It feels really alien to discuss this in terms of "taxing AI", like an economic abstraction completely breaking down. Ultimately when you take automation to its logical conclusion we have people with needs and we have machines and automation capable of meeting those needs with minimal human labor.
No matter how you try to resolve this economically, it should hold that if something can be produced with minimal human labor, it shouldn't require substantial human labor to buy (in "reasonable" quantities, however you want to define and enforce that).
Without understanding the "end game" of automation (decades+ from now) it feels like we're just sleepwalking into an absurd reality where a few trillionaires own the world's fully automated food supply chain, but buying food somehow requires just as much labor as it does today.
You're completely omitting externalised cost, though. As it stands, all this production requires gargantuan amounts of energy that have to come from somewhere, and cause pollution and waste that must be accounted for. As long as these factors aren't solved—if they can be solved in the first place—either the prices for consumers or the manufacturing cost must reflect this, I don't see the increased degree automation affecting prices much.
That has nothing to do with this, those are things that should still be solved at a much higher level of abstraction. Tax the energy, pollution, waste - those have problems regardless of what caused them.
The point I am making is that the reason artefacts of highly automated production (even with minimal human labor required) will never become accessible for very low human labor, because all that automation has its own cost. We can externalise that as long as possible and defer the bill to somewhere or someone else, but it will have to be paid eventually.
> […] those are things that should still be solved at a much higher level of abstraction […]
I don't think that makes much sense. If a data center consumes all available electricity in a given municipality, it may provide AI services at a very low cost, but thereby makes the region uninhabitable. There is no way to "solve" this at a higher abstraction level. Or alternatively, consider a factory producing consumer goods, which emits toxic fumes; we can limit the amount of fumes the vicinity of the factory is exposed to by implementing very expensive filters—thus increasing the final price of the goods—or externalise all the negative effects—such as health risks in the population, ecological demise, and subsequently lower property values—to society, achieving a lower final price.
Currently, we often pick the latter option, because it usually has the better profit margin. I agree that it's a systemic issue that must be addressed holistically, but the actual solutions have to be implemented at all levels of the production chain. And this means the cost attached will have to be included in the price of all goods.
> The point I am making is that the reason artefacts of highly automated production (even with minimal human labor required) will never become accessible for very low human labor, because all that automation has its own cost
While I'm not sure I agree, this is not solved by tackling things at a low level and should be done at a higher level of abstraction - that's what they were saying.
> don't think that makes much sense. If a data center consumes all available electricity in a given municipality, it may provide AI services at a very low cost, but thereby makes the region uninhabitable.
If the data center was providing streaming services would you want to manage that differently? Imagine you had a data center that solved some user problem X, and another one that solves the same problem. Data center A uses AI, B does not but uses more power. Would you want to tax B less? Given what you've said so far I'd assume the answer is no - you'd want to tax that more because it's not really the AI part you care about, it's the power usage/emissions/local impact/externality X you want to avoid.
> I agree that it's a systemic issue that must be addressed holistically, but the actual solutions have to be implemented at all levels of the production chain.
Actually the more abstract sometimes the fewer places you have to deal with it. You don't have to figure out what cars everyone has, the specific MPG of each, driving patterns, how far your delivery driver went, whether they had other packages, etc - you can tax gasoline. This automatically flows through and avoids lots of wrangling about details and loopholes.
> Currently, we often pick the latter option, because it usually has the better profit margin.
Yes - and this drive makes it hard to manage when you put very precise rules around it. Tax AI and watch things rebrand as whatever falls just outside the limits of AI. See how products are built, deconstructed and remade exactly based on specific tariffs. Ford used to ship vans with windows and seats installed, then take them out again after they arrived!
How does more pollution/waste equal higher consumer cost? Do you mean because we'll have to pay more taxes because we'll need more publicly funded resources to clean up the excess waste? Or because corporations would pass the price of the fines for violating environmental regulations onto consumers?
It doesn't matter much how exactly those costs are passed on; someone has to eventually pay for them. That includes the energy itself, which doesn't come for free, but also the bill for environmental damage and resource exhaustion that we will have to pay at some point. You can argue that that'll be the case only in the far future, but then you're just externalising the cost—again—to future generations. It's all moot: Someone will pay for it eventually.
I'm not omitting anything, it's included. If the total sum of human labor required to produce food (which includes energy production and pollution mitigation) goes down by 90%, then the total sum of human labor required to buy food should go down by 90% as well.
My point is simply this: As automation advances, the vast majority of that value (e.g. 98%) should flow back to society. Right now it's being soaked up by capitalists. There should be a reward for improving efficiency but it should be sensible.
I think the idea is that the trillionaires won't need us at all when the food supply is fully automated. They might keep a small population for genetic diversity, but that's about it.
I see a natural equilibrium with a tension: automation (also through AI) causes unit economics to drop and results in cheaper prices. At the same time, salaries for contributors grow because their impact is so high. So you end up with a new equilibrium of much cheaper prices and much higher salaries. What, however, about the people who can’t contribute? IMO the most natural and fair approach is to support (through whatever means) people’s “education”, allowing them to upgrade their skills so that they can contribute. IMO this leads to a new tension: not rich vs poor, or useful vs useless, but people who can up-level their skills vs those who don’t. And I think, at its extreme, it boils down to this: how much plasticity does your brain have? Because every other constraint, society can adapt or accommodate for.
That is a game thoery approach but it completely fails in the face of reality.
The reality is that the floor to become "useful" is relatively low, which means the few billioanires have a large pool of potentially useful people of which they only employ some, leading to no greater salaries due to labour competition.
The other potentially useful workers cannot pool together and compete as the barrier of entry in the sector is prohibitely high.
So a natural moat emerges over cost of setting up a company, workers beg for a job of which they will take for a small wage and a few billioanires control the market.
This is a much closer approximation to the market we currently see
Yeah, that definitely won't work at scale. The bar for what constitutes being "educated" keeps increasing. Previously it was knowing how to code, now it is having an ML PhD, for example. At the same time, AI keeps getting more and more capable, so no matter how much "education" you have, AI will eventually get to you.
In any case, the argument won't work for majority of the population without a college degree. Are you going to have 50+ year old truck drivers upskilling in a fancy new tool to keep a job? And again, how long until that new skill you upgraded them to is now done by AI as well.
There has been extensive debate around that topic since that paper came out. Some points to discuss:
1. Even the article you shared mentions that starting in 2003, earnings has stopped tracking productivity. "Total compensation remains close until 2003, but does not follow 2003’s uptick in productivity growth (behavior which remains a topic for future research)."
2. They use average earnings and not median earnings. Average earnings include people like CEOs. This by consequence shows that inequality among workers has also increased. Check out chart 4 here to see how much smaller median wages are compared to average: (https://www.csls.ca/ipm/23/IPM-23-Mishel-Gee.pdf)
3. Apart from the average vs median difference, the biggest point of contention between that study and more recent ones is the measure of inflation used. The 2007 study you cite uses a measure of inflation that also includes things paid by employers like medical insurance. It turns out that using that one leads to significantly lower inflation. If you use consumer price index, what workers actually pay out of pocket, the difference again becomes larger. Citing page 37 of the study above: "In other words, that the prices of consumer items has risen faster than a broader index of prices that includes net exports, government goods and services, and investment goods. Therefore, for a given increase in income, the purchasing power of the consumer has fallen faster than that of business for investment goods and foreigners for U.S. exports."
The article I shared before plus this other one describe all the discrepancies (https://www.epi.org/productivity-pay-gap/). Specially see chart 10 in the PDF study. That shows all possible variations of how you measure productivity and income. No matter how you look at it, the most substantiated conclusion is that income has NOT matched productivity.
Transition periods are always difficult, but they've always reached some equilibrium. Right now the spread between the two seem higher (not sure they are) but the system will bring them closer together.
No matter how you try to resolve this economically, it should hold that if something can be produced with minimal human labor, it shouldn't require substantial human labor to buy (in "reasonable" quantities, however you want to define and enforce that).
Without understanding the "end game" of automation (decades+ from now) it feels like we're just sleepwalking into an absurd reality where a few trillionaires own the world's fully automated food supply chain, but buying food somehow requires just as much labor as it does today.