I really like Eno’s take on capitalism. I think about capitalism often. And I think it’s part of the problem … if not the entire problem his take that the design optimizes for very negative things, and I can’t help but agree … However, playing “devil’s advocate”for a bit …
I spent multiple decades in the semiconductor industry. The costs back in 1983 were high, but TODAY?
The cost of a modern fab is so astronomical that it’s hard to imagine any system other than capitalism pulling it off. We’re talking $20 billion+ per fab at the cutting edge, with individual machines (like ASML’s EUV lithography systems) costing upwards of $300 million each. The sheer financial scale means that any alternative system has to answer a brutal question: who takes the risk, and how do they get rewarded?
TSMC has nine (9!!!) fabs in Taiwan alone … their capital investment is on the order of $200 billion. Who else is going to do that?
And modern life as we know it would be impossible without semiconductors—and by extension, without the capitalist system that built them.
Take away advanced semiconductors, and we lose nearly everything:
No modern computers or smartphones.
No internet, because servers, networking gear, and fiber optics all rely on silicon.
No advanced medical imaging (MRI, CT scans, ultrasound).
No GPS, meaning no modern transportation or logistics.
No advanced power grids, which rely on microcontrollers, sensors, and power semiconductors.
No modern industry—factories, automation, robotics, and even agriculture are all deeply dependent on semiconductors.
Even cars would revert to mechanical-only systems, and the global economy would be thrown back to the mid-20th century (or earlier). A world without semiconductors isn’t just “slightly less convenient”—it’s a collapse of nearly every major system that keeps society running.
From a musician’s perspective, no synthesizers, no recording equipment: even tape based multitrack relied on discrete semiconductors.
Looking at history, capitalism wasn’t the only system that tried to industrialize, but it was the only one that succeeded in pushing semiconductor tech to where it is now. The Soviet Union, for example, had a state-controlled electronics industry, but it lagged behind Western semiconductor development by at least a decade—and that gap widened over time. Even today, China’s state-driven efforts haven’t been able to break free from reliance on capitalist supply chains.
The fundamental problem is that semiconductor development requires insane levels of risk-taking, competition, and specialization. Every node shrink, every new fabrication process, and every innovation in chip design comes from a brutal process of iterative failure, financial risk, and cutthroat competition. Governments can throw money at it, but they struggle to match the efficiency of private-sector-driven innovation.
I mean … mind you, I am no fan of what capitalism has become, but modern life would be impossible without it … maybe people wouldn’t mind going back to the 1940s though. We’d certainly have to revert back 85 years … and people would have to learn to read paper maps again. :^)
It is completely understandable that an an engineer would love the development of chips in the last fifty years. It is indeed impressive and remarkable. But it has come at a cost. I would point out that all of that incredible progress you described has in fact pushed us closer to the edge. Species extinction, climate change, AI use of electricity, all these are related. And the real bottom line is that people aren’t happier. People die younger than they did thirty years ago. Social media, doom scrolling, looking for the next dopamine hit, the rise of fascism all over the planet, and more inequality that ever… I’ve seen comparisons that show inequality is worse than at the time of the French Revolution.
We are not very good at taking the long view, are we? We charge ahead and then hope to fix things later. We get so drunk on our power that we miss the exit ramp and suddenly we hurtle down the path so fast…
>We are not very good at taking the long view, are we? We charge ahead and then hope to fix things later. We get so drunk on our power that we miss the exit ramp and suddenly we hurtle down the path so fast…
The root cause of this seems (to me) to be what Paul Humphreys calls “epistemic opacity” from his book “Extending Ourselves: Computational Science, Empiricism, and Scientific Method.” – i.e., “processes, systems, or knowledge structures are not fully understandable or transparent to human cognition, either due to their complexity, inaccessibility, or the limitations of human reasoning.” ( I would add “greed” but Humphreys didn’t)
At least, that’s the nice version. The not-so-nice version is we don’t care about the future and future consequences … as an aggregate species. We’re just not interested. At least that’s how it seems from my seat out here on the plains in “flyover country.”
But also … Those of us that are in science and applied science (engineering) really don’t consider future consequences due to this epistemic opacity. I don’t really think all those scientists involved in the Manhattan Project were evil … I just think they didn’t really have an appreciation for what their work would propagate into the future. Same with computers: all that was being looked for was a way to switch telephone calls faster as the telephone network grew. And the same is true of the internet. When I got on it there were ~800 nodes on the whole thing … world wide. No one contemplated “social media” … all we wanted to do was exchange files electronically instead of mailing 9-track data tape to each other via UPS.
So … all the things you list: “… Species extinction, climate change, AI use of electricity, all these are related. And the real bottom line is that people aren’t happier. People die younger than they did thirty years ago. Social media, doom scrolling, looking for the next dopamine hit, the rise of fascism all over the planet, and more inequality that ever… I’ve seen comparisons that show inequality is worse than at the time of the French Revolution …”
To be honest, I cannot disagree with anything you wrote.
As a 23 year old newly minted BSEE/CS student, I was gonna do all kinds of “cool stuff” … The (technical) world was “my oyster” and all that … design new devices with exotic new materials … super quiet amplification devices with infinite bandwidth … blah, blah, blah … That was 42 years ago. But … four+ decades is a long time, and things definitely went sideways: the big HARD pivot took place at two significant points in fairly recent history- 1995: The National Science Foundation decommissioned NSFNET, and private companies (like AT&T, MCI, and Sprint) took over the backbone of the internet, allowing full commercialization. This facilitated what took place a decade later: Facebook, YouTube, Twitter. And here we are. It is not coincidence that most of what you indicate above has taken place between 1995 and today. (Well … except for climate change … that one has a much longer event horizon going back to late 1800s)
I think it will get much worse before the trajectory changes direction in the slightest. Which gets back to what Eno says in the interview: if you optimise a system for a specific thing (such as “engagement” or “profit” instead of “happiness” or “human improvement” ) don’t be surprised if you get extreme negative externalities.
I really like Eno’s take on capitalism. I think about capitalism often. And I think it’s part of the problem … if not the entire problem his take that the design optimizes for very negative things, and I can’t help but agree … However, playing “devil’s advocate”for a bit …
I spent multiple decades in the semiconductor industry. The costs back in 1983 were high, but TODAY?
The cost of a modern fab is so astronomical that it’s hard to imagine any system other than capitalism pulling it off. We’re talking $20 billion+ per fab at the cutting edge, with individual machines (like ASML’s EUV lithography systems) costing upwards of $300 million each. The sheer financial scale means that any alternative system has to answer a brutal question: who takes the risk, and how do they get rewarded?
TSMC has nine (9!!!) fabs in Taiwan alone … their capital investment is on the order of $200 billion. Who else is going to do that?
And modern life as we know it would be impossible without semiconductors—and by extension, without the capitalist system that built them.
Take away advanced semiconductors, and we lose nearly everything:
No modern computers or smartphones.
No internet, because servers, networking gear, and fiber optics all rely on silicon.
No advanced medical imaging (MRI, CT scans, ultrasound).
No GPS, meaning no modern transportation or logistics.
No advanced power grids, which rely on microcontrollers, sensors, and power semiconductors.
No modern industry—factories, automation, robotics, and even agriculture are all deeply dependent on semiconductors.
Even cars would revert to mechanical-only systems, and the global economy would be thrown back to the mid-20th century (or earlier). A world without semiconductors isn’t just “slightly less convenient”—it’s a collapse of nearly every major system that keeps society running.
From a musician’s perspective, no synthesizers, no recording equipment: even tape based multitrack relied on discrete semiconductors.
Looking at history, capitalism wasn’t the only system that tried to industrialize, but it was the only one that succeeded in pushing semiconductor tech to where it is now. The Soviet Union, for example, had a state-controlled electronics industry, but it lagged behind Western semiconductor development by at least a decade—and that gap widened over time. Even today, China’s state-driven efforts haven’t been able to break free from reliance on capitalist supply chains.
The fundamental problem is that semiconductor development requires insane levels of risk-taking, competition, and specialization. Every node shrink, every new fabrication process, and every innovation in chip design comes from a brutal process of iterative failure, financial risk, and cutthroat competition. Governments can throw money at it, but they struggle to match the efficiency of private-sector-driven innovation.
I mean … mind you, I am no fan of what capitalism has become, but modern life would be impossible without it … maybe people wouldn’t mind going back to the 1940s though. We’d certainly have to revert back 85 years … and people would have to learn to read paper maps again. :^)
It is completely understandable that an an engineer would love the development of chips in the last fifty years. It is indeed impressive and remarkable. But it has come at a cost. I would point out that all of that incredible progress you described has in fact pushed us closer to the edge. Species extinction, climate change, AI use of electricity, all these are related. And the real bottom line is that people aren’t happier. People die younger than they did thirty years ago. Social media, doom scrolling, looking for the next dopamine hit, the rise of fascism all over the planet, and more inequality that ever… I’ve seen comparisons that show inequality is worse than at the time of the French Revolution.
We are not very good at taking the long view, are we? We charge ahead and then hope to fix things later. We get so drunk on our power that we miss the exit ramp and suddenly we hurtle down the path so fast…
>We are not very good at taking the long view, are we? We charge ahead and then hope to fix things later. We get so drunk on our power that we miss the exit ramp and suddenly we hurtle down the path so fast…
The root cause of this seems (to me) to be what Paul Humphreys calls “epistemic opacity” from his book “Extending Ourselves: Computational Science, Empiricism, and Scientific Method.” – i.e., “processes, systems, or knowledge structures are not fully understandable or transparent to human cognition, either due to their complexity, inaccessibility, or the limitations of human reasoning.” ( I would add “greed” but Humphreys didn’t)
At least, that’s the nice version. The not-so-nice version is we don’t care about the future and future consequences … as an aggregate species. We’re just not interested. At least that’s how it seems from my seat out here on the plains in “flyover country.”
But also … Those of us that are in science and applied science (engineering) really don’t consider future consequences due to this epistemic opacity. I don’t really think all those scientists involved in the Manhattan Project were evil … I just think they didn’t really have an appreciation for what their work would propagate into the future. Same with computers: all that was being looked for was a way to switch telephone calls faster as the telephone network grew. And the same is true of the internet. When I got on it there were ~800 nodes on the whole thing … world wide. No one contemplated “social media” … all we wanted to do was exchange files electronically instead of mailing 9-track data tape to each other via UPS.
So … all the things you list: “… Species extinction, climate change, AI use of electricity, all these are related. And the real bottom line is that people aren’t happier. People die younger than they did thirty years ago. Social media, doom scrolling, looking for the next dopamine hit, the rise of fascism all over the planet, and more inequality that ever… I’ve seen comparisons that show inequality is worse than at the time of the French Revolution …”
To be honest, I cannot disagree with anything you wrote.
As a 23 year old newly minted BSEE/CS student, I was gonna do all kinds of “cool stuff” … The (technical) world was “my oyster” and all that … design new devices with exotic new materials … super quiet amplification devices with infinite bandwidth … blah, blah, blah … That was 42 years ago. But … four+ decades is a long time, and things definitely went sideways: the big HARD pivot took place at two significant points in fairly recent history- 1995: The National Science Foundation decommissioned NSFNET, and private companies (like AT&T, MCI, and Sprint) took over the backbone of the internet, allowing full commercialization. This facilitated what took place a decade later: Facebook, YouTube, Twitter. And here we are. It is not coincidence that most of what you indicate above has taken place between 1995 and today. (Well … except for climate change … that one has a much longer event horizon going back to late 1800s)
I think it will get much worse before the trajectory changes direction in the slightest. Which gets back to what Eno says in the interview: if you optimise a system for a specific thing (such as “engagement” or “profit” instead of “happiness” or “human improvement” ) don’t be surprised if you get extreme negative externalities.