I haven’t read this book Power and Progress, but I couldn't agree more with Noah here. Honestly, I’m so grateful we have younger, sharper minds like Noah Smith doing the heavy lifting on these genuinely critical modern issues, saying what many of our generation peers think but are too busy or lacks the platform to say.
I loved the other Why Nations Fail, which makes this trend of Nobel-level giants losing their edge when facing AI all the more disappointing. It's a sort of "vulgar caution"—when confronted with AI, they offer nothing fresh, just boilerplate warnings and predictable pleas for regulation. The same thing happened with the late Henry Kissinger. He was fascinated by AI in his final years and wrote extensively about it, but reading his take felt remarkably similar: these giants lived and shaped history, yet when confronted with a paradigm-shaping reality like AI, their insights don't actually sound any wiser than those of a regular young person living here in California.
More people need to be reading Noah's sober, data-driven critiques.
I read it and the margins of my book are similarly filled with skeptical comments. The book is every bit as bad as Noah suggests.
Speaking of a “council of mandarins” my take on the book is that the authors’ just want the power for a council made up of themselves or people who think like them in charge of technological innovation. The true meaning of “power and progress” (or the increase in power and the death of progress)
Brilliant. But is there any evidence that any of our political masters, or their advisors, read essays like this?
The creators of great ideas in technology can form companies, or sell their ideas to companies, and see them adopted, by the simple operation of the pursuit of self-interest on the part of the adopters. But great ideas regarding society and politics? (I'm not saying this book review's are correct as opposed to the ideas of the authors it criticizes. Just that this is the sort of debate to which our political leaders should be exposed, directly or indirectly. How can we achieve this?)
Brilliant point, Doug. You’ve highlighted a painful asymmetric reality: technology scales through the sheer force of market self-interest, but great policy ideas have to fight through a thick wall of bureaucratic inertia.
To your first question—is there any concrete evidence? Probably no, although Substack’s own marketing fluff said, think-tank researchers, congressional staffers, and policy advisors are subscribers to this tier of intellectual writing. I certainly hope that’s true and that they are absorbing it invisibly. But as you implied, hoping isn't enough—it would be a lot better if we actually saw these people engaging in these critical debates openly.
Sir your second question is a much bigger and wiser one. Let’s be blunt—politicians don't read essays out of intellectual curiosity; they only change when they feel the heat. Right now, the ultimate reality check is the fierce tech competition with China. If Washington falls for the "vulgar caution" of ivory-tower academics and ties its own AI sector up in endless regulatory red tape, the US simply loses the race. Real change will happen when the fear of falling behind China forces the pragmatic factions in government to actively hunt for level-headed, pro-innovation arguments—like the ones Noah provides—to dismantle the ideological nonsense.
I'm probably being hopelessly utopian here, but I wonder if there is any way that our intelligentsia, Left and Right, could unite to try to get our political masters at least aware of the sort of issues discussed here? For example, pick one essay a month, alternating among Left, Right and neither, and make a massive effort to publicize it, aiming especially at the political class.
We would need a small governing council to pick the monthly essay, and to co ordinate publicity efforts. (For example, collecting the email addresses of all political people; co-ordinating saturating online political forums with the monthly essay to read; finding other publicity outlets. )
Previous philosophers have only interpreted the world. The point is to change it.
With all due respect, Sir, I deeply admire your idealism—but LOL, allow me to say, you truly are a hopelessly utopian thinker!
First, I actually love your concept of tossing aside the rigid Left-Center-Right labels and having a council determine the most rigorous essays for the public. C-SPAN was originally to serve exactly that kind of objective, non-partisan role for the political class. But nobody watches those antique cable channels anymore; everyone—including politicians and their staffers—has moved to TikTok and algorithm-driven echo chambers.
Second, while fostering cross-ideological debate is noble, in today’s hyper-polarized climate, it is politically suicidal. Look at Gavin Newsom: when he tried to cross the aisle for high-profile debates with Charlie Kirk and Sean Hannity, he failed to win over moderates and was instead viciously cannibalized by the far-left of his own party and it severely damaged his capital for 2028. When a politician gets burned just for trying to have an open conversation, who would dare to emulate them?
Finally, your closing quote from Karl Marx—"..the point is to change it", is it?—is a magnificent piece of unintentional irony. If we were to actually implement it, we would find ourselves in a hilarious paradox: utilizing a highly disciplined, top-down, Vanguard-style propaganda machine to aggressively force-feed politicians Noah Smith’s message of "let business be business, let tech be tech, and please just leave the market alone." It’s an absolute masterpiece of black comedy.
I probably didn't make myself clear. I'm NOT advocating 'tossing aside' political labels. Of course every good advocate for his side knows that the other side is pure evil, the embodiment of motiveless malignancy, and must be treated as such. I'm only proposing that periodically, a group of Good and Evil people choose a recently-published essay (alternating among Good, Evil, and Indifferent authors) which they agree is important -- even if only on a know-your-enemy basis -- and try to get our political masters to read it.
As the fellow who wanted to change, and not just interpret, the world once said, "Ignorance never yet helped anybody." Surely we could convince at least a few political leaders that it's worth knowing what the smartest people on their side, and on the other side, are saying about contemporary issues. (Okay, maybe not 'surely'. Substitute 'It's just faintly possible that' for 'surely'.)
Well let's stay on 'surely'! I really appreciate your humor and your great knowledge of Marx (and your impressive quotes!). You’ve made yourself perfectly clear. I sincerely share your hope, utopian or not, that such a dynamic could exist. At the end of the day, regardless of which political camp one occupies, a willingness to engage with the smart minds from the opposing side is always a victory over ignorance, always a good thing. Thank you for the brilliant exchange!
I think Noah's criticism of Acemoglu's definition of power is actually one of the most interesting parts of the review. If power simply means "the ability to achieve objectives," then almost any successful outcome can be explained by power, making the concept difficult to test or even discuss.
But perhaps the more interesting question isn't "Who has power?" It's "Why does power keep reappearing?" People gain it, lose it, and are replaced, yet power itself never seems to disappear. It simply moves to the next person because the structure that produces it is still there.
I tend to think of power a bit like fruit on a tree. Whoever climbs high enough can pick the fruit, but the fruit isn't created by the climber. It is produced by the tree itself. Likewise, positions of power emerge within all kinds of organizations, from relationships and families to companies and states. Individuals may lose power, but the structure that generates power often remains unless the structure itself changes.
If that's true, then asking who has power is only half the story. The other half is understanding the structure itself. Perhaps that's where AI becomes most important, not simply because it makes individuals more powerful, but because it may be reinforcing some power structures while transforming others. Perhaps the real disruption isn't AI alone. It's that AI is changing the environment in which power structures evolve.
Perhaps that's what makes this moment so unsettling. If AI is changing the environment in which power structures evolve, are we also moving toward a new power structure in which the decisions of a very small number of people determine the fate of almost everyone else?
Surely, providing a brief synopsis of an essay is exactly the sort of thing AI can do. And it will make the author's ideas, if not his actual words, more widely known -- an example of AI increasing overall human welfare. (Depending on what you think of the author's ideas, of course.)
I find it amusing how the hatred of Elon Musk makes liberal economists write nonsense. I doubt they would be so skeptical of technological progress if X was still Twitter, held by previous owners, and Musk was still a Dem supporter. Liberals just cannot forgive that Silicon Valley switched allegiance and nobody embodies this more powerfully than Musk. The fact that he cannot be categorized neatly as an antisocial robber baron (works hard, does not have mansions or yachts, pro-clean energy, put the car industry on a sustainable path, Starlink helps Ukraine) and defies conventional narratives just adds to the animosity towards him. Musk derangement syndrome reaches extreme levels: Krugman wants to prosecute Musk, while Acemoglu wants to second guess technological choices entrepreneurs make. For me this is just intellectual hubris.
In basic science research creatures writing things that violate facts and logic would be treated as the worthless creatures they are. By rejecting verifiable facts and proven math and logic applied to those facts they are in fact sub-humans, creatures without a functioning human brain. But rejecting facts and math is common in economics. Krugman now freely interchanges wealth and income even though you couldn’t pass a high school economics class if you didn’t understand that they are different so much that they don’t mathematically have the same dimensions. Then Krugman goes on equating wealth with political power proving he couldn’t pass 5th grade where we learn that the US is one vote per person and there is little evidence of vote buying. As a human with a human brain I find political spending is a valuable signal that often indicates the economic value to some group is high which almost always means that many not in that group will be economically burdened.
On inequality the biggest change over my 6+ decades is that the income of the top 20%, or top 25% has increased relative to the 2nd quintile by 50%. The chart in this post shows similar data. Anyone who can do 7th grade algebra and has the data understands that while the very top earners have increased their relative share of national income it is a fraction of the total captured by the top 20%. Yet people on the left, supported by economists like Dean Baker, think all the change is from the 1% or 0.01%.
Facts and math and logic separate humans from subhumans. It turns out that people who understand facts and logic and math are valuable economically and not surprisingly their relative incomes have increased. A lot of inequality is likely due to the value of people with human brains. A friend with a research team at a world leading technology hardware company said demand for the PhD physicists he hires to work for him has made starting offers right out of grad school (no post doc) around $200k/year. And when your research team leads to major new products the lead is compensated over $1million/year. It classic supply and demand, and there just are not that many people with the skills to do the work. As technology increases in complexity and scale the pyramid gets taller and narrower.
Overall, I really liked this. But I think you need to say more about the alleged gap between pay and productivity. Schneider’s critique doesn’t kill EPI’s point. He shows average pay tracks average productivity when you match sectors and workers correctly. EPI asks whether typical workers shared in those gains. Both can be true, so it does not refute EPI’s point that the difference is rising inequality.
It seems to me that there is history behind this book. I used to read the magazine "Technology Review", back when it was about "technology and policy". The general idea seemed to be that there was an academic discipline of "technology policy" which was about the sort of thing the book recommends, that some government organ can regulate which technologies are deployed and how for the public good. It was somewhat Bolshie in that it never recommended leaving something to the market. In one article, it argued that airline deregulation was a failure because in the first five years or so, airline companies lost a lot of money. Somehow the democratizing of air travel wasn't significant.
Looking at the quoted examples, I suspect that the book's concept of social good is minimizing economic disruption for workers, ensuring that each group remains where it is on the socioeconomic totem pole. Notice that the authors don't object to Japanese real wages falling for decades but they do object to sharp changes in relative wages between groups in the US.
The only version of this concept I've heard of being successful was AT&T's management of the introduction of new technologies into the telephone industry. Reports are that AT&T was careful to regulate which technologies would be deployed and when to ensure minimum disruption (including that no capital investment was obsoleted before it was paid off). That came to an abrupt end when AT&T was broken up and ceased to be a monopoly that could control telephone technology.
> Subsequently, Haber and other scientists used the same ideas to design chemical weapons that killed and maimed hundreds of thousands on World War I battlefields.
This passage is even odder because making ordinary explosives requires fixed nitrogen, which during WW I, Germany used the Haber-Bosch process to supply to circumvent the British blockade of guano imports. So if you want to complain about "weapons enabled by the Haber-Bosch process", the major category was ordinary munitions.
'Set out to prove' books always risk not being true. Books written by two authors always risk self-contradiction and confusion. This was the perfect storm, by the sounds of it.
"Acemoglu and Johnson have a reason for employing a definition [of power] this infinitely broad; it allows them to include persuasion and compulsion in a single category of “power”."
I agree. This is the Foucauldian move though; I love Foucault's work but I hate that the term he came up for what he was trying to illustrate -- how science and scientific experts regulate our lives -- he ended up calling "power." Now anything you don't like -- you can just call it "power." https://computingandsociety.substack.com/p/what-was-michel-foucault-guilty-of
Looks like Acemoglu and Johnson's use of "power" is even worse!
I haven’t read this book Power and Progress, but I couldn't agree more with Noah here. Honestly, I’m so grateful we have younger, sharper minds like Noah Smith doing the heavy lifting on these genuinely critical modern issues, saying what many of our generation peers think but are too busy or lacks the platform to say.
I loved the other Why Nations Fail, which makes this trend of Nobel-level giants losing their edge when facing AI all the more disappointing. It's a sort of "vulgar caution"—when confronted with AI, they offer nothing fresh, just boilerplate warnings and predictable pleas for regulation. The same thing happened with the late Henry Kissinger. He was fascinated by AI in his final years and wrote extensively about it, but reading his take felt remarkably similar: these giants lived and shaped history, yet when confronted with a paradigm-shaping reality like AI, their insights don't actually sound any wiser than those of a regular young person living here in California.
More people need to be reading Noah's sober, data-driven critiques.
I read it and the margins of my book are similarly filled with skeptical comments. The book is every bit as bad as Noah suggests.
Speaking of a “council of mandarins” my take on the book is that the authors’ just want the power for a council made up of themselves or people who think like them in charge of technological innovation. The true meaning of “power and progress” (or the increase in power and the death of progress)
Brilliant. But is there any evidence that any of our political masters, or their advisors, read essays like this?
The creators of great ideas in technology can form companies, or sell their ideas to companies, and see them adopted, by the simple operation of the pursuit of self-interest on the part of the adopters. But great ideas regarding society and politics? (I'm not saying this book review's are correct as opposed to the ideas of the authors it criticizes. Just that this is the sort of debate to which our political leaders should be exposed, directly or indirectly. How can we achieve this?)
Brilliant point, Doug. You’ve highlighted a painful asymmetric reality: technology scales through the sheer force of market self-interest, but great policy ideas have to fight through a thick wall of bureaucratic inertia.
To your first question—is there any concrete evidence? Probably no, although Substack’s own marketing fluff said, think-tank researchers, congressional staffers, and policy advisors are subscribers to this tier of intellectual writing. I certainly hope that’s true and that they are absorbing it invisibly. But as you implied, hoping isn't enough—it would be a lot better if we actually saw these people engaging in these critical debates openly.
Sir your second question is a much bigger and wiser one. Let’s be blunt—politicians don't read essays out of intellectual curiosity; they only change when they feel the heat. Right now, the ultimate reality check is the fierce tech competition with China. If Washington falls for the "vulgar caution" of ivory-tower academics and ties its own AI sector up in endless regulatory red tape, the US simply loses the race. Real change will happen when the fear of falling behind China forces the pragmatic factions in government to actively hunt for level-headed, pro-innovation arguments—like the ones Noah provides—to dismantle the ideological nonsense.
I'm probably being hopelessly utopian here, but I wonder if there is any way that our intelligentsia, Left and Right, could unite to try to get our political masters at least aware of the sort of issues discussed here? For example, pick one essay a month, alternating among Left, Right and neither, and make a massive effort to publicize it, aiming especially at the political class.
We would need a small governing council to pick the monthly essay, and to co ordinate publicity efforts. (For example, collecting the email addresses of all political people; co-ordinating saturating online political forums with the monthly essay to read; finding other publicity outlets. )
Previous philosophers have only interpreted the world. The point is to change it.
With all due respect, Sir, I deeply admire your idealism—but LOL, allow me to say, you truly are a hopelessly utopian thinker!
First, I actually love your concept of tossing aside the rigid Left-Center-Right labels and having a council determine the most rigorous essays for the public. C-SPAN was originally to serve exactly that kind of objective, non-partisan role for the political class. But nobody watches those antique cable channels anymore; everyone—including politicians and their staffers—has moved to TikTok and algorithm-driven echo chambers.
Second, while fostering cross-ideological debate is noble, in today’s hyper-polarized climate, it is politically suicidal. Look at Gavin Newsom: when he tried to cross the aisle for high-profile debates with Charlie Kirk and Sean Hannity, he failed to win over moderates and was instead viciously cannibalized by the far-left of his own party and it severely damaged his capital for 2028. When a politician gets burned just for trying to have an open conversation, who would dare to emulate them?
Finally, your closing quote from Karl Marx—"..the point is to change it", is it?—is a magnificent piece of unintentional irony. If we were to actually implement it, we would find ourselves in a hilarious paradox: utilizing a highly disciplined, top-down, Vanguard-style propaganda machine to aggressively force-feed politicians Noah Smith’s message of "let business be business, let tech be tech, and please just leave the market alone." It’s an absolute masterpiece of black comedy.
I probably didn't make myself clear. I'm NOT advocating 'tossing aside' political labels. Of course every good advocate for his side knows that the other side is pure evil, the embodiment of motiveless malignancy, and must be treated as such. I'm only proposing that periodically, a group of Good and Evil people choose a recently-published essay (alternating among Good, Evil, and Indifferent authors) which they agree is important -- even if only on a know-your-enemy basis -- and try to get our political masters to read it.
As the fellow who wanted to change, and not just interpret, the world once said, "Ignorance never yet helped anybody." Surely we could convince at least a few political leaders that it's worth knowing what the smartest people on their side, and on the other side, are saying about contemporary issues. (Okay, maybe not 'surely'. Substitute 'It's just faintly possible that' for 'surely'.)
Well let's stay on 'surely'! I really appreciate your humor and your great knowledge of Marx (and your impressive quotes!). You’ve made yourself perfectly clear. I sincerely share your hope, utopian or not, that such a dynamic could exist. At the end of the day, regardless of which political camp one occupies, a willingness to engage with the smart minds from the opposing side is always a victory over ignorance, always a good thing. Thank you for the brilliant exchange!
best wishes to your rabbit for a speedy recovery
I think Noah's criticism of Acemoglu's definition of power is actually one of the most interesting parts of the review. If power simply means "the ability to achieve objectives," then almost any successful outcome can be explained by power, making the concept difficult to test or even discuss.
But perhaps the more interesting question isn't "Who has power?" It's "Why does power keep reappearing?" People gain it, lose it, and are replaced, yet power itself never seems to disappear. It simply moves to the next person because the structure that produces it is still there.
I tend to think of power a bit like fruit on a tree. Whoever climbs high enough can pick the fruit, but the fruit isn't created by the climber. It is produced by the tree itself. Likewise, positions of power emerge within all kinds of organizations, from relationships and families to companies and states. Individuals may lose power, but the structure that generates power often remains unless the structure itself changes.
If that's true, then asking who has power is only half the story. The other half is understanding the structure itself. Perhaps that's where AI becomes most important, not simply because it makes individuals more powerful, but because it may be reinforcing some power structures while transforming others. Perhaps the real disruption isn't AI alone. It's that AI is changing the environment in which power structures evolve.
Perhaps that's what makes this moment so unsettling. If AI is changing the environment in which power structures evolve, are we also moving toward a new power structure in which the decisions of a very small number of people determine the fate of almost everyone else?
Henry Farrell has a critique of this review that you might want to respond to: https://www.programmablemutter.com/p/noah-smiths-review-of-power-and-progress
And best wishes to your rabbit. Hope everything gets better soon!
Hi. Respect.
Now, honesty.
If you think you are getting paid by the word .. lovely.
I don't reward anyone for wasting my attention.
So. Please. Provide a synopsis. A "tldr". Axios gets it.
Don't drag me through your painfully indulgent narrative. If I want narrative I can open a Harry Potter book. ( I *NEVER* do that. )
Please respect my time and tell me - UP FRONT - what it is that you wan to tell me.
Surely, providing a brief synopsis of an essay is exactly the sort of thing AI can do. And it will make the author's ideas, if not his actual words, more widely known -- an example of AI increasing overall human welfare. (Depending on what you think of the author's ideas, of course.)
I find it amusing how the hatred of Elon Musk makes liberal economists write nonsense. I doubt they would be so skeptical of technological progress if X was still Twitter, held by previous owners, and Musk was still a Dem supporter. Liberals just cannot forgive that Silicon Valley switched allegiance and nobody embodies this more powerfully than Musk. The fact that he cannot be categorized neatly as an antisocial robber baron (works hard, does not have mansions or yachts, pro-clean energy, put the car industry on a sustainable path, Starlink helps Ukraine) and defies conventional narratives just adds to the animosity towards him. Musk derangement syndrome reaches extreme levels: Krugman wants to prosecute Musk, while Acemoglu wants to second guess technological choices entrepreneurs make. For me this is just intellectual hubris.
In basic science research creatures writing things that violate facts and logic would be treated as the worthless creatures they are. By rejecting verifiable facts and proven math and logic applied to those facts they are in fact sub-humans, creatures without a functioning human brain. But rejecting facts and math is common in economics. Krugman now freely interchanges wealth and income even though you couldn’t pass a high school economics class if you didn’t understand that they are different so much that they don’t mathematically have the same dimensions. Then Krugman goes on equating wealth with political power proving he couldn’t pass 5th grade where we learn that the US is one vote per person and there is little evidence of vote buying. As a human with a human brain I find political spending is a valuable signal that often indicates the economic value to some group is high which almost always means that many not in that group will be economically burdened.
On inequality the biggest change over my 6+ decades is that the income of the top 20%, or top 25% has increased relative to the 2nd quintile by 50%. The chart in this post shows similar data. Anyone who can do 7th grade algebra and has the data understands that while the very top earners have increased their relative share of national income it is a fraction of the total captured by the top 20%. Yet people on the left, supported by economists like Dean Baker, think all the change is from the 1% or 0.01%.
Facts and math and logic separate humans from subhumans. It turns out that people who understand facts and logic and math are valuable economically and not surprisingly their relative incomes have increased. A lot of inequality is likely due to the value of people with human brains. A friend with a research team at a world leading technology hardware company said demand for the PhD physicists he hires to work for him has made starting offers right out of grad school (no post doc) around $200k/year. And when your research team leads to major new products the lead is compensated over $1million/year. It classic supply and demand, and there just are not that many people with the skills to do the work. As technology increases in complexity and scale the pyramid gets taller and narrower.
Book review: "How Africa works" by Joe Studwell
Overall, I really liked this. But I think you need to say more about the alleged gap between pay and productivity. Schneider’s critique doesn’t kill EPI’s point. He shows average pay tracks average productivity when you match sectors and workers correctly. EPI asks whether typical workers shared in those gains. Both can be true, so it does not refute EPI’s point that the difference is rising inequality.
It seems to me that there is history behind this book. I used to read the magazine "Technology Review", back when it was about "technology and policy". The general idea seemed to be that there was an academic discipline of "technology policy" which was about the sort of thing the book recommends, that some government organ can regulate which technologies are deployed and how for the public good. It was somewhat Bolshie in that it never recommended leaving something to the market. In one article, it argued that airline deregulation was a failure because in the first five years or so, airline companies lost a lot of money. Somehow the democratizing of air travel wasn't significant.
Looking at the quoted examples, I suspect that the book's concept of social good is minimizing economic disruption for workers, ensuring that each group remains where it is on the socioeconomic totem pole. Notice that the authors don't object to Japanese real wages falling for decades but they do object to sharp changes in relative wages between groups in the US.
The only version of this concept I've heard of being successful was AT&T's management of the introduction of new technologies into the telephone industry. Reports are that AT&T was careful to regulate which technologies would be deployed and when to ensure minimum disruption (including that no capital investment was obsoleted before it was paid off). That came to an abrupt end when AT&T was broken up and ceased to be a monopoly that could control telephone technology.
> Subsequently, Haber and other scientists used the same ideas to design chemical weapons that killed and maimed hundreds of thousands on World War I battlefields.
This passage is even odder because making ordinary explosives requires fixed nitrogen, which during WW I, Germany used the Haber-Bosch process to supply to circumvent the British blockade of guano imports. So if you want to complain about "weapons enabled by the Haber-Bosch process", the major category was ordinary munitions.
Well done... it is not easy to be critical of the anointed leaders in a field. It takes courage.
You lost me on this. Zero interest. Way too long.
'Set out to prove' books always risk not being true. Books written by two authors always risk self-contradiction and confusion. This was the perfect storm, by the sounds of it.
I'm just here to send love to Giggles. <3
"Acemoglu and Johnson have a reason for employing a definition [of power] this infinitely broad; it allows them to include persuasion and compulsion in a single category of “power”."
I agree. This is the Foucauldian move though; I love Foucault's work but I hate that the term he came up for what he was trying to illustrate -- how science and scientific experts regulate our lives -- he ended up calling "power." Now anything you don't like -- you can just call it "power." https://computingandsociety.substack.com/p/what-was-michel-foucault-guilty-of
Looks like Acemoglu and Johnson's use of "power" is even worse!