosophical question — how a journalist decides he knows something in a field where, as Taubes argues, the science itself is unreliable. Taubes explains why he went beyond “we just don’t know,” why he thinks nutrition never inherited the rigorous scientific culture of physics, and why public-health messaging demands a certainty science can’t honestly supply.
Key points
- Taubes opens with his Bill Gates analogy: saying Gates got rich because he took in more money than he spent is true but explains nothing — and he argues energy-balance reasoning about obesity has the same defect.
- Taubes explains why his conclusions go past agnosticism: a journalist, he says, must keep researching until he can decide who is right, while writing in a way that leaves open the possibility he is wrong.
- Taubes argues physics and biology inherited a culture of rigorous, critical science from the German and Austrian researchers who fled Europe before World War II — “Hitler’s gift” — while nutrition and public health, rebuilt in the US without that mentorship lineage, never did.
- Taubes says Good Calories, Bad Calories ends by separating belief from proof: its epilogue lists what he would tell his own family — conclusions he says he would “bet my children’s lives on” — while insisting readers judge the evidence critically for themselves.
- Taubes argues doctors are trained to project certainty because patients and public-health campaigns demand it, but that dogmatic certainty is “contrary to good science on a very profound level.”
- Using seat belts as the example, Taubes argues public health must oversell tiny individual effects — one in 600 drivers benefiting from a seat belt in the 1980s figures he cites — to move whole populations, which is harmless for seat belts but dangerous, he says, when the underlying advice (such as the saturated-fat campaign) may itself be wrong.
- Taubes then lays out the carbohydrate-insulin model as his alternative to energy-balance thinking: the question is not how much energy people take in, but what hormonal signals tell their fat tissue to store or release it.