“Thankfully, today we have a more accurate map of the brain. We now know that people with mental health problems have chemical imbalances in their brains which can be solved with a lifelong prescription to certain pharmaceuticals…” you almost had me there :D Great essay reminds me of Goodhart’s law in a weird way.
I think there’s a deeper layer here. When you build a complex system, like a software system, the blueprint isn’t all of it. It’s just part of it. Sometimes you have to go out and “draw a map” ie, interact with the system experimentally and try to see what it’s doing. I don’t think this contradicts your point at all, so much as, even for “totally designed” systems like computer networks, there are often emergent phenomena that weren’t in the blueprint but show up as your plans unfold and collide with the reality outside of them. The interaction between bike tires and gravel is an example here. Even a blueprint won’t tell you precisely which surfaces the bike tires can grip on - at best it can help you guess.
One tiny note: I couldn't tell if you were being sarcastic or not when you said, "We now know that people with mental health problems have chemical imbalances in their brains..."
I assume you know that there are no such "chemical imbalances"? That's another myth.
Interesting way of looking at it. Hubris is right. Nature appears to have intricacies we’ve barely come to appreciate. Maybe part of the problem is that we credit everything that’s ‘natural’ to blind chance and everything that humans invent to some process. A process has got to be better and smarter than random chance, which by definition has no brains whatsoever behind it.
This is thought-provoking, but I think for this to be usefully applied we'd have to better distinguish between lobotomy-style intervention (bad), and vaccine or antibiotic intervention (good). If it's just "interventions sometimes have downsides" then that's not useful unless we have some inclination that these downsides erode 100% (or 200%) of the gains, or 10% of the gains.
With ozempic, if you're pre-diabetic., 65 and obese, then the risks of taking it seem small relative to the gains. You'd be crazy not to take it. But if you're 25 and 10lbs overweight you'd be crazy (I think) to take it. I hope no one uses this argument to tell the 65 year old not to take the drug.
More generally, I worry this is a steelmanned version of the naturalism fallacy. You've picked two definitely bad interventions (trans fats and lobotomies), added synthetic fertiliser, which is *much* less clear (standard position today would be this has been a huge boon for society), and then extrapolated to a new drug (ozempic). Why not mRNA vaccines? Why not cognitive behavioural therapy? (Also based on science about how a natural process - our thought patterns - can be changed).
I think you could easily extend this to mRNA vaccines yes, the idea that they are or were a completely risk-free intervention that everyone should take advantage of was ludicrous. If you were high-risk it's probably a worthwhile tradeoff, but the notion that kids should be getting them and getting constantly boosted is just silly given the risk profile.
And I think you're kind of committing the error I'm talking about with your mention of Ozempic.
"The risks of taking it seem small relative to the gain," this is the precise problem, we don't know what the risks are yet, and the 'gain' is something that can be achieved without Ozempic. That's very different from, say, smallpox immunity, which you cannot achieve through lifestyle changes.
I suspect we've got different views of how easily a 65 y/o might be able to get out of the danger zone through a strategy of lifestyle changes (which for lots of people seems to be very hard). But I agree with you re: kids and covid vaccines.
Map vs Blueprint is a great mental model which I will keep to hand!
That said the iteration of bad map, solve issues, unexpected off-map effects on system create new issues make better map to come up with solution that solves those issues too... sounds like progress! The bad part is how long it takes for the new issues to be recognized and acted on, in part due to financial incentives.
The lesson I would take is for the solvers and regulators to more on the lookout for the off- map effects and take appropriate action (ban, regulate, label, iterate on mapping and creating).
granted, the larger point is the focus. but, one can avoid shaky categorical statements like that (which happens to not be nearly as solid as one might think reading it) while still achieving the larger goal.
No there isn't, go do your own research and form your own opinion on the subject if you disagree.
It's waste of time to meticulously cite sources in a piece like this. People will simply latch onto them and quibble over p-values and sample sizes while missing the much larger point.
“Thankfully, today we have a more accurate map of the brain. We now know that people with mental health problems have chemical imbalances in their brains which can be solved with a lifelong prescription to certain pharmaceuticals…” you almost had me there :D Great essay reminds me of Goodhart’s law in a weird way.
Hahaha, thanks Felix
What a great distinction!! Well-thought-out and well-written. I feel like you gave a pretty legit explanation about the cycles of the market.
Thank you Rima!
Phenomenal essay. I have a feeling I’ll be using the map/blueprint distinction many times in life.
Thanks Mark! Glad you enjoyed it
I think there’s a deeper layer here. When you build a complex system, like a software system, the blueprint isn’t all of it. It’s just part of it. Sometimes you have to go out and “draw a map” ie, interact with the system experimentally and try to see what it’s doing. I don’t think this contradicts your point at all, so much as, even for “totally designed” systems like computer networks, there are often emergent phenomena that weren’t in the blueprint but show up as your plans unfold and collide with the reality outside of them. The interaction between bike tires and gravel is an example here. Even a blueprint won’t tell you precisely which surfaces the bike tires can grip on - at best it can help you guess.
And AI is demonstrating these emergent property aspects more and more
Fantastic read, thanks Nat!
One tiny note: I couldn't tell if you were being sarcastic or not when you said, "We now know that people with mental health problems have chemical imbalances in their brains..."
I assume you know that there are no such "chemical imbalances"? That's another myth.
Yup, it was sarcasm :)
Wait, there aren't?
Interesting way of looking at it. Hubris is right. Nature appears to have intricacies we’ve barely come to appreciate. Maybe part of the problem is that we credit everything that’s ‘natural’ to blind chance and everything that humans invent to some process. A process has got to be better and smarter than random chance, which by definition has no brains whatsoever behind it.
Very useful mental model. A suggestion for complementary reading: "The Map Is Not the Territory" on Farnam Street Blog.
Was going to share this! “The simplified version of reality, is not reality.”
This is thought-provoking, but I think for this to be usefully applied we'd have to better distinguish between lobotomy-style intervention (bad), and vaccine or antibiotic intervention (good). If it's just "interventions sometimes have downsides" then that's not useful unless we have some inclination that these downsides erode 100% (or 200%) of the gains, or 10% of the gains.
With ozempic, if you're pre-diabetic., 65 and obese, then the risks of taking it seem small relative to the gains. You'd be crazy not to take it. But if you're 25 and 10lbs overweight you'd be crazy (I think) to take it. I hope no one uses this argument to tell the 65 year old not to take the drug.
More generally, I worry this is a steelmanned version of the naturalism fallacy. You've picked two definitely bad interventions (trans fats and lobotomies), added synthetic fertiliser, which is *much* less clear (standard position today would be this has been a huge boon for society), and then extrapolated to a new drug (ozempic). Why not mRNA vaccines? Why not cognitive behavioural therapy? (Also based on science about how a natural process - our thought patterns - can be changed).
I think you could easily extend this to mRNA vaccines yes, the idea that they are or were a completely risk-free intervention that everyone should take advantage of was ludicrous. If you were high-risk it's probably a worthwhile tradeoff, but the notion that kids should be getting them and getting constantly boosted is just silly given the risk profile.
And I think you're kind of committing the error I'm talking about with your mention of Ozempic.
"The risks of taking it seem small relative to the gain," this is the precise problem, we don't know what the risks are yet, and the 'gain' is something that can be achieved without Ozempic. That's very different from, say, smallpox immunity, which you cannot achieve through lifestyle changes.
I suspect we've got different views of how easily a 65 y/o might be able to get out of the danger zone through a strategy of lifestyle changes (which for lots of people seems to be very hard). But I agree with you re: kids and covid vaccines.
Map vs Blueprint is a great mental model which I will keep to hand!
That said the iteration of bad map, solve issues, unexpected off-map effects on system create new issues make better map to come up with solution that solves those issues too... sounds like progress! The bad part is how long it takes for the new issues to be recognized and acted on, in part due to financial incentives.
The lesson I would take is for the solvers and regulators to more on the lookout for the off- map effects and take appropriate action (ban, regulate, label, iterate on mapping and creating).
Mind blowing
Fantastic read 🔥
Great piece, Nat!
granted, the larger point is the focus. but, one can avoid shaky categorical statements like that (which happens to not be nearly as solid as one might think reading it) while still achieving the larger goal.
“People are eating more because they’re nutrient-starved”
citation or nontrivial qualification needed
No there isn't, go do your own research and form your own opinion on the subject if you disagree.
It's waste of time to meticulously cite sources in a piece like this. People will simply latch onto them and quibble over p-values and sample sizes while missing the much larger point.
Solve this type of nitpicking (the comment you're replying to here), Nat, by framing your statement as a question.
But I guess an anonymous commenter named "Hello" isn't really here to think much more than he or she needs to.
Brilliant!
Hitting home. Thanks Nat!
Great article. We must realize there’s almost always something lurking beneath the surface.