Otis Kim: Hey, quick question before we start — if you spent eight hundred million dollars engineering your own body, would you eat a steak at your girlfriend's parents' house?
Iris Holm: That's a very specific hypothetical.
Otis Kim: Bryan Johnson did. That's what we're talking about today — he went to Australia to meet Kate's parents, broke a six-year meat-free streak that was a core pillar of Project Blueprint, and then immediately posted his gut bacteria data to X. Roseburia and Agathobacter down sixty-one percent. Bile-tolerant species up two-point-five times. Thousands of replies. And then — and this is the line — he writes that he wonders if he's taken 'this whole longevity thing too far.'
Iris Holm: The man who launched a protocol called 'Don't Die' is now wondering if he went too far.
Otis Kim: Which — I mean, that framing is doing a lot of work. Blueprint costs two million a year, involves over a hundred daily measurements, launched October 2021. The whole premise is that aging is an engineering problem. And the crack in that premise didn't come from a lab result. It came from a dinner table in Australia.
Iris Holm: Hold on — is the crack the diagnosis or the dinner?
Otis Kim: That's actually the right question. And I think the honest answer is neither — the crack is that he can't tell which one broke it either.
Iris Holm: Okay, that's where I want to dig in.
Otis Kim: Okay but the microbiome numbers are real. Roseburia down sixty-one percent, that's a real marker. Fiber-fermenting bacteria, they produce short-chain fatty acids, protective stuff. Bile-tolerant species up two-point-five times, also real, associated with higher-fat dietary patterns. The biology isn't fake. The attribution is.
Iris Holm: Right. And that's exactly the problem.
Otis Kim: Think of your gut bacteria as a crowd at a venue. The crowd changes depending on the neighborhood, the time of night, and what's on the menu. Johnson changed all three simultaneously — flew to Australia, blew up his circadian rhythm, ate animal protein for the first time in six years. The crowd shifted. But which one moved it?
Iris Holm: And he has no control condition. That's the actual gap.
Otis Kim: None. Academic literature on intercontinental travel — Dutch cohorts traveling to Asia, that kind of work — shows gut bacterial communities shift rapidly just from jet lag and a new food environment, no dietary deviation required. Johnson went to Australia. Jet lag, new water, different ambient diet. Roseburia and Agathobacter were already under pressure before the first bite of meat.
Iris Holm: So this isn't a diet experiment. It's a travel disruption with a meat narrative bolted on.
Otis Kim: And the reporting actually gives this away — the bile-tolerant bacteria rise gets cited as both two-point-five times and a hundred and fifty-four percent in different places. Same number, different framing. But no source specifies the baseline measurement period or the testing methodology. Which means the baseline was never clean to begin with. You can't dress up a noisy observation in two different units and call it precision.
Iris Holm: N=1, no control, ambiguous baseline. The question is — what is this actually evidence of?
Otis Kim: It's evidence that Bryan Johnson was in Australia. That's genuinely all you can say with confidence. Everything else — the meat caused this, plant-based is vindicated, omnivores rejoice — that's the X replies talking, not the data.
Iris Holm: And that's where the bacteria story actually dissolves into something worse — because the gastritis doesn't care about the attribution problem. Johnson was diagnosed with autoimmune gastritis in 2026. Incurable. And he'd had chronically low ferritin for eleven years before that diagnosis landed.
Otis Kim: Eleven years. While Blueprint was running over a hundred daily measurements and fifty-plus supplements and sixty-seven interventions citing a thousand clinical trials.
Iris Holm: Right. So the question is — did it catch it, or did it miss it?
Otis Kim: And that's actually — okay, I want to be fair here, because there's a real version of the pro-Blueprint argument. Picture someone who isn't Johnson — say, a woman in her forties running a hundred biomarkers a day for five years, and she gets diagnosed with an incurable stomach condition. She might genuinely be the person who caught it before it became catastrophic. Most people with autoimmune gastritis have no idea they have it until it's late-stage. That's a real thing. Blueprint surfacing it early, that could be the win.
Iris Holm: Sure. And that's the trap.
Otis Kim: Exactly — I mean, actually no, the trap is that both stories are available. If Blueprint succeeds, it claims the detection. If it fails, the failure gets reframed as early detection anyway. Johnson literally cannot distinguish between those two outcomes with his own data. The protocol is unfalsifiable. Whatever happens, there's a narrative that fits.
Iris Holm: He posted publicly that he wonders if he's taken 'this whole longevity thing too far.' That's the first time the doubt isn't quantifiable into a win.
Otis Kim: And that's what makes the gastritis the actual rupture — not the bacteria numbers, not the steak at Kate's parents' table. It's the one data point that the measurement infrastructure cannot retrofit into a success story. Not yet, anyway.
Iris Holm: The part that makes this even stranger — and we'll get to this — is what the granularity of the measurement infrastructure is actually doing versus what it claims to be doing. Because I think that's where the social context angle becomes the tell.
Otis Kim: And the social context is actually the cleanest data point in this whole story. Not the Roseburia numbers, not the gastritis — the fact that a man who compared eating animals to future AI-human oppression ate animal protein because he was at Kate's parents' table. That's not a protocol decision. That's a person being a person.
Iris Holm: And then he posted it to X.
Otis Kim: And then he posted it to X. Which is — I mean, that's where the measurement infrastructure thing actually lands. Blueprint can read fiber-bacteria ratios to the decimal. It cannot tell him whether what just happened was a biological experiment or a content opportunity. And I'm not sure he can either.
Iris Holm: The tools are too granular and the interpretive layer is completely loose. Two million dollars a year, a hundred daily measurements, and the output is — what? An X thread that starts a plant-based versus omnivore war. That's the demonstration.
Otis Kim: And the brutal part is that the pillars that actually have evidence behind them — diet quality, sleep, exercise — those are largely free. Blueprint doesn't demonstrate what optimization costs. It demonstrates what one person with eight hundred million dollars from selling Braintree to PayPal does with his anxiety about dying.
Iris Holm: Wait — so the 'Don't Die' philosophy is actually just... grief? Expensive, quantified grief?
Otis Kim: I think that's actually closer to true than uncomfortable. He's described a decade of depression before Blueprint launched in 2021. The protocol isn't vanity — I'll grant him that. But the measurement infrastructure can't answer its own central question. Whatever happens, there's a narrative available. Caught early. Prevented worse. Proves the method. And that's — yeah, that's not falsifiable. That's belief with a spreadsheet.
Iris Holm: Belief with a spreadsheet. That's the calibrated take. Blueprint isn't science and it isn't fraud — it's n=1 self-experimentation where the experimenter is also the brand, and no outcome can falsify it. The dinner in Australia was actually the most honest data point he's produced.
Otis Kim: Which circles back to the steak, actually. You open with — would you eat at your girlfriend's parents' table if you'd spent eight hundred million dollars engineering your body. And the honest answer, the answer the whole episode lands on, is: yeah. Probably. Because Bryan Johnson did. And the most monitored body on earth could not tell him afterward whether that meal was a protocol deviation, a microbiome experiment, or just — Tuesday in Australia.
Iris Holm: And that's the design problem. Not a data problem. 'Don't Die' is the philosophy. Autoimmune gastritis is the result. Blueprint could not prevent it, could not definitively flag it, and cannot tell you whether it helped. If no outcome falsifies it, it's not science. It's a belief system with a very expensive spreadsheet.
Otis Kim: Belief with a spreadsheet. Yeah. I think that's where I actually land — not that Johnson was wrong to try, but that the frame was always going to win. Whatever happened, the frame survived. Anyway. Good conversation to have had.
