3 ms·

Interesting how this argument is only popping now that technology is threatening white collar workers. Automation has been shoving blue collars out of the job

Interesting how this argument is only popping now that technology is threatening white collar workers. Automation has been shoving blue collars out of the job market for a century. A single farmer can do with his machinery today what took a dozen of people just 50 years ago. Manufacturing has been super automated long ago. Even in commerce automated checkout has been replacing workers for more than a decade. In any case such a tax is not only pointless but actively dangerous, as all it achieves is making countries without such a tax more competitive.

cubefox 11mo ago
It's dangerous but not pointless. If we get human mass unemployment, someone has to pay for them, and if the companies offering the AI sit in the US, most countries won't be able tax OpenAI & Co directly.
- HPsquared 11mo ago
  If anything, US AI companies will be eating the lunches of workers from all over the world. It'll be like a service export.
  - seg_lol 11mo ago
    Highly automated and connected business infrastructure will make replacing first world jobs the easiest, but yes as automation spreads, the spigot of job replacement will flow to every corner of the globe.
    - cubefox 11mo ago
      That's not the problem. The problem is that the revenue will flow to the US, and when those US companies are taxed, this will at most pay for an UBI for US Americans only. Which will mean the US gets very rich while the other countries get very poor, because there is no large AI company to tax.
  - epolanski 11mo ago
    I'm not sure about that to be honest. 1. Financial perspective. People are too focused on what "the best" is, rather than what is the most financially viable at scale, which is what really matters. At the peak of the Ethereum mining craze in 2016/2017 the GTX Titan X was the best performing GPU. But buying 200$ AMD Polaris GPUs was what gave you the most performance per $ and per watt. 2. Open source models keep being impressive and lagging only so much behind the closed source ones. It's hard to predict the future, but few years from now the most viable application might be to internally fine tune and deploy on whatever cloud or internal infra open source models. I have already many use cases in prod where Gemini Flash 2.0 did a great job, and that's an old model by today's standards (summarizing news/translation). Now I have in production a service that reviews pull requests and updates documentation/JIRA accordingly when they are merged. That requires quite more plumbing, agentic approach and thinking, but yet again open source models can do a terrific job already there. 3. At the end of the day, the lion share is going to be eaten by whoever provides the best applications, not models, but conversely we're also living in a space where more and more you can just build roughly-the-same-feature with few $ worth of APIs. 4. Even more, the biggest benefit will lie among those who will leverage AI in the best way. Companies and individuals able to really delegate successfully complex tasks making crazy savings. Who knows who's really gonna take the biggest advantage. Maybe US companies, maybe not. Thus, in essence, I envy your certainties around the future, I personally have lots and lots of doubts and have no clue who's gonna eat whoever's lunch.
    - cubefox 11mo ago
      > 1. Financial perspective. People are too focused on what "the best" is, rather than what is the most financially viable at scale, which is what really matters. I don't think so: People flocked to ChatGPT because it was the best, even though there are far cheaper options. If you are 10% better than any competitor, you don't get just 10% more market share, you get far more. It's a winner-takes-most situation.
      - epolanski 11mo ago
        That's a money loser, I'm talking about who's gonna be able to monetize and leverage this.
mlrtime 11mo ago
What makes you think this argument is popping? Or that it's only being discussed now, because of 1 article on a domain I've never heard of?
- msopena 11mo ago
  ElPais is the most important (by number of readers) Spanish newspaper.
veunes 11mo ago
I think the current debate is less about whether automation happens and more about where the gains go
- Ajedi32 11mo ago
  The same place they've always gone: to consumers purchasing the products made by those automated processes. The average person now is far wealthier in terms of actual purchasing power than the average person 100 years ago, and that's largely because of automation making everything cheaper.
  - Arkhaine_kupo 11mo ago
    > The average person now is far wealthier in terms of actual purchasing power than the average person 100 years ago access to housing is incredibly expensive. Measuring their purchasing power for how many scented candles they can buy is pretty meaningless when they are much closer to homelessness than ever before.
    - Ajedi32 11mo ago
      Land is one thing you can't automate the production of, and construction still consists mostly of manual skilled labor. But I think despite that you'll find housing today still compares very favorably to housing 100 years ago (in terms of median square footage, safety, amenities, etc).
    - jlawson 11mo ago
      This is actually _because_ automation has been so effective. It's called Baumol's cost disease. https://en.wikipedia.org/wiki/Baumol_effect https://en.wikipedia.org/wiki/Baumol_effect
  - int_19h 11mo ago
    > The same place they've always gone: to consumers purchasing the products made by those automated processes. The people who are "automated away" and have reduced income are also less of a consumer. In a society where most jobs are automated, who are the consumers?
ozmodiar 11mo ago
I'm open to the idea that companies need much larger barriers to functioning internationally. Not just because of their ability to pump money overseas, but also because they are often used to blatantly further the security goals of their parent countries (looking at you two USA and China, but everyone able to, seems to do it). Maybe the current system would've worked if it was built on many more small companies. These monolithic corporations funneling power upward are the death of civilization, and leadership are clearly high on their own farts. Or just want to be on top in a new feudal age.
- zanellato19 11mo ago
  This also makes labor laws much less effective because the companies move countries to avoid them.
estearum 11mo ago
This has been a question since at least the industrial revolution. Perhaps the more interesting bit is that you only seemed to have noticed it when it is asked about white collar workers?
shagie 11mo ago
It's been around for nearly a decade. Taxing Robots : Easier Said Than Done (2017) https://www.ctf.ca/EN/EN/Newsletters/Canadian_Tax_Focus/2017/2/170204.aspx https://www.ctf.ca/EN/EN/Newsletters/Canadian_Tax_Focus/2017... Robots, technological change and taxation (2017) https://www.taxjournal.com/articles/robots-technological-change-and-taxation-14092017 https://www.taxjournal.com/articles/robots-technological-cha... Why robots should be taxed if they take people's jobs (2017) https://www.theguardian.com/business/2017/mar/22/robots-tax-bill-gates-income-inequality https://www.theguardian.com/business/2017/mar/22/robots-tax-...

cubefox 11mo ago

It's dangerous but not pointless. If we get human mass unemployment, someone has to pay for them, and if the companies offering the AI sit in the US, most countries won't be able tax OpenAI & Co directly.

HPsquared 11mo ago
If anything, US AI companies will be eating the lunches of workers from all over the world. It'll be like a service export.
- seg_lol 11mo ago
  Highly automated and connected business infrastructure will make replacing first world jobs the easiest, but yes as automation spreads, the spigot of job replacement will flow to every corner of the globe.
  - cubefox 11mo ago
    That's not the problem. The problem is that the revenue will flow to the US, and when those US companies are taxed, this will at most pay for an UBI for US Americans only. Which will mean the US gets very rich while the other countries get very poor, because there is no large AI company to tax.
- epolanski 11mo ago
  I'm not sure about that to be honest. 1. Financial perspective. People are too focused on what "the best" is, rather than what is the most financially viable at scale, which is what really matters. At the peak of the Ethereum mining craze in 2016/2017 the GTX Titan X was the best performing GPU. But buying 200$ AMD Polaris GPUs was what gave you the most performance per $ and per watt. 2. Open source models keep being impressive and lagging only so much behind the closed source ones. It's hard to predict the future, but few years from now the most viable application might be to internally fine tune and deploy on whatever cloud or internal infra open source models. I have already many use cases in prod where Gemini Flash 2.0 did a great job, and that's an old model by today's standards (summarizing news/translation). Now I have in production a service that reviews pull requests and updates documentation/JIRA accordingly when they are merged. That requires quite more plumbing, agentic approach and thinking, but yet again open source models can do a terrific job already there. 3. At the end of the day, the lion share is going to be eaten by whoever provides the best applications, not models, but conversely we're also living in a space where more and more you can just build roughly-the-same-feature with few $ worth of APIs. 4. Even more, the biggest benefit will lie among those who will leverage AI in the best way. Companies and individuals able to really delegate successfully complex tasks making crazy savings. Who knows who's really gonna take the biggest advantage. Maybe US companies, maybe not. Thus, in essence, I envy your certainties around the future, I personally have lots and lots of doubts and have no clue who's gonna eat whoever's lunch.
  - cubefox 11mo ago
    > 1. Financial perspective. People are too focused on what "the best" is, rather than what is the most financially viable at scale, which is what really matters. I don't think so: People flocked to ChatGPT because it was the best, even though there are far cheaper options. If you are 10% better than any competitor, you don't get just 10% more market share, you get far more. It's a winner-takes-most situation.
    - epolanski 11mo ago
      That's a money loser, I'm talking about who's gonna be able to monetize and leverage this.

mlrtime 11mo ago

What makes you think this argument is popping? Or that it's only being discussed now, because of 1 article on a domain I've never heard of?

msopena 11mo ago
ElPais is the most important (by number of readers) Spanish newspaper.

veunes 11mo ago

I think the current debate is less about whether automation happens and more about where the gains go

Ajedi32 11mo ago
The same place they've always gone: to consumers purchasing the products made by those automated processes. The average person now is far wealthier in terms of actual purchasing power than the average person 100 years ago, and that's largely because of automation making everything cheaper.
- Arkhaine_kupo 11mo ago
  > The average person now is far wealthier in terms of actual purchasing power than the average person 100 years ago access to housing is incredibly expensive. Measuring their purchasing power for how many scented candles they can buy is pretty meaningless when they are much closer to homelessness than ever before.
  - Ajedi32 11mo ago
    Land is one thing you can't automate the production of, and construction still consists mostly of manual skilled labor. But I think despite that you'll find housing today still compares very favorably to housing 100 years ago (in terms of median square footage, safety, amenities, etc).
  - jlawson 11mo ago
    This is actually _because_ automation has been so effective. It's called Baumol's cost disease. https://en.wikipedia.org/wiki/Baumol_effect https://en.wikipedia.org/wiki/Baumol_effect
- int_19h 11mo ago
  > The same place they've always gone: to consumers purchasing the products made by those automated processes. The people who are "automated away" and have reduced income are also less of a consumer. In a society where most jobs are automated, who are the consumers?

ozmodiar 11mo ago

I'm open to the idea that companies need much larger barriers to functioning internationally. Not just because of their ability to pump money overseas, but also because they are often used to blatantly further the security goals of their parent countries (looking at you two USA and China, but everyone able to, seems to do it). Maybe the current system would've worked if it was built on many more small companies. These monolithic corporations funneling power upward are the death of civilization, and leadership are clearly high on their own farts. Or just want to be on top in a new feudal age.

zanellato19 11mo ago
This also makes labor laws much less effective because the companies move countries to avoid them.

estearum 11mo ago

This has been a question since at least the industrial revolution. Perhaps the more interesting bit is that you only seemed to have noticed it when it is asked about white collar workers?

shagie 11mo ago

It's been around for nearly a decade. Taxing Robots : Easier Said Than Done (2017) https://www.ctf.ca/EN/EN/Newsletters/Canadian_Tax_Focus/2017/2/170204.aspx https://www.ctf.ca/EN/EN/Newsletters/Canadian_Tax_Focus/2017... Robots, technological change and taxation (2017) https://www.taxjournal.com/articles/robots-technological-change-and-taxation-14092017 https://www.taxjournal.com/articles/robots-technological-cha... Why robots should be taxed if they take people's jobs (2017) https://www.theguardian.com/business/2017/mar/22/robots-tax-bill-gates-income-inequality https://www.theguardian.com/business/2017/mar/22/robots-tax-...