"Shipping containers didn't make anyone rich" is quite a take. Entire new categories of trade were born out of them. Sure, being the one making the containers didn't make one rich, so in that way it's different than the railroads, but it probably had a similar effect in terms of the new businesses that were made suddenly possible as a result.
I guess the result for AI is don't try to get rich inventing AI. Try to get rich using AI to enable things that weren't previously possible. I think that space has barely been explored - so far most of the activity is not net new things, but just trying to do the same thing we always did before, but faster/cheaper.
The shipping container insight is good. There was no one big winner, but many medium-sized ones. Levinson's "The Box" (2006) is the classic book on how that worked out. It's worth reading in the AI era.
"The Box" points out that shipping did not convert to containers as much as replace ports and shipping lines with new builds built for containers. The Port of San Francisco used to be a major port. But, when containers came along, a new container port was built in Oakland. One that didn't have an entrenched longshoreman's union.
A few things helped. BART was under construction, and digging up dirt that had to go somewhere. That dirt, as fill, became the container port.
London went in a similar direction. London never got a container port. Instead, a minor non-union port on the eastern coast was built up into a container port.
The London Docklands died, to come back decades later as office and residential space.
Containers needed all new ships, and made existing ships obsolete. Some shipping lines converted; others were superseded. There was no one big winner in containers.
It took quite a while to get trucking and railroads on board. Railroads had tried various intermodal schemes - trailer on flat car, trailers that could run on rails, and such. These took less infrastructure for loading and unloading than the big maritime boxes. But eventually the dumb boxes won out. Union Pacific even increased its clearances to allow double-stack container trains.
The US longshoremens' union got deals for lifetime support for their existing members. But there were not many new people entering the industry. If you visit a container port today (if you can get in) there won't be many people around. Here's what one looks like.[1] That's not a render; it's the port of Rotterdam, a decade ago. (The video is sped up; the AGVs are not that fast. This is real speed.[2]) All recent ports are at least that automated.
It's a good comparison. Perhaps everyone here saved $500 or more (perhaps much more? Depending how you want to do the accounting) due to containerized shipping.
But also a bunch of people lost their whole job and a historically powerful union lost a lot of clout, part of a broader trend of organized labor becoming less influential in politics.
Did we manage the transition well? Were we intentional at all? Did we maximize benefits and minimize losses? It's hard to say and depends on your values.
> The Port of San Francisco used to be a major port. But, when containers came along, a new container port was built in Oakland. One that didn't have an entrenched longshoreman's union.
Unlike the east coast longshoremen's unions, the west coast unions saw the writing on the wall. The Port of San Francisco containerized (the first one to start, IIRC), but couldn't handle the volume, thus the shift to Oakland. Other west coast ports managed to expand and prosper. By contrast, the Port of New York and other east coast port unions strongly resisted, and ended up with nothing as new containerized ports came online.
I don't remember if I read it in The Box or Philip Dray's There is Power in a Union, but notwithstanding the early radicalism of the longshoremen's union in SF and other lore, west coast unions, at least mid century, were much more cooperative with industry than on the east coast and their members faired technology transitions fairly well. It had alot to do with smart leadership during that period, but perhaps also just the growth of the economy on the west coast.
The metaphor doesn't hold, since, if gold was what was rushed for in the gold rush era, then AI is what is being rushed for in the AI era. AI models are the gold.
It anything, GPUs and data centers are the shovels. Building AI models is digging for gold. Unlike gold, AI models dont have preexisting or intrinsic value. So the money is in figuring out what to apply AI to.
I've been looking at leaving tech for various reasons, and my dad recently told me to go into geology to do site geology for hydro dams and geothermal and other kinds of energy.
Maybe he's just trying to get me to start selling shovels.
Shipping containers also just had the effect of making everyone a little bit richer. Everything that needed to get moved around the world, up and down the supply chain, got there faster and cheaper as a result.
does that really follow? I potentially get stuff faster, but is that "richer" value actually going to me, or the shipping company?
I get paid 50k a year. shipping stuff is expensive and i can only afford to buy 8k worth of stuff. that pays for 40 widgets.
I get paid 50k a year. shipping is cheaper but i can only afford to buy 8k worth of stuff. that pays for 55 widgets now. I have more crap, but my actual take home isn't more, I'm not richer, the costs of goods went down.
this is also why US wages are stagnant and why the US (and much of the world) is dependant on illegal or immigrant labor.
Every product got cheaper, labour also got cheaper. Could people afford more or less? Probably depends on how you measure. More porcelain knick-knacks and TVs. Less housing and food.
I would posit that it made both housing and food cheaper. A lot of what's required to build houses (plus the infrastructure they need) comes from overseas and would have come down in cost as a result. Unfortunately those savings don't do much in the face of an inability to actually build things, but houses would almost certainly be more expensive than they are now in a world with no shipping containers.
Same with food - even domestic food gets cheaper because a ton of potash and other stuff needed to grow it is shipped in. Also made people wealthier in the sense that it made it economical to bring in a much wider variety of foods from all over the world.
A lot of this sounded (from a quick read) interesting and insightful.
I wasn't sure about these two paragraphs:
> If AI had started a new wave, there would have been an extended period of uncertainty and experimentation. There would have been a population of early adopters experimenting with their own models. When thousands or millions of tinkerers use the tech to solve problems in entirely new ways, its uses proliferate. But because they are using models owned by the big AI companies, their ability to fully experiment is limited to what’s allowed by the incumbents, who have no desire to permit an extended challenge to the status quo.
> This doesn’t mean AI can’t start the next technological revolution. It might, if experimentation becomes cheap, distributed and permissionless—like Wozniak cobbling together computers in his garage, Ford building his first internal combustion engine in his kitchen, or Trevithick building his high-pressure steam engine as soon as James Watt’s patents expired. When any would-be innovator can build and train an LLM on their laptop and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption.
> ... In the 1970s, it took time for people to wrap their heads around the desirability of powerful and ubiquitous computing. But in 2025, machines that think better than previous machines are easy for people to understand.
Even if tinkerers aren't currently training frontier LLMs from scratch, there's still a lot of experimentation happening, and I don't think everyone even realizes yet how transformative this technology will be. On the contrary, there are a lot of skeptics.
That said, it's a good question, who captures the value.
On the question of tinkerers, I'm reminded of the man who just copy-pasted the middle layers of an LLM on top of each other: https://dnhkng.github.io/posts/rys/.
The computer version of this would read something like:
> When any would-be innovator can grow and etch a silicon CPU in their kitchen and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption.
Put that way, it seems clear this whole reaction is a category error.
The author clearly sees "AI" only as training models and ignores everything built on top of models. It's a common mistake that gets made for every new technology.
>Even if tinkerers aren't currently training frontier LLMs from scratch, there's still a lot of experimentation happening, and I don't think everyone even realizes yet how transformative this technology will be.
The big issue of vendor lock-in remains for the frontier models. We saw a small piece of this in the games industry when people started reverse compiling games at unprecedented speeds, and companies' answer was to push Anthropic to refusing to process such prompts. A very easy workaround for now, but they played their hand and revealed their allegiance.
Any innovation that threatens to disrupt with current structures will be stifled by the old guard.
> I don't think everyone even realizes yet how transformative this technology will be. On the contrary, there are a lot of skeptics.
Same occurred back at the start of the cloud era where sysadmins were certain no sane business owner would up load their business to the cloud
And well...all those businesses were run by nut jobs
Whole lot of IT people are hung up on linguistics and refuse to acknowledge there's a physical component of the machine and ultimately we're managing machine states for the users need not contextualizing machine needs with layers of dev abstraction
If devs were carpenters they'd have diamond crusted hammers and debate the visual splendor all day never finishing the house
I mean, it destroys far more value than it captures.
You can use it to clone Adobe tools, emulate games, copy whatever software you want.
I assume the original idea was that the big software companies would replace all their pesky, expensive engineers. Not that we’d simply burn down their companies on accident because we had a tool to clone whatever we wanted.
> If AI had started a new wave, there would have been an extended period of uncertainty and experimentation. There would have been a population of early adopters experimenting with their own models. When thousands or millions of tinkerers use the tech to solve problems in entirely new ways, its uses proliferate
Maybe I'm reading this incorrectly, but isn't that... exactly what's happening right now? The author goes on to write other related assumptions.
I can't say I agree or disagree, but the article seems a bit disconnected from the true state of the technology in question.
Right now, I see the application of ai at a strategic level to accomplish two things:
1. To clear out the last 40 years of technical debt that wasn't dealt with because "omg we don't have time for that".
2. To search for interesting problems that have never been attacked before in a much faster way than ever before.
---
AI coding has let me launch SaaS for non classical logic reasoning ( https://arbiter.housecarl.cloud/ ), along with experimenting with game ideas, plus implementing a Lisp 1.5 with JIT compilation. And more.
I spend a lot of time coding after work with ideas. Always have. AI supercharges that.
Great piece, but I think he's missing something, the next phase of AI adoption is hiding it, because AI-made work increasingly all looks the same. If everyone cuts costs with the same models, cheap output becomes a commodity, and the real edge may go to companies whose work doesn’t look AI-made at all.
That's what I see. I run a small e-signature service. I built it mostly with AI. Building is inexpensive, but it's like that for everyone, so I can't charge for it. I price at cost plus a little and the saving goes to the customer.
Got my first paying customer today, $19 a month. They emailed a list of questions before signing up. It was quick and easy to build a few things they asked for. The part I think actually sold it was answering every question straight, including the ones where the answer was no.
Little nugget at the start that turned out to be very accurate at least with regards to data centers:
> In contrast, society did not need a bubble to pop to start excoriating AI. Given that the backlash to tech has been going on for a decade, this seems normal to us. But the AI backlash differs from the general high regard, earlier in the cycle, enjoyed by the likes of Bill Gates, Steve Jobs, Jeff Bezos, and others who built big tech businesses. The world hates change, and only gave tech a pass in the ‘80s and ‘90s because it all still seemed reversible: it could be made to go away if it turned out badly. This gave the early computer innovators some leeway to experiment. Now that everyone knows computers are here to stay, AI is not allowed the same wait-and-see attitude. It is seen as part of the ICT revolution.
One of the things that keeps surprising me about the AI build out is that there’s assumptions about usage which run directly counter to the public’s general hatred of this technology. It’s possible it’ll get more hidden and people will use it but unlike PCs or the internet it’s not that there’s an early adopter and a late adopter who just isn’t exposed yet. Instead there’s the early adopter who loves the technology and everyone else who actively dislikes the technology.
1) Your opinion is based on the political fringes in the west. Look at the AI adoption in India and China, which hold the majority of the worlds population. It looks completely different.
2) There was more demand than supply for datacenters before the AI boom. Statistical models for things like medicine, protein folding, ad-serving models, astronomy were already using more computation than was available. This is not going anywhere and was already here.
3) Every incursion by big tech into the private life of the individual happens in steps until it seems inevitable. Zhuboff writes about this in "The age of Surveillance Capitalism". The first steps are always unreasonable and triggers a backlash, but it still normalizes the technology. The amount of data collection and personal data collection we have would have seen absolute crazy to someone 20 years ago. Cameras and filming people in public was not okay 10 years ago, now its commonplace, the next step is glasses that film, that will have a backlash. Luckely for the big tech, people have a short attention span, and often forget about these things, and the next time an incursion arrives, people react less and less.
This is me. I am learning it, running it locally, trying to figure out how to adapt it to me and me to it.
For work.
For things I am interested in personally and creatively, it has almost zero use.
Not exactly zero, but very nearly.
Sometimes the fact that it answers in a completely average way has paradoxical value. When you are trying not to make or do the average thing, trying to explore something deeper, knowing that an AI can be relied upon to respond from the average perspective can be useful. Particularly if you are swimming in a very big pond.
> instead there’s the early adopter who loves the technology and everyone else who actively dislikes the technology.
I generally don't agree here. I think the narrative is fairly interesting in this regard, because AI threatened the pop culture bubble.
Early and obvious profit centers for AI usage were Art, Music, Video, Game production, News reporting, etc. Basically - Media narratives for AI were generally negative on most social media news in the US (incl hn) because the people making that content felt threatened.
But that's an echo chamber. My general experience with most in-person encounters has often gone the exact opposite direction: People often quite like what it can do for them when they try using it, and keep using it once they start.
---
So there's a lot of professional angst floating around in circles where people are worried AI is going to take their job (which I won't discount, and frankly also somewhat fear...) and that gets amplified because a lot of those most scared happen to control a lot of social media content production.
But the rest of society? They don't really care much. All I see is a staggering amount of AI generated content replacing b-roll and poster content everywhere. Forms getting made by AI, sites getting generated by AI, my kids grandpa using AI for trip planning, my mother-in-law using AI for her dance class music, teachers at my kids schools using AI for school posters and signs, restaurants using AI for menu design, kids using chatgpt for homework, folks looking up compatible car parts with ai, etc...
If the thing they're using AI for isn't the central part of their business, they're usually quite happy with the results.
---
Basically - I don't see the dislike in the general populace. Especially once you exit the US media bubble.
Datacenters aren't super popular, but frankly - I think that has more to do with folks being convinced their power bill is going to go up because of them than anything related to AI. It also helps that they don't really employ any meaningful number of people, so no one is beating the "pro-datacenter" drum all that loudly in the general public.
> Early and obvious profit centers for AI usage were Art, Music, Video, Game production, News reporting, etc. Basically - Media narratives for AI were generally negative on most social media news in the US (incl hn) because the people making that content felt threatened.
You are trivialising this: they didn't just feel threatened, they were threatened.
Lets not replace objective fact by subjective feels.
> But that's an echo chamber. My general experience with most in-person encounters has often gone the exact opposite direction: People often quite like what it can do for them when they try using it, and keep using it once they start.
Only until you ask them what value are they providing? My interactions, especially with young folk (19 - 25 age group) show negativity to AI, because they are looking forward to a lifetime of devalued skills.
Your in-person encounters are probably with owners of capital (even if it is a little bit), because they can now produce stuff without needing to employ any knowledge workers, or with people who have not yet realised that their job can be done for next to no cost by an AI that will only require an unskilled baby-sitter.
Actualy, 'ai' is trying to make me 'rich', as all my work gets sent through the web, and many of my clients reference search standing and 5star reviews, not one bit of which I in any way set up or manage, it has all coalessed ,ha!, from some cloud of processing.
The truth is that all of these people would happily click on a solution, but for the real built world, that is difficult to impossible, but the fantasy is so strong that many abandon there plans when it becomes clear that there will be a whole process involving design ,manufacture, and instalation, with them bieng required to decide what they want with no possibility
of picking from a menue,and there fumbling attempts to use 'ai' to prompt me result in
AI will completely destroy capitalism. It is going to be the most idiotic technological development ever if understood in human terms. Massive unemployment, economic crisis without a precedent in history, and of course, as consequence, war, revolutions.
Or maybe not, and 99,99% of us will have the privilege of living in a world that is a mix of Soylent Green and 1984.
Oh please. $1M is a lot of money already. It's almost $100k a month. A lot of people make that only in a year. I refuse to believe you ran a company with $29M in funding and went home broke after shutting shop.
The "cool" thing about AI is that is that if it works out as well as possible it eventually makes everyone poor by entirely breaking our economic system but in a moatless way.
1) Massive job loss
2) Reduction of need in companies of any employee size including 1 (when you don't need employees, you don't need a company, companies are just a collection of experts).
People like to talk about "1 person billion dollar companies" but if that person is just driving LLMs what value do they really bring? Even if you assume they have some unique quality of taste or very esoteric knowledge, anything they do can be easily replicated by other people copying them using LLMs causing an immediate race to the bottom.
3) Nobody to sell even the intelligence to. Capitalism implodes.
In a way, this is just a negative reframing of what one might call "the goal of technology". A piece of technology so efficient that it removes all labor needs. A utopia where all of ones needs are met and individuals can spend their times more recreationally.
Now, I would like to point out that even if AI is this good, we still haven't fully figured out robotics. So, we would only wipe out knowledge work. Thats not all work, and is the counterargument to points 2 and 3.
Right, when labor has no value, there's no point in society permitting ownership of capital. Currently, you, the worker, are trading today's labor for capital, which grows, then you trade back in for consumption some time later. But if there's no labor, what are you trading for the capital? Sure, you could imagine the people who have capital just keep accruing capital. But why would society stand for that, if the winners are not adding anything? AI puts even the captains of industry out of a job.
If this resolves peacefully, it demands UBI, and you can laugh at UBI as being a fantasy, but it actually already works in Norway, where there is a geyser of money in the ocean and everyone gets a cut and actually the society works pretty well overall. UBI doesn't work if you still need labor, but when there is infinite self sustaining capital, it can work. Or it can fail and we get war until we bomb ourselves backwards in the tech tree.
I don't know about Norway, but there is "UBI" in Alaska. Everybody gets a few hundred dollars per year from "the geyser of money in the ocean". Via Claude:
-----
YEAR DIVIDEND
2017 $1,100
2018 $1,600
2019 $1,606
2020 $992
2021 $1,114
2022 $3,284
2023 $1,312
2024 $1,702
2025 $1,000
2026 $1,200
-----
I am not sure whether getting $1,200 from Amodei will make up for me losing my six-figure job (knock on wood, I am still employed).
I know Amodei would like to be the great Altruist on top of the mountain but I don't see him replacing government. That is not a stable configuration for society, imo only.
- Ai lowers the cost of work massively
- Ai companies cannot capture enough value in the long term to make the investment worth it. Leading to a debt crisis and deflation.
They aren't my claims. I'm asking if they are unfalsifiable because if that's true then no amount of testing will ever suffice.. an unfalsifiable claim is itself a fallacy.
"When any would-be innovator can build and train an LLM on their laptop and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption."
Sophistry. AI is going to replace more and more jobs whether individuals can make their own or not. There is no reason to believe that individuals need to be able to create their own AIs for revolutionary changes to happen, and none is given by the author.
That is not ironic, it has always been the case that the elite take all the wealth and then become "benefactors" of the arts, society, etc.
Amodei, like Carnegie, Gates, Rockefeller, the Tudors, etc, does not care who he grinds into dust on the way to become the Most Altruistic Person Of All Time.
Tomorrows Anthropic IPO will create something like 4,000 millionaires just from the employee base.
Some people just believe whatever they want to believe, and people whose value to society comes from their ability to write want to believe that AI is without value.
In a decade this will all be quietly forgotten, like the people who claimed the dotcom thing was 'just a fad' after the bubble burst.
"Shipping containers didn't make anyone rich" is quite a take. Entire new categories of trade were born out of them. Sure, being the one making the containers didn't make one rich, so in that way it's different than the railroads, but it probably had a similar effect in terms of the new businesses that were made suddenly possible as a result.
I guess the result for AI is don't try to get rich inventing AI. Try to get rich using AI to enable things that weren't previously possible. I think that space has barely been explored - so far most of the activity is not net new things, but just trying to do the same thing we always did before, but faster/cheaper.
The shipping container insight is good. There was no one big winner, but many medium-sized ones. Levinson's "The Box" (2006) is the classic book on how that worked out. It's worth reading in the AI era.
"The Box" points out that shipping did not convert to containers as much as replace ports and shipping lines with new builds built for containers. The Port of San Francisco used to be a major port. But, when containers came along, a new container port was built in Oakland. One that didn't have an entrenched longshoreman's union. A few things helped. BART was under construction, and digging up dirt that had to go somewhere. That dirt, as fill, became the container port.
London went in a similar direction. London never got a container port. Instead, a minor non-union port on the eastern coast was built up into a container port. The London Docklands died, to come back decades later as office and residential space.
Containers needed all new ships, and made existing ships obsolete. Some shipping lines converted; others were superseded. There was no one big winner in containers.
It took quite a while to get trucking and railroads on board. Railroads had tried various intermodal schemes - trailer on flat car, trailers that could run on rails, and such. These took less infrastructure for loading and unloading than the big maritime boxes. But eventually the dumb boxes won out. Union Pacific even increased its clearances to allow double-stack container trains.
The US longshoremens' union got deals for lifetime support for their existing members. But there were not many new people entering the industry. If you visit a container port today (if you can get in) there won't be many people around. Here's what one looks like.[1] That's not a render; it's the port of Rotterdam, a decade ago. (The video is sped up; the AGVs are not that fast. This is real speed.[2]) All recent ports are at least that automated.
[1] https://www.youtube.com/watch?v=zm_rlLyelQo
[2] https://www.youtube.com/watch?v=SmOVUeS5Byk
It's a good comparison. Perhaps everyone here saved $500 or more (perhaps much more? Depending how you want to do the accounting) due to containerized shipping.
But also a bunch of people lost their whole job and a historically powerful union lost a lot of clout, part of a broader trend of organized labor becoming less influential in politics.
Did we manage the transition well? Were we intentional at all? Did we maximize benefits and minimize losses? It's hard to say and depends on your values.
I’m not sure that the previous system benefited laborers much — they were already exploited by organized crime, unfair hiring, corruption, etc.
There were people making a lot of money at the ports, but it wasn’t necessarily the people doing the physical labor.
> The Port of San Francisco used to be a major port. But, when containers came along, a new container port was built in Oakland. One that didn't have an entrenched longshoreman's union.
Unlike the east coast longshoremen's unions, the west coast unions saw the writing on the wall. The Port of San Francisco containerized (the first one to start, IIRC), but couldn't handle the volume, thus the shift to Oakland. Other west coast ports managed to expand and prosper. By contrast, the Port of New York and other east coast port unions strongly resisted, and ended up with nothing as new containerized ports came online.
I don't remember if I read it in The Box or Philip Dray's There is Power in a Union, but notwithstanding the early radicalism of the longshoremen's union in SF and other lore, west coast unions, at least mid century, were much more cooperative with industry than on the east coast and their members faired technology transitions fairly well. It had alot to do with smart leadership during that period, but perhaps also just the growth of the economy on the west coast.
How can you not love a website with this caliber of comments
So… don’t sell shovels but dig for gold?
I think you can find examples and counter-examples for any of these point of views, and we simply don’t know yet.
The metaphor doesn't hold, since, if gold was what was rushed for in the gold rush era, then AI is what is being rushed for in the AI era. AI models are the gold.
It anything, GPUs and data centers are the shovels. Building AI models is digging for gold. Unlike gold, AI models dont have preexisting or intrinsic value. So the money is in figuring out what to apply AI to.
Incidentally, it’s the guys that sold shovels for the gold rush that have lasting family wealth.
AI isn’t the shovel, hardware and datacenters are the shovel.
Hardware and data centres aren’t the shovel. Hydroelectric generators are the shovel.
Hydroelectric generators aren’t the shovel. The hydrologic cycle is the shovel.
The hydrologic cycle isn’t the shovel. The sun is the shovel.
Sell the sun. Can’t go wrong.
I've been looking at leaving tech for various reasons, and my dad recently told me to go into geology to do site geology for hydro dams and geothermal and other kinds of energy.
Maybe he's just trying to get me to start selling shovels.
I’m going to privatize and rent extract the sun in my sphere.
Shipping containers also just had the effect of making everyone a little bit richer. Everything that needed to get moved around the world, up and down the supply chain, got there faster and cheaper as a result.
does that really follow? I potentially get stuff faster, but is that "richer" value actually going to me, or the shipping company?
I get paid 50k a year. shipping stuff is expensive and i can only afford to buy 8k worth of stuff. that pays for 40 widgets.
I get paid 50k a year. shipping is cheaper but i can only afford to buy 8k worth of stuff. that pays for 55 widgets now. I have more crap, but my actual take home isn't more, I'm not richer, the costs of goods went down.
this is also why US wages are stagnant and why the US (and much of the world) is dependant on illegal or immigrant labor.
Every product got cheaper, labour also got cheaper. Could people afford more or less? Probably depends on how you measure. More porcelain knick-knacks and TVs. Less housing and food.
I would posit that it made both housing and food cheaper. A lot of what's required to build houses (plus the infrastructure they need) comes from overseas and would have come down in cost as a result. Unfortunately those savings don't do much in the face of an inability to actually build things, but houses would almost certainly be more expensive than they are now in a world with no shipping containers.
Same with food - even domestic food gets cheaper because a ton of potash and other stuff needed to grow it is shipped in. Also made people wealthier in the sense that it made it economical to bring in a much wider variety of foods from all over the world.
It also greatly made offshore labor cost effective.
The guy who made docker which is what people think containers are, got pretty rich
A lot of this sounded (from a quick read) interesting and insightful.
I wasn't sure about these two paragraphs:
> If AI had started a new wave, there would have been an extended period of uncertainty and experimentation. There would have been a population of early adopters experimenting with their own models. When thousands or millions of tinkerers use the tech to solve problems in entirely new ways, its uses proliferate. But because they are using models owned by the big AI companies, their ability to fully experiment is limited to what’s allowed by the incumbents, who have no desire to permit an extended challenge to the status quo.
> This doesn’t mean AI can’t start the next technological revolution. It might, if experimentation becomes cheap, distributed and permissionless—like Wozniak cobbling together computers in his garage, Ford building his first internal combustion engine in his kitchen, or Trevithick building his high-pressure steam engine as soon as James Watt’s patents expired. When any would-be innovator can build and train an LLM on their laptop and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption.
> ... In the 1970s, it took time for people to wrap their heads around the desirability of powerful and ubiquitous computing. But in 2025, machines that think better than previous machines are easy for people to understand.
Even if tinkerers aren't currently training frontier LLMs from scratch, there's still a lot of experimentation happening, and I don't think everyone even realizes yet how transformative this technology will be. On the contrary, there are a lot of skeptics.
That said, it's a good question, who captures the value.
On the question of tinkerers, I'm reminded of the man who just copy-pasted the middle layers of an LLM on top of each other: https://dnhkng.github.io/posts/rys/.
Turns out you can just do that!
The computer version of this would read something like:
> When any would-be innovator can grow and etch a silicon CPU in their kitchen and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption.
Put that way, it seems clear this whole reaction is a category error.
The author clearly sees "AI" only as training models and ignores everything built on top of models. It's a common mistake that gets made for every new technology.
>Even if tinkerers aren't currently training frontier LLMs from scratch, there's still a lot of experimentation happening, and I don't think everyone even realizes yet how transformative this technology will be.
The big issue of vendor lock-in remains for the frontier models. We saw a small piece of this in the games industry when people started reverse compiling games at unprecedented speeds, and companies' answer was to push Anthropic to refusing to process such prompts. A very easy workaround for now, but they played their hand and revealed their allegiance.
Any innovation that threatens to disrupt with current structures will be stifled by the old guard.
> I don't think everyone even realizes yet how transformative this technology will be. On the contrary, there are a lot of skeptics.
Same occurred back at the start of the cloud era where sysadmins were certain no sane business owner would up load their business to the cloud
And well...all those businesses were run by nut jobs
Whole lot of IT people are hung up on linguistics and refuse to acknowledge there's a physical component of the machine and ultimately we're managing machine states for the users need not contextualizing machine needs with layers of dev abstraction
If devs were carpenters they'd have diamond crusted hammers and debate the visual splendor all day never finishing the house
I mean, it destroys far more value than it captures.
You can use it to clone Adobe tools, emulate games, copy whatever software you want.
I assume the original idea was that the big software companies would replace all their pesky, expensive engineers. Not that we’d simply burn down their companies on accident because we had a tool to clone whatever we wanted.
> If AI had started a new wave, there would have been an extended period of uncertainty and experimentation. There would have been a population of early adopters experimenting with their own models. When thousands or millions of tinkerers use the tech to solve problems in entirely new ways, its uses proliferate
Maybe I'm reading this incorrectly, but isn't that... exactly what's happening right now? The author goes on to write other related assumptions.
I can't say I agree or disagree, but the article seems a bit disconnected from the true state of the technology in question.
Right now, I see the application of ai at a strategic level to accomplish two things:
1. To clear out the last 40 years of technical debt that wasn't dealt with because "omg we don't have time for that".
2. To search for interesting problems that have never been attacked before in a much faster way than ever before.
---
AI coding has let me launch SaaS for non classical logic reasoning ( https://arbiter.housecarl.cloud/ ), along with experimenting with game ideas, plus implementing a Lisp 1.5 with JIT compilation. And more.
I spend a lot of time coding after work with ideas. Always have. AI supercharges that.
Great piece, but I think he's missing something, the next phase of AI adoption is hiding it, because AI-made work increasingly all looks the same. If everyone cuts costs with the same models, cheap output becomes a commodity, and the real edge may go to companies whose work doesn’t look AI-made at all.
That's what I see. I run a small e-signature service. I built it mostly with AI. Building is inexpensive, but it's like that for everyone, so I can't charge for it. I price at cost plus a little and the saving goes to the customer.
Got my first paying customer today, $19 a month. They emailed a list of questions before signing up. It was quick and easy to build a few things they asked for. The part I think actually sold it was answering every question straight, including the ones where the answer was no.
Little nugget at the start that turned out to be very accurate at least with regards to data centers:
> In contrast, society did not need a bubble to pop to start excoriating AI. Given that the backlash to tech has been going on for a decade, this seems normal to us. But the AI backlash differs from the general high regard, earlier in the cycle, enjoyed by the likes of Bill Gates, Steve Jobs, Jeff Bezos, and others who built big tech businesses. The world hates change, and only gave tech a pass in the ‘80s and ‘90s because it all still seemed reversible: it could be made to go away if it turned out badly. This gave the early computer innovators some leeway to experiment. Now that everyone knows computers are here to stay, AI is not allowed the same wait-and-see attitude. It is seen as part of the ICT revolution.
One of the things that keeps surprising me about the AI build out is that there’s assumptions about usage which run directly counter to the public’s general hatred of this technology. It’s possible it’ll get more hidden and people will use it but unlike PCs or the internet it’s not that there’s an early adopter and a late adopter who just isn’t exposed yet. Instead there’s the early adopter who loves the technology and everyone else who actively dislikes the technology.
1) Your opinion is based on the political fringes in the west. Look at the AI adoption in India and China, which hold the majority of the worlds population. It looks completely different.
2) There was more demand than supply for datacenters before the AI boom. Statistical models for things like medicine, protein folding, ad-serving models, astronomy were already using more computation than was available. This is not going anywhere and was already here.
3) Every incursion by big tech into the private life of the individual happens in steps until it seems inevitable. Zhuboff writes about this in "The age of Surveillance Capitalism". The first steps are always unreasonable and triggers a backlash, but it still normalizes the technology. The amount of data collection and personal data collection we have would have seen absolute crazy to someone 20 years ago. Cameras and filming people in public was not okay 10 years ago, now its commonplace, the next step is glasses that film, that will have a backlash. Luckely for the big tech, people have a short attention span, and often forget about these things, and the next time an incursion arrives, people react less and less.
What is AI adoption like in India and China?
> Instead there’s the early adopter who loves the technology and everyone else who actively dislikes the technology.
As an early adopter I would not say I love the technology. I just keep it around because I know I have to.
This is me. I am learning it, running it locally, trying to figure out how to adapt it to me and me to it.
For work.
For things I am interested in personally and creatively, it has almost zero use.
Not exactly zero, but very nearly.
Sometimes the fact that it answers in a completely average way has paradoxical value. When you are trying not to make or do the average thing, trying to explore something deeper, knowing that an AI can be relied upon to respond from the average perspective can be useful. Particularly if you are swimming in a very big pond.
> instead there’s the early adopter who loves the technology and everyone else who actively dislikes the technology.
I generally don't agree here. I think the narrative is fairly interesting in this regard, because AI threatened the pop culture bubble.
Early and obvious profit centers for AI usage were Art, Music, Video, Game production, News reporting, etc. Basically - Media narratives for AI were generally negative on most social media news in the US (incl hn) because the people making that content felt threatened.
But that's an echo chamber. My general experience with most in-person encounters has often gone the exact opposite direction: People often quite like what it can do for them when they try using it, and keep using it once they start.
---
So there's a lot of professional angst floating around in circles where people are worried AI is going to take their job (which I won't discount, and frankly also somewhat fear...) and that gets amplified because a lot of those most scared happen to control a lot of social media content production.
But the rest of society? They don't really care much. All I see is a staggering amount of AI generated content replacing b-roll and poster content everywhere. Forms getting made by AI, sites getting generated by AI, my kids grandpa using AI for trip planning, my mother-in-law using AI for her dance class music, teachers at my kids schools using AI for school posters and signs, restaurants using AI for menu design, kids using chatgpt for homework, folks looking up compatible car parts with ai, etc...
If the thing they're using AI for isn't the central part of their business, they're usually quite happy with the results.
---
Basically - I don't see the dislike in the general populace. Especially once you exit the US media bubble.
Datacenters aren't super popular, but frankly - I think that has more to do with folks being convinced their power bill is going to go up because of them than anything related to AI. It also helps that they don't really employ any meaningful number of people, so no one is beating the "pro-datacenter" drum all that loudly in the general public.
> Early and obvious profit centers for AI usage were Art, Music, Video, Game production, News reporting, etc. Basically - Media narratives for AI were generally negative on most social media news in the US (incl hn) because the people making that content felt threatened.
You are trivialising this: they didn't just feel threatened, they were threatened.
Lets not replace objective fact by subjective feels.
> But that's an echo chamber. My general experience with most in-person encounters has often gone the exact opposite direction: People often quite like what it can do for them when they try using it, and keep using it once they start.
Only until you ask them what value are they providing? My interactions, especially with young folk (19 - 25 age group) show negativity to AI, because they are looking forward to a lifetime of devalued skills.
Your in-person encounters are probably with owners of capital (even if it is a little bit), because they can now produce stuff without needing to employ any knowledge workers, or with people who have not yet realised that their job can be done for next to no cost by an AI that will only require an unskilled baby-sitter.
That feels more like assigning value to loud people on the Internet then engaging with the evidence.
Actualy, 'ai' is trying to make me 'rich', as all my work gets sent through the web, and many of my clients reference search standing and 5star reviews, not one bit of which I in any way set up or manage, it has all coalessed ,ha!, from some cloud of processing. The truth is that all of these people would happily click on a solution, but for the real built world, that is difficult to impossible, but the fantasy is so strong that many abandon there plans when it becomes clear that there will be a whole process involving design ,manufacture, and instalation, with them bieng required to decide what they want with no possibility of picking from a menue,and there fumbling attempts to use 'ai' to prompt me result in
Previous discussion a year ago:
Will AI be the basis of many future industrial fortunes, or a net loser?
https://news.ycombinator.com/item?id=45235676
AI will completely destroy capitalism. It is going to be the most idiotic technological development ever if understood in human terms. Massive unemployment, economic crisis without a precedent in history, and of course, as consequence, war, revolutions.
Or maybe not, and 99,99% of us will have the privilege of living in a world that is a mix of Soylent Green and 1984.
Tell that to the note-taking app that got $29M and then closed shop.
https://get.mem.ai/blog/mem-is-shutting-down
29m is just not a lot of money in the tech world these days. There are so many AI companies with 1b+ valuations that nobody heard of.
Oh please. $1M is a lot of money already. It's almost $100k a month. A lot of people make that only in a year. I refuse to believe you ran a company with $29M in funding and went home broke after shutting shop.
Who got rich here?
Take a wild guess. It's not the investors for sure.
Shipping containers do not recursively improve themselves
That's true. They're in great company.
If it was so good I should be able to give it 16bx and say make me rich. It is a new turing test. Ok maybe it should be 255bx.
The "cool" thing about AI is that is that if it works out as well as possible it eventually makes everyone poor by entirely breaking our economic system but in a moatless way.
1) Massive job loss
2) Reduction of need in companies of any employee size including 1 (when you don't need employees, you don't need a company, companies are just a collection of experts).
People like to talk about "1 person billion dollar companies" but if that person is just driving LLMs what value do they really bring? Even if you assume they have some unique quality of taste or very esoteric knowledge, anything they do can be easily replicated by other people copying them using LLMs causing an immediate race to the bottom.
3) Nobody to sell even the intelligence to. Capitalism implodes.
In a way, this is just a negative reframing of what one might call "the goal of technology". A piece of technology so efficient that it removes all labor needs. A utopia where all of ones needs are met and individuals can spend their times more recreationally.
Now, I would like to point out that even if AI is this good, we still haven't fully figured out robotics. So, we would only wipe out knowledge work. Thats not all work, and is the counterargument to points 2 and 3.
Right, when labor has no value, there's no point in society permitting ownership of capital. Currently, you, the worker, are trading today's labor for capital, which grows, then you trade back in for consumption some time later. But if there's no labor, what are you trading for the capital? Sure, you could imagine the people who have capital just keep accruing capital. But why would society stand for that, if the winners are not adding anything? AI puts even the captains of industry out of a job.
If this resolves peacefully, it demands UBI, and you can laugh at UBI as being a fantasy, but it actually already works in Norway, where there is a geyser of money in the ocean and everyone gets a cut and actually the society works pretty well overall. UBI doesn't work if you still need labor, but when there is infinite self sustaining capital, it can work. Or it can fail and we get war until we bomb ourselves backwards in the tech tree.
I don't know about Norway, but there is "UBI" in Alaska. Everybody gets a few hundred dollars per year from "the geyser of money in the ocean". Via Claude:
-----
YEAR DIVIDEND
2017 $1,100
2018 $1,600
2019 $1,606
2020 $992
2021 $1,114
2022 $3,284
2023 $1,312
2024 $1,702
2025 $1,000
2026 $1,200
-----
I am not sure whether getting $1,200 from Amodei will make up for me losing my six-figure job (knock on wood, I am still employed).
I know Amodei would like to be the great Altruist on top of the mountain but I don't see him replacing government. That is not a stable configuration for society, imo only.
Yup. Fully in this camp. I don’t see how LLMs are anything other than deflationary.
I'm already feeling this way at work. Why do anything if next year's model will do it 10x better and faster?
Because it has to be done this year?
I work at a business that is not facing any existential crises at the moment. What do you work on that has to be done yesterday?
Why do you work at a business if it's better to just shut it down and resume next year?
It’s not just LLMs, it’s also the bubble.
One or both of these is happening:
- Ai lowers the cost of work massively - Ai companies cannot capture enough value in the long term to make the investment worth it. Leading to a debt crisis and deflation.
are these unfalsifiable claims?
Only one way to know, test them and see if they break. What scenario do you propose testing the claims against?
They aren't my claims. I'm asking if they are unfalsifiable because if that's true then no amount of testing will ever suffice.. an unfalsifiable claim is itself a fallacy.
"When any would-be innovator can build and train an LLM on their laptop and put it to use in any way their imagination dictates, it might be the seed of the next big set of changes—something revolutionary rather than evolutionary. But until and unless that happens, there can be no irruption."
Sophistry. AI is going to replace more and more jobs whether individuals can make their own or not. There is no reason to believe that individuals need to be able to create their own AIs for revolutionary changes to happen, and none is given by the author.
[flagged]
[dead]
[dead]
Ironic if Anthropic, which is a public benefit corp, ends up being the firm that cashes in the most on AI.
That is not ironic, it has always been the case that the elite take all the wealth and then become "benefactors" of the arts, society, etc.
Amodei, like Carnegie, Gates, Rockefeller, the Tudors, etc, does not care who he grinds into dust on the way to become the Most Altruistic Person Of All Time.
Among overly confident takes rife with total failure to reason with appropriate caution about non-linear tightly-coupled systems,
this is sure a remarkably pure example.
This is right up there with "humans will never achieve flight" and "humans will die of asphyxiation at 100 mph" or whatever.
Tomorrows Anthropic IPO will create something like 4,000 millionaires just from the employee base.
Some people just believe whatever they want to believe, and people whose value to society comes from their ability to write want to believe that AI is without value.
In a decade this will all be quietly forgotten, like the people who claimed the dotcom thing was 'just a fad' after the bubble burst.
A millionaire isn't what it used to be. When you see "millionaire" read "house and pension owner". Something we used to assume we'd all have.