Econ blogger/Substack royalty Noah Smith tried his best to defend Marc Andreessen’s techno-optimist manifesto last week, with a 4,000+ word post titled “Thoughts on techno-optimism.” It’s a thorough post, in which Smith offers a more intellectually defensible version of
Once upon a time, chemistry was the field of miracles. Aniline dyes, nylon, Teflon, the green revolution in agriculture and others were great advances and proof of DuPont’s slogan, “Better Things for Better Living ... Through Chemistry.” And there was a lot of good that came from chemical progress in the 20th century, likely doing more to improve the lives of more people worldwide than the Silicon Valley has achieved up till now. But during the last third of the 20th century, the toll from poor oversight and lax regulations began to be noticed. Love Canal and Bhopal are famous, but there are dozens of other superfund sites and fatal accidents that occurred during that era. The OSHA Process Safety Management (PSM) standard was issue early in my career in chemical manufacturing(1994). It was a direct response to Bhopal and other chemical disasters. My initial reaction, and that of many others, was that the new rule would make chemical manufacturing almost impossible, since it required so much new work that we weren’t doing. But over time, as we learned what the 14 sections of the regulation really meant, it made us better engineers. Rather than moving fast and breaking things (and killing people), we learned to plan better, train better, and made sure everyone went home to their families after work. I expect amazing technological advances to be made in the 21st century, but we don’t need to let tech billionaires experiment of the rest of us without regulations or consequences in order to create a better future.
This is a really interesting comparison. I don't know much about the chemical industry though, and I've always wondered whether there is some good long form article that would give me a potted history and compare it to the promises of tech?
Sorry, I can't point you to such an article. It's not an exact parallel with tech, because rather than being disruptive of the industry, it was big companies like Dow and DuPont that were funding research and then celebrating their new products in glowing television adds. Perhaps the best cultural touchstone is the one word piece of advice that Dustin Hoffman is given in the movie The Graduate: "Plastics."
Thanks for your reply. I am really interesting the comparison between the two industries' development and how they differed. But yes, I will keep looking.
Great piece, and I totally agree that the "passive" framing is mistaken, as it is only passive if the norms & regulations of the world fit your priors (e.g. unfettered capitalism). In my realm of studying television, I try to debunk the term "deregulation" which became so popular in the 1980s, as more accurately "regulatory shift" - the media industries wanted to eliminate some regulations, like ownership caps and educational programming mandates, but not those that let them make profits, like copyright. Same with Andreessen's ilk - they love the "passive" existing regulations and norms that work for them, decrying those that threaten them (like safety and accountability).
"So I think that the manifesto does have a particular political purpose. It’s intended to split and divide - to drive out the heretics - rather than to unite people under a single banner. Very likely, from Andreessen’s perspective, the angry response is a feature, not a bug, creating a line of division that he wants to use to drive his own political agenda in internecine disputes within Silicon Valley."
Myself, I am doubtful. I freely concede that Farrell is a lot smarter than me. And sure, Andreesen likely *had a purpose in mind* when he sat down to write, and no doubt that purpose was political. But I am not seeing how heightening the contradictions helps Andreesen, nor why Andreesen might think that it would. I think it likely that the purpose was less cerebral and more reflexive, something as simple as "leave me alone, and definitely don't sue me for liability in the Compound DAO case."
Farrell may be smarter than you, and I know you are smarter than me, but Andreesen is a very lucky man with a knack for the VC game. The argument that regulation and oversight inherently stifle innovation is the purest of hokum. It's the equivalent of a salesman telling you that contract is just boilerplate, go ahead and sign, we can add an addendum later if we need to. I don't think Andreesen is enough of a Steve Bannon to write a manifesto to encourage division, I'd bet the seed corn he believes every word in that cheese log of a document. Why not? It's worked out great for him and his freinds.
One thing I never see in the breathless pronouncements and manifestos of techno-utopian cheerleaders is a solid, evidence-based discussion of how their prescriptions would reverse the trend of resource overuse.
The basic problem is that every one of these optimists is one precisely because they handwave away the elephant in the room: Jevons' Paradox. Everybody loves a good green/clean tech story, but how many technological innovations have ended up creating induced demand and increasing our bootprint on the ecosphere? Spoiler alert: almost every single one. Look at the most recent waves of techbro hype (crypto scams and weaponised plagiarism) and plot the electricity demand that they spawned. This is not just a theoretical exercise...power prices, blackouts, resource misallocation, and increased emissions are all happening because of credulous followers looking to grab some unearned income by shuffling bored apes in pixel form.
What we need to have a conversation about is how all of this is an emergent property of late-stage catabolic capitalism, and the demonstrable fact that market failures like climate chaos will never be solved by markets themselves. Technology as saviour is just another deus ex machina and simply because the latest starry-eyed TED talker doesn't breathe a word about biophysical limits to growth does not erase their existence.
I remember a report from the Club of Rome (I think? I was a teenager) that we'd run out of oil. There was scoffing at this, and then fracking was invented, kind of making the exact point of the scoffers. As prices go up, the incentive to find more of whichever resource would increase, and *poof*, problem solved. Human ingenuity to the rescue!
I am probably not smart enough, certainly not hard-working enough, to come up with a good argument that, market forces notwithstanding, the damage caused by resource overuse is so grave that even if some Randian hero comes up with a brilliant new energy source, they'll be marketing it to a room full of people in hazmat suits with built-in cooling.
I do think that warning people against a hazard that is not currently killing them is pretty much impossible. I'm just glad I'm going to be dead within 30 years (less probably).
The first use of "fracking" ("hydraulic fracturing") dates back to 1947. The first commercially successful application of the technique dates to 1950. Awareness of the existence of massive shale in various locations in North America, particularly Alberta, is even more ancient. But until recently, these were classified as "unconventional oil" sources, precisely because it was understood that the yields from exploiting these highly diffuse deposits would be orders of magnitude smaller than relative to the investments that would be required than earlier generations of fossil fuels. As large as these reserves were, because the cost of recovery, they were refarded as scraping the bottom of the barrel. And that has largely held true: if you look at ratios of energy invested to in recovery to energy recovered, we are talking about something around an order of magnitude decline (ie, from around a barrel of oil equivalent invested to forty recovered in the 1950s, down to about a barrel invested to only about four recovered today).
I suppose to be more precise we could talk about the advent of computer-aided directional drilling, which was the technological game changer that made fracking feasible at scale.
A lot of people say that, but the actual predictions were pretty close to what happened if one compares predictions and empirical time series. The report included more than one model, but they were all roughly on track. It's like Cassandra in the Iliad or Ehrlich's Population Bomb. People don't like to hear forecasts that make them question their assumptions. Who can blame them? I suggest re-reading the original report and then looking at some of the recent assessments.
Nope. His doomsday scenarios may have been exaggerated, but the problems he pointed out remain. A lot of that is because a lot of his suggestions for action were taken. I recently reread his book, and was surprised.
I love the visuals in your Randian hero example. Have you seen any of the Yes Men videos?
The fracking boom shows up another potent factor that gets left out of the conversations: EROEI. As we scrape deeper and harder at the bottom of the fossil barrel, the amount of useful energy left over will always be diminishing. Shale oil and gas has been like a kid spending his allowance at the candy store and bingeing the entire purchase. Now that the Bakken is going into terminal decline, with the Permian soon to follow, the oil-dependent sectors of the economy are passed out on the sidewalk in a metabolic coma with all the empty wrappers piled around them.
"Present-day techno-optimists like Andreessen, Musk, and Sam Altman are saying the same thing, hoping we notice how the benefits of digital futures’ past have been apportioned." Did you mean the techno-optimists are hoping we _don't_ notice how the benefits have been apportioned? Otherwise, I don't see how this sentence fits with (what I understood to be) your thesis re Luddites and so on.
"regulators and journalists and trust & safety professionals and sustainability advocates stay out of the way." Or stay only optimally in the way.
There should be some consideration of environmental impact of drilling holes in the ground, but it should not be more difficult to drill a hole looking to heat or to pump CO2 IN than it is to get oil and gas OUT. There should not be less risk of killing people with a nuclear power plant than a solar or a coal power plant. FDA should consider the costs of delaying drug and therapy development by not having reciprocal agreement with other regulators. ETC.
I think the regulatory pendulum has swung far past the cost benefit optimum. And considering the spillovers, we vastly underfund R&D.
I really don't think it is important what you call this position, "optimistic" "realistic" "pragmatic" "neoliberal, "moderate," "traditional," "Centrist."
I recommend Justin Joque, "Revolutionary Mathematics: Algorithms, Statistics, and the Logic of Capitalism", out recently from Verso Books, for an important corrective tonic to a lot of today's "techno-optimism". Joque's book in some ways recapitulates a lot of analogous arguments made in the past about capitalism's propensities for breakdown and crises (eg, Henryk Grossman, Michael Hudson, etc), but he focuses usefully on the current, especially hottest commodities of techno-optimism, ie, "artificial intelligence", algorithmic optimization, etc.
What do the academic "replication crisis", VW and other automakers' systematic cheating on emissions controls, and fears about "unintended consequences" of AI have to do with each other? Joque argues that they are all driven by an underlying "logic of capitalism" and its imperative of generating profits for its ownership class that sets up inevitable collisions with the common good, and hard limits on the benefits of innovation.
Spot on. The fact that so much of Smith's argument could be rewritten, the only edit being to replace the word optimist with the word pessimist, and it would basically mean the same thing, is what first struck me while reading it too. The key thing to note with the flavour of Tech Optimism Andreessen would have us adopt is that he wishes us to be fully passive, to join him as helpless observers in his Techno Optimism and leave the companies and tech he backs and their ability to make profits, unencumbered by pesky regulation. And he devises a nasty way of bullying us into that, as I wrote about here: https://atomless.substack.com/p/adreessens-basilisk
Scientific achievements and world changing technologies have never needed independent public or Government funding. Nor do they need help from Silicon Valley. Meta did not sponsor the creation of the steam engine. Public funding did not create the airplane. GPS was built directly by the US Government and not given grants as an external project to some third person dreaming of the future.
Point is the way people view technology now is as if only Government or massive investors like companies can create it or birth it into existence. World changing technology is created by individuals with a dream and the skills and contact to create that dream. It is then up to the public and industry to adopt the new technology, not Government policies, Silicon Valley, or idiot investors looking to extract all revenue from an idea
And yes there is a difference between Government projects and Government investment in external projects. JWST would not be possible at that scale without NASA and the US Gov and contractors. Also investments to scale and investment to create are two different things. Again, investors don't create and are only involved at the end of the process
"[The 'Active' statements] strike me as little more than a framing exercise!"
Another way to look at this would be to try to assess the chance of success. Let's say that we decide that we want to increase the amount of technological discovery (or the distribution of benefits from technology), how likely is it that basically normal policy and management decisions will lead to success.
Will it require good luck (or heroic leadership) or is it something that we more or less know how to do?
I think the optimist position would give much higher odds of success.
And a related bit (which I didn't get into here because I have midterms to grade this week and only have time for a shorter piece) is the underlying theory-of-change.
I think there's an optimist theory-of-change that postulates the most effective route to better policies is by engaging the better angels of our existing elites' nature.
Then there's an agonist theory-of-change, arguing that you're more likely to arrive at improved policy choices by negative external pressure on existing elites. (I touched on this a few years ago actually, in my essay criticizing the 10,000 year clock). https://www.wired.com/story/the-10000-year-clock-is-a-waste-of-time/
I think the agonistic view of politics is more accurate. But it's another legitimate dimension of disagreement.
One charitable interpretation of Stewart Brand's whole career is that he foresaw the nexus of unaccountable money and power that computer technology was actually going to be and thought the way forward might be wishcasting it as egalitarian to the powerful to make it more so.
Once upon a time, chemistry was the field of miracles. Aniline dyes, nylon, Teflon, the green revolution in agriculture and others were great advances and proof of DuPont’s slogan, “Better Things for Better Living ... Through Chemistry.” And there was a lot of good that came from chemical progress in the 20th century, likely doing more to improve the lives of more people worldwide than the Silicon Valley has achieved up till now. But during the last third of the 20th century, the toll from poor oversight and lax regulations began to be noticed. Love Canal and Bhopal are famous, but there are dozens of other superfund sites and fatal accidents that occurred during that era. The OSHA Process Safety Management (PSM) standard was issue early in my career in chemical manufacturing(1994). It was a direct response to Bhopal and other chemical disasters. My initial reaction, and that of many others, was that the new rule would make chemical manufacturing almost impossible, since it required so much new work that we weren’t doing. But over time, as we learned what the 14 sections of the regulation really meant, it made us better engineers. Rather than moving fast and breaking things (and killing people), we learned to plan better, train better, and made sure everyone went home to their families after work. I expect amazing technological advances to be made in the 21st century, but we don’t need to let tech billionaires experiment of the rest of us without regulations or consequences in order to create a better future.
This is a really interesting comparison. I don't know much about the chemical industry though, and I've always wondered whether there is some good long form article that would give me a potted history and compare it to the promises of tech?
Sorry, I can't point you to such an article. It's not an exact parallel with tech, because rather than being disruptive of the industry, it was big companies like Dow and DuPont that were funding research and then celebrating their new products in glowing television adds. Perhaps the best cultural touchstone is the one word piece of advice that Dustin Hoffman is given in the movie The Graduate: "Plastics."
Thanks for your reply. I am really interesting the comparison between the two industries' development and how they differed. But yes, I will keep looking.
Great piece, and I totally agree that the "passive" framing is mistaken, as it is only passive if the norms & regulations of the world fit your priors (e.g. unfettered capitalism). In my realm of studying television, I try to debunk the term "deregulation" which became so popular in the 1980s, as more accurately "regulatory shift" - the media industries wanted to eliminate some regulations, like ownership caps and educational programming mandates, but not those that let them make profits, like copyright. Same with Andreessen's ilk - they love the "passive" existing regulations and norms that work for them, decrying those that threaten them (like safety and accountability).
So what did you think of Henry Farrell's theory that the whole point of Andreesen's screed was to provoke angry responses? (https://programmablemutter.substack.com/p/marc-andreessen-wanted-to-make-people)
Here is the summary:
"So I think that the manifesto does have a particular political purpose. It’s intended to split and divide - to drive out the heretics - rather than to unite people under a single banner. Very likely, from Andreessen’s perspective, the angry response is a feature, not a bug, creating a line of division that he wants to use to drive his own political agenda in internecine disputes within Silicon Valley."
Myself, I am doubtful. I freely concede that Farrell is a lot smarter than me. And sure, Andreesen likely *had a purpose in mind* when he sat down to write, and no doubt that purpose was political. But I am not seeing how heightening the contradictions helps Andreesen, nor why Andreesen might think that it would. I think it likely that the purpose was less cerebral and more reflexive, something as simple as "leave me alone, and definitely don't sue me for liability in the Compound DAO case."
Farrell may be smarter than you, and I know you are smarter than me, but Andreesen is a very lucky man with a knack for the VC game. The argument that regulation and oversight inherently stifle innovation is the purest of hokum. It's the equivalent of a salesman telling you that contract is just boilerplate, go ahead and sign, we can add an addendum later if we need to. I don't think Andreesen is enough of a Steve Bannon to write a manifesto to encourage division, I'd bet the seed corn he believes every word in that cheese log of a document. Why not? It's worked out great for him and his freinds.
One thing I never see in the breathless pronouncements and manifestos of techno-utopian cheerleaders is a solid, evidence-based discussion of how their prescriptions would reverse the trend of resource overuse.
The basic problem is that every one of these optimists is one precisely because they handwave away the elephant in the room: Jevons' Paradox. Everybody loves a good green/clean tech story, but how many technological innovations have ended up creating induced demand and increasing our bootprint on the ecosphere? Spoiler alert: almost every single one. Look at the most recent waves of techbro hype (crypto scams and weaponised plagiarism) and plot the electricity demand that they spawned. This is not just a theoretical exercise...power prices, blackouts, resource misallocation, and increased emissions are all happening because of credulous followers looking to grab some unearned income by shuffling bored apes in pixel form.
What we need to have a conversation about is how all of this is an emergent property of late-stage catabolic capitalism, and the demonstrable fact that market failures like climate chaos will never be solved by markets themselves. Technology as saviour is just another deus ex machina and simply because the latest starry-eyed TED talker doesn't breathe a word about biophysical limits to growth does not erase their existence.
I remember a report from the Club of Rome (I think? I was a teenager) that we'd run out of oil. There was scoffing at this, and then fracking was invented, kind of making the exact point of the scoffers. As prices go up, the incentive to find more of whichever resource would increase, and *poof*, problem solved. Human ingenuity to the rescue!
I am probably not smart enough, certainly not hard-working enough, to come up with a good argument that, market forces notwithstanding, the damage caused by resource overuse is so grave that even if some Randian hero comes up with a brilliant new energy source, they'll be marketing it to a room full of people in hazmat suits with built-in cooling.
I do think that warning people against a hazard that is not currently killing them is pretty much impossible. I'm just glad I'm going to be dead within 30 years (less probably).
"Then fracking was invented". Uhhh, no.
The first use of "fracking" ("hydraulic fracturing") dates back to 1947. The first commercially successful application of the technique dates to 1950. Awareness of the existence of massive shale in various locations in North America, particularly Alberta, is even more ancient. But until recently, these were classified as "unconventional oil" sources, precisely because it was understood that the yields from exploiting these highly diffuse deposits would be orders of magnitude smaller than relative to the investments that would be required than earlier generations of fossil fuels. As large as these reserves were, because the cost of recovery, they were refarded as scraping the bottom of the barrel. And that has largely held true: if you look at ratios of energy invested to in recovery to energy recovered, we are talking about something around an order of magnitude decline (ie, from around a barrel of oil equivalent invested to forty recovered in the 1950s, down to about a barrel invested to only about four recovered today).
https://en.wikipedia.org/wiki/Fracking?wprov=sfla1
I suppose to be more precise we could talk about the advent of computer-aided directional drilling, which was the technological game changer that made fracking feasible at scale.
That Club of Rome report recently celebrated its 50th anniversary. The general verdict is that the forecasts were pretty accurate.
They seem to be totally wrong to me.
A lot of people say that, but the actual predictions were pretty close to what happened if one compares predictions and empirical time series. The report included more than one model, but they were all roughly on track. It's like Cassandra in the Iliad or Ehrlich's Population Bomb. People don't like to hear forecasts that make them question their assumptions. Who can blame them? I suggest re-reading the original report and then looking at some of the recent assessments.
https://www.resilience.org/stories/2021-08-16/revisiting-the-limits-to-growth/#:~:text=The%20paper%20is%20by%20Gaya%20Herrington%20%28formerly%20Gaya,1972%2C%20and%20the%2030-year%20update%20published%20in%202002.
https://mahb.stanford.edu/wp-content/uploads/2021/07/yale-publication-1.pdf
“Ehrlich's Population Bomb”.
Surely has been totally disproven.
Nope. His doomsday scenarios may have been exaggerated, but the problems he pointed out remain. A lot of that is because a lot of his suggestions for action were taken. I recently reread his book, and was surprised.
I love the visuals in your Randian hero example. Have you seen any of the Yes Men videos?
The fracking boom shows up another potent factor that gets left out of the conversations: EROEI. As we scrape deeper and harder at the bottom of the fossil barrel, the amount of useful energy left over will always be diminishing. Shale oil and gas has been like a kid spending his allowance at the candy store and bingeing the entire purchase. Now that the Bakken is going into terminal decline, with the Permian soon to follow, the oil-dependent sectors of the economy are passed out on the sidewalk in a metabolic coma with all the empty wrappers piled around them.
"Present-day techno-optimists like Andreessen, Musk, and Sam Altman are saying the same thing, hoping we notice how the benefits of digital futures’ past have been apportioned." Did you mean the techno-optimists are hoping we _don't_ notice how the benefits have been apportioned? Otherwise, I don't see how this sentence fits with (what I understood to be) your thesis re Luddites and so on.
Yep that's exactly what I meant. Fixing the text now, thanks for catching it.
Glad I could help.
"regulators and journalists and trust & safety professionals and sustainability advocates stay out of the way." Or stay only optimally in the way.
There should be some consideration of environmental impact of drilling holes in the ground, but it should not be more difficult to drill a hole looking to heat or to pump CO2 IN than it is to get oil and gas OUT. There should not be less risk of killing people with a nuclear power plant than a solar or a coal power plant. FDA should consider the costs of delaying drug and therapy development by not having reciprocal agreement with other regulators. ETC.
I think the regulatory pendulum has swung far past the cost benefit optimum. And considering the spillovers, we vastly underfund R&D.
I really don't think it is important what you call this position, "optimistic" "realistic" "pragmatic" "neoliberal, "moderate," "traditional," "Centrist."
I recommend Justin Joque, "Revolutionary Mathematics: Algorithms, Statistics, and the Logic of Capitalism", out recently from Verso Books, for an important corrective tonic to a lot of today's "techno-optimism". Joque's book in some ways recapitulates a lot of analogous arguments made in the past about capitalism's propensities for breakdown and crises (eg, Henryk Grossman, Michael Hudson, etc), but he focuses usefully on the current, especially hottest commodities of techno-optimism, ie, "artificial intelligence", algorithmic optimization, etc.
What do the academic "replication crisis", VW and other automakers' systematic cheating on emissions controls, and fears about "unintended consequences" of AI have to do with each other? Joque argues that they are all driven by an underlying "logic of capitalism" and its imperative of generating profits for its ownership class that sets up inevitable collisions with the common good, and hard limits on the benefits of innovation.
Spot on. The fact that so much of Smith's argument could be rewritten, the only edit being to replace the word optimist with the word pessimist, and it would basically mean the same thing, is what first struck me while reading it too. The key thing to note with the flavour of Tech Optimism Andreessen would have us adopt is that he wishes us to be fully passive, to join him as helpless observers in his Techno Optimism and leave the companies and tech he backs and their ability to make profits, unencumbered by pesky regulation. And he devises a nasty way of bullying us into that, as I wrote about here: https://atomless.substack.com/p/adreessens-basilisk
Scientific achievements and world changing technologies have never needed independent public or Government funding. Nor do they need help from Silicon Valley. Meta did not sponsor the creation of the steam engine. Public funding did not create the airplane. GPS was built directly by the US Government and not given grants as an external project to some third person dreaming of the future.
Point is the way people view technology now is as if only Government or massive investors like companies can create it or birth it into existence. World changing technology is created by individuals with a dream and the skills and contact to create that dream. It is then up to the public and industry to adopt the new technology, not Government policies, Silicon Valley, or idiot investors looking to extract all revenue from an idea
And yes there is a difference between Government projects and Government investment in external projects. JWST would not be possible at that scale without NASA and the US Gov and contractors. Also investments to scale and investment to create are two different things. Again, investors don't create and are only involved at the end of the process
"[The 'Active' statements] strike me as little more than a framing exercise!"
Another way to look at this would be to try to assess the chance of success. Let's say that we decide that we want to increase the amount of technological discovery (or the distribution of benefits from technology), how likely is it that basically normal policy and management decisions will lead to success.
Will it require good luck (or heroic leadership) or is it something that we more or less know how to do?
I think the optimist position would give much higher odds of success.
And a related bit (which I didn't get into here because I have midterms to grade this week and only have time for a shorter piece) is the underlying theory-of-change.
I think there's an optimist theory-of-change that postulates the most effective route to better policies is by engaging the better angels of our existing elites' nature.
Then there's an agonist theory-of-change, arguing that you're more likely to arrive at improved policy choices by negative external pressure on existing elites. (I touched on this a few years ago actually, in my essay criticizing the 10,000 year clock). https://www.wired.com/story/the-10000-year-clock-is-a-waste-of-time/
I think the agonistic view of politics is more accurate. But it's another legitimate dimension of disagreement.
One charitable interpretation of Stewart Brand's whole career is that he foresaw the nexus of unaccountable money and power that computer technology was actually going to be and thought the way forward might be wishcasting it as egalitarian to the powerful to make it more so.