AI used well can restore middle class jobs [pdf]
nber.org
nber.org
It's not quite "expertise" that justifies the wage gap, it's supply and demand. Almost anyone can be a crossing guard--not everyone can be an air traffic controller. If AI theoretically could make anyone an air traffic controller, one would expect the salary to collapse as well.
Additionally, the notion of a middle class relies on a wage differential. If AI levels the playing field so dramatically, the notion of middle class will entirely evaporate since everyone's purchasing power equalizes.
>The contemporary challenge is the high and rising price of essential services like healthcare, higher education and law, that are monopolized by guilds of highly educated experts.
The rising cost of education and healthcare are not exactly going to the "guilds of highly educated experts" as much as they are ballooned by second-order effects from an overly complicated system. For example, the higher cost of education is not going to professors as much as it's going to the administrative costs as colleges continue to compete for a growing supply of students willing to pay (in part because the mass availability of student loans allows them an extensive line of credit and also because the labor market makes it seem like they must get a degree to be competitive).
We can study this by looking at examples!
Slovenia (where I'm from) has the 3rd lowest gini index in the world[1]. Meaning we are the 3rd most egalitarian country in the world with close to zero income inequality.
You would think this is fantastic, but in practice it means everyone's equally poor-ish. There's a lot of talent drain. Everyone in my millenial generation that's great at their job has either moved out of the country, works remotely for foreign companies, or has built a business that primarily targets foreign markets. They're starting to build a proper middle class, but there's nowhere for their money to go.
There aren't enough higher middle class folk to support all the services they'd want. So the extra money goes to foreign companies or into inflated real estate that has become unaffordable to anyone employed locally.
And there are no rich people to tax or ask for philanthropy to break the status quo. A net worth of ~$20mil puts you on the list of top 100 richest Slovenians as of 2023. Most of them made their money abroad and have also moved their wealth out of the country to avoid high taxation.
The concept of "upwards mobility" is barely worth talking about.
[1] https://en.wikipedia.org/wiki/List_of_countries_by_income_eq...
Whether the government wants to tax it is a different question. Usually they do not.
We're talking about a formerly socialist European country here. Not everywhere is USA :)
Income tax in Slovenia goes up to 50% and VAT (similar to sales tax) is an eyebleeding 22%. Capital gains taxes can be as high as 25%. Slovenia is not afraid to tax people don't worry.
Otherwise, there will be an asymmetry in which the "brains and capital" will get all the benefits of taxes with little to no skin in the game. That implies society serves the economy and not the other way around.
People like my mom who bought their place 20 years ago. Or like my grandparents who bought their house 50 years ago.
But there's not enough new investment to increase supply. In large part because there aren't enough people with liquid wealth to provide said supply.
And none of the people who bought decades ago are selling because they need somewhere to live and can't afford to buy at current prices. Any gains they've made would be immediately eaten away by the new purchase.
You couldn't do something like a wealth tax on these people because they have very little cashflow to pay those taxes with. And it's not like they aren't using their homes, so an under-use tax wouldn't work either. You could try a land value tax, but again there's no cashflow to collect that from and nowhere for these people to go if they're forced to sell.
- “prompt engineer - accounting” who has vast history of accounting practices
- “prompt engineer - lawyer”. No more farming out writing up complicated EULAs to Skadden. Have your general counsel do all of this for you with the help of “AI”. Or maybe have an IPO prepared without the help of a middleman such as “JPM” or “Goldman Sachs”. Billable hours drop significantly.
- “prompt engineer - swe”. Instead of multiple teams of engineers. Have “Devin” scaffold out the basic application while you focus on architecting an end to end solution.
All of those “middle class jobs” once held by mid tier specialists have evaporated.
Then the code review will still require a human senior developer and maybe the architect. The testing after that can be fuzzed by AI but will still need to be confirmed by a human senior QA specialist.
In other words, AI can lower the bar for entry level work (not by much), but it does not eliminate or create any new jobs. It's very similar to what search did for entry level work. Search results "powered by AI" are also over a decade old.
AI is very good at doing things that can be unit tested, or doing small implementation details.
It falls flat on its face when you ask it to understand architectural patterns, or even reference other classes/modules in a broader codebase.
The hard part about coding is rarely the code, more often it is understanding the human side of things, and how you can fulfill real world needs of users.
So if AI enables every unskilled worker to produce $132k of value instead of $33k, who gets the $99k surplus? Marxist economics teaches us that in a capitalist society, without additional state intervention, the employer gets all of it. Too bad for the worker.
The good news, though, is that most labour economists wouldn't go as far as Marx. In modern societies the working man seems to get at least some benefit from productivity gains that don't come directly from him working harder. And even if you do believe Marx, note the caveat about "without state intervention". The modern state has many tools to intervene and is not afraid to use them: taxes, minimum wage laws, mandating bullshit jobs. In this scenario, doubling minimum wage wouldn't hurt economic productivity - the workers are all producing $132k of value, so not a single one will be laid off if they need to be paid $66k each.
Of course, there are some coordination problems to solve...
That may have been the case 30, 40 years ago in the US. In reality, the minimum wage in almost the entire US (unchanged since 2009, to add!) is not enough to afford to have a family [1]. Companies and the ultra rich aren't taxed shit - when you have people like Warren Buffett calling to tax him and his ilk more [2], it's obvious that the situation is clearly out of control.
If you raised a minimum wage to that level, you would eliminate a lot of low-level jobs completely, and leave people who don't aspire or need this non-trivial level of achievement without jobs.
Society needs children to survive - someone has to pay the pensions after all, at least in common "societal contract" systems, otherwise people will have to literally work until they die.
For that in turn, society needs at least a replacement fertility rate, and usually there is a close negative correlation between wealth and amount of children (i.e. the wealthier people get, the less children they have), so it makes sense to strengthen the ability particularly of the lower classes to have children.
This is true. However, what society definetly doesn't need is
> the ability particularly of the lower classes to have children
That's how you breed poverty and other systemic problems that you later have to deal with. People should only have children when they can afford to invest into real care for them. If a person earns a low wage, he shouldn't be able to afford children because he shouldn't have them.
The right to found a family is enshrined in Article 12 of the European Convention on Human Rights [1] and Article 16 of the Universal Declaration of Human Rights [2].
I agree it's a bit of a stretch, but at least the UDHR definition explicitly states that "the family is the natural and fundamental group unit of society and is entitled to protection by society and the State"... for me, that clearly implies economic protection, i.e. a society has to make sure that people can actually exercise that right - because how useful is a theoretical right to something (be it to found a family, or to have the right to due process and fair treatment before the courts, or whatever else), when large swaths of the population cannot exercise that right, or have access to it effectively denied by circumstances outside of their but inside the government's control?
[1] https://www.echr.coe.int/documents/d/echr/convention_ENG
[2] https://www.un.org/en/about-us/universal-declaration-of-huma...
This is an important distinction that the author misses. People are paid according to their value to the economy, not their value to society. So, even if the jobs were similar in expertise, the crossing guard would likely not make as much because the ATC provides more to the economy.
*FWIW, I'm not saying this is moral or fair, it's just is the way the economic cookie crumbles.
E.g., a hedge fund manager is paid handsomely for if they extract value, but that same extraction is value generation for their clients. It feels like two sides of the same zero sum coin.
PhDs provide an immense amount of value but have very little leverage.
Workers who unionize and use their power effectively have more leverage than those who do not.
Land owners contribute precisely 0 value but hold leverage in the form of a legal title.
(I would also push back a bit that a PhD adds immense value. Many that do get paid well via IP and start-ups, but that’s a better interface with the economy. There are many in my experience that add little more than a credential)
For employment the most it does is set a ceiling on wages. Pushing up your ceiling doesn't necessarily push up your wages.
>I would also push back a bit that a PhD adds immense value.
Of course, that's a common opinion. Blue sky research is vastly underrated in the west. This is partly why the Chinese are overtaking us.
I’m not discounting fundamental research in the way you may be implying. I’m just saying it has little economic value until it has an application. By definition, it doesn't have an economically marketable use, so it can't have economic value.
I also don’t think China is overtaking the US on a research front. They publish a lot (although I would question how much 'blue sky' vs. derivative), but that’s a bad metric for a lot of reasons.
That's a bit like saying that employees create very little value until a product is sold. Realized value != created value.
It is precisely the lack of competition on this front (research) for decades and the abstract nature of the value created that has led it to be so badly undervalued.
If you could quantify fundamental research's value to the economy (without hindsight), it ceases to be fundamental research. Again, by definition, fundamental research does not have application. If it has no application, it has no interface to the economy. I'm not saying R&D or fundamental research has no potential economic value. It has no current avenue for realized economic value. And our system is incentivized for short-term gain. You could be a magnificent fundamental researcher, but if your work doesn't add to the economy in a relatively short timeframe, you probably won't be paid very well compared to a lesser researcher who does.
This is exactly my point. This ^ is categorically false.
>If you could quantify fundamental research's value to the economy
It seems like you're committing the https://en.wikipedia.org/wiki/McNamara_fallacy
I think you are misunderstanding the fallacy you linked. I've already said multiple times there are multiple dimensions to value. The economy is specifically about goods and services sold, so yes, we should measure production. It says nothing of other dimensions. You can have a job that is of immense social value (e.g., clergy or dog fostering or whatever) while only having marginal economic value. So what other non-productive economic dimensions do you suggest? I suspect anything you name would go into a different category than economics (e.g., quality of life). That's fundamentally misunderstanding my point.
You seem to be saying "See that idle factory over there!? Look at everything it's contributing to the economy!" and I'm saying "It's not contributing anything because it's idle." And there may still be other dimensions to its value. Maybe the architecture has artistic value, or it's history has cultural value. But none of those domains move the needle at all for economic value. I have been very clear in constraining my argument to economic value for a good reason, and it's not about your linked fallacious argument.
Since this discussion concerns wages, my claim was that economic value addition is the most highly correlated with pay. You may say we should compensate people more for what they bring to society rather than just what they bring to the economy, and I would probably agree. But that's not the system we're working under and that is my point.
I'm saying that factory has value - sometimes quite a lot of economic value.
The specific example you chose is actually interesting because in the west we largely do view idled factories as valueless and have shifted to a just-in-time manufacturing philosophy whereas Russia has taken the opposite approach of building in a lot of slack into their industrial capacity so that they can scale up quickly.
The practical upshot of this is that our differing conceptions of value are a large part of why we are no losing the largest land war in Europe since WW2. We simply don't have the capacity to scale production to necessary levels.
This is an aside, though.
>Since this discussion concerns wages, my claim was that economic value addition is the most highly correlated with pay.
Yes, this is still false. It correlates with leverage, of which your ability to give the impression of economic value addition is just one part.
This is exactly what I was cautioning against. Slack in a system is not economically productive. It is risk mitigation. You're conflating economically productive value with risk-reduction value. For the same reason, insurance is considered a liability, not an asset. That doesn't mean adding extra capacity isn't warranted from a risk perspective, but it doesn't mean you can add that excess capacity to the plus side of your ledger. It really seems like we're talking past each other because you aren't able to parse the different types of value I'm talking about.
>It correlates with leverage
I've already conceded this point. I used economic leverage because it is the most generalizable method to get leverage. Sure, you could instead blackmail the boss or date the boss's daughter to get leverage. But I don't think those hold as a general rule for a good strategy so we're mainly left with providing differential economic value. It's like quibbling that offense/defense doesn't matter in football, only points matter. In the same way, harping on leverage is both true and not particularly distinguishing in its helpfulness. Or to tie this into the OP, adding value to society does not give you much leverage in a capitalist system. Adding economic value does. And I think people focus on the wrong value system and get frustrated when the expected results don't follow.
Yup, and that is going to be the defining problem of the next decade - IMHO even more than climate change and most other environmental issues.
The lower classes aren't going to be threatened by AI, not for a looong time until we get "I, Robot"/Star Trek TNG Data-like robots with precision dexterity comparable to a human. The jobs they do aren't automatable at all, have been automated long ago so it's not an issue any more (most of manufacturing, mining) or human labor will be cheaper than a robot replacing it (which is sad enough - shouldn't automation actually free the masses from toiling in hard and rewardless jobs?).
The higher classes (depending on definition, usually the 1-10 top % of wealth) aren't going to be threatened by AI either. Those who have the money have the power after all, and almost everyone with a fully paid off home and a second one to rent out should be set even for the worst cases.
But the wide masses? Realistically, as you said most of them will fall to lower classes in income and lifestyle, and a very few luckshotters (="AI prompt engineers") manage to raise up. We've seen in the last decade or so just how powerful and reactionary the search of these masses for a scapegoat for their externally-caused misfortune can be, and it can take on a lot of different forms: nationalism to far-right xenophobia, antisemitism, anti-muslim, anti-intellectual ("antivaxxers", a ton of "homeschoolers")... that's a lot to take on even for a stable society, and external influence (enemy nation propaganda, financially motivated propaganda such as the Macedonian troll farms hunting for ad placements, domestic media moguls) makes it even worse.
Honestly I have zero idea how the fuck humanity is supposed to continue to exist as a civilization longer than 10-20 years. Even the best of our democracies are falling apart not just at the fringes (you always had and will have loons) but in the center.
>or human labor will be cheaper than a robot replacing it
One option is to create an automation tax, which makes human labor more competitive while also supplanting the income taxes that are lost due to automation. That's just one example, but like so many problems that are of man-made origin, there are also potential solutions of man-made origin. They are not natural laws.
As if those in power would ever agree to such a thing. "Trickle down" is a myth, US Congress has been gridlocked for decades, and most other parliaments in the world are so infested with lobbyists that you will not ever see an automation tax without widespread violent unrest. People living outside of democracies have it even worse.
There are zero signs that democracy is capable of fixing the issues at hand, it's all just too damn corrupt, and fighting for utter and bare survival against the rise of authoritarianism. Yes, you sometimes have "social democrats" gain power (or authoritarians getting booted off their posts like PiS in Poland or what will sooner or later happen to Erdogan and Netanyahu), but they spend almost all of their political energy on undoing the worst "accomplishments" of their predecessors, and all too often get booted themselves by a population unwilling or financially unable to go through change (as we're seeing in Germany).
Since everything is going to get automated, this is one of the more complex ways of getting to the same place: industry taxes that pay for a universal income.
My take is just as radical, but makes more philosophical and moral sense (to me).
All newly discovered non-human made resources are considered a common inheritance. Industry (whoever) gets bounties for discovery, and bids to extract, with that money getting distributed to everyone evenly.
What I like about it: inheritance is considered a moral transaction, joint inheritance of the planet is an ideal we often give lip service to, flat distribution does not favor or disfavor anyone, and this distribution method doesn't take anything from anyone (after a gradual transition to the new regime).
The value of total natural resources extracted is going to increase over time, even as unit pricing goes down. Access to resources beyond Earth will give the value of new resources a big boost.
Something will have to radically change, regardless.
https://nces.ed.gov/programs/coe/indicator/cba/annual-earnin...
To relate this to your bachelor's degree example, if everyone earned a bachelor's degree in e.g. physics then the earnings premium for STEM graduates would drop if not vanish. But I expect that society's collective material prosperity would increase. I say "earned" rather than "got mailed a certificate" because getting a certificate without developing the corresponding abilities is not useful.
The point of the experience is that it allows you to produce more value with the same inputs.
To wit: nations with higher educational attainment do not see the averaging that you suggest, instead they see rather remarkable standards of living.
If it slides too far to the signal side, it becomes a bad proxy. Considering grade inflation is a thing, I wonder how much of that relationship gets eroded
We choose yo ignore wealth ineqaulty for delusional reasons
"Recall that the advent of pre-AI computing made the expert judgment of professional decision- makers more consequential and more valuable by speeding the task of acquiring and organizing information. Simultaneously, computerization devalued and displaced the procedural expertise that was the stock-in-trade of many middle-skill workers. But imagine a technology that could invert this process: what would it look like? It would support and supplement judgment, thus enabling a larger set of non-elite workers to engage in high-stakes decision-making. It would simultaneously temper the monopoly power that doctors hold over medical care, lawyers over document production, software engineers over computer code, professors over undergraduate education, etc."
"Artificial Intelligence is this inversion technology. By providing decision support in the form of real-time guidance and guardrails, AI could enable a larger set of workers possessing complementary knowledge to perform some of the higher-stakes decision-making tasks currently arrogated to elite experts like doctors, lawyers, coders and educators. This would improve the quality of jobs for workers without college degrees, moderate earnings inequality, and — akin to what the Industrial Revolution did for consumer goods — lower the cost of key services such as healthcare, education and legal expertise."
"Moderate earnings inequality" means "fewer high-paying jobs", as someone pointed out in another comment. From where does the pressure come to raise incomes across the board? That's what unions were for. In the US, the unions were crushed. The whole idea of paying people more than they are worth as an economic unit, a key goal of the union movement, is almost forgotten. Yet that's what this paper assumes will happen. Somehow.
This is "trickle-down" economics with an "AI" label pasted on it.
The author has better papers. His "Why are there still so many jobs" (2014) [1] is worth a read. He makes predictions one can now check.
[1] https://economics.mit.edu/sites/default/files/publications/w...
No, it wasn't. A company can't survive long paying all its employees more than they're worth. The goal of unions was, at least as far as pay is concerned, only to reduce the margin that was skimmed off of employee value by employers.
"What we want to consider is, first, to make our employment more secure, and, secondly, to make wages more permanent, and, thirdly, to give these poor people a chance to work. The laborer has been regarded as a mere producing machine ... but back of labor is the soul of man and honesty of purpose and aspiration. Now you cannot, as the political economists and college professors, say that labor is a commodity to be bought and sold. I say we are American citizens with the heritage of all the great men who have stood before us; men who have sacrificed all in the cause except honor. . . . I say the labor movement is a fixed fact. It has grown out of the necessities of the people, and, although some may desire to see it fail, still the labor movement will be found to have a strong lodgment in the hearts of the people, and we will go on until success has been achieved!"
Gompers was noted for answering, when asked what labor wanted, "More".
[1] http://historymuse.net/readings/GompersWhatdoestheworkingman...
I would say at most 1 out of 50 are genuinely really unique in some way, and that’s being generous, the rest of the population are just regular average Joes or not too far off.
That few people today even know this was once mainstream is the result of a very effective lobbying campaign started in the early 1970s. Milton Friedman provided the theory, and the U.S. Chamber of Commerce providing the money. There was a master plan for this, the Lewis Powell Memorandum.[2] This is where the organized conservative movement came from, from think tanks to talk shows.
[1] https://www.ilo.org/dyn/normlex/en/f?p=1000:62:0::NO:62:P62_...
e.g. Wages can be both inputs and outputs. Because there’s no ultimate judge deciding how these things should be considered.
Millions of writers can write billions of words about it one way or the other, but at the end of the day they don’t count for anything, when compared to 8 billion lives worth of daily decisions.
No, it assumes that companies can't pull money out of the ether. There definitionally isn't enough value produced by the collective labor of all the employees for them to each be paid more than their value.
To pick a few things we could do: -> criminal liability for white collar crime -> muscular antitrust -> roll back citizen's united -> public option for health insurance -> taxes on real estate as investment
I don't understand why people think this is all or nothing. Plenty of countries are doing much better than the US on these fronts.
The problem is that for the people on the 1000000 end of the spectrum, the same mechanics that encourage investment in productive activities also encourage investment in corruption, as they make no distinction between the two. With hundreds of billions in present-value-of-future-cash-flows at stake, almost any expenditure is justified. The laws of economics say that for you lobbying is a hobby, but for them it's a job with ROI (a whole office full of jobs, in fact). That's a steep advantage to overcome, but we've done it before and we'll do it again.
The world is mostly inhabited by people who work for a living but mostly run by people who own things for a living.
These two sources of income (or capital gains -- heaven forbid we tax our rich like we tax our plebs) lead to highly distinct and often directly opposed objective functions. The extent to which this leads to a self-serving policy spiral for the rich is a function of the degree to which the inhabited-by/run-by distinction is true: with a steep enough wealth distribution, yeah, it's fair for people at the bottom to say that people at the top actually are all colluding.
The self-serving policy spiral doesn't look like "Lidl hikes prices and Tesco doesn't, so you can just shop at Tesco," it looks like "merger approved between Lidl and Tesco, now listen to the surviving PR department tell you how this will surely save you money."
That's just for the hypothetical streaming agency. When you add up all these little increases of just about every good and compare that to wage growth, you see that wages are stagnating relative to other metrics like cost of living or business/national productivity, in the US or the UK or just about anywhere with these sort of metrics available. You can raise minimum wages to fend of a total economic collapse every now and then but these increases find their way to these other hands reaching into your pocket before long as well. A constant sucking noise and a march to what, I'm not sure, but it won't be pretty I expect.
Then came deregulation. Within five years, PG&E went bankrupt.[1]
[1] https://en.wikipedia.org/wiki/2000%E2%80%932001_California_e...
It's not like deregulation was the only issue, that just accelerated the bullshit. PG&E had neglected ("deferred") maintenance for many decades - the deadly 2018 wildfire was caused by a cable hook that was around a century old and had never been inspected for a very VERY long time until it eventually failed [1]. IMHO, that amount of maintenance neglect should have been classified as financial fraud:
> PG&E has said in recent court filings it didn’t track when its hooks were installed or how long they had been exposed to weather. It also didn’t routinely inspect them to look for corrosion or wear. Before the Camp Fire, PG&E inspection forms didn’t even include space to note the condition of hooks and other hardware.
How PG&E dared to pay dividends when they didn't even know how much liabilities they had on the books, that is just utterly insane.
[1] https://www.wsj.com/articles/this-old-metal-hook-could-deter...
[1] https://www.theguardian.com/books/2014/nov/20/ursula-k-le-gu...
Ok, and I claim that the problem with the current model of capitalism, as opposed to the post-WWII model, is that it's oppositional to a middle class - the incentives are opposed.
Now the question is about the societal outcomes we expect from the systems people create. Why should we prefer greater inequity and an even more complete loss of the middle class? These things are all being imposed by human-manufactured systems - we can control them.
We don't even need to throw AI or human desperation or planned economies on this bonfire of ideas to recognize that, as initially posed, the problem is itself incorrect.
Control of these systems lies in the capital class not in the working class. They prefer greater inequality, you and I probably don't but you and I have zero agency to do anything at all about how the world works. We lack support and we lack money to advertise our position and garner support relative to the groups empowered and enriched by the status quo.
My understanding is that most people's understanding of "middle class" is "about 20% richer than me", and that this remains true regardless of how rich one is.
Most people, for most of the years since Joseph Marie Jacquard realised he could control a loom with punched cards, the Red Queen race of automation has given us more stuff and experiences per hour of labour — if you wanted to live the 1820s idea of a middle class life, you can retire as soon as you get around €$£ 25-50k of savings (just don't start a family, the rest of society considers this standard to be unacceptable for that, what with no electricity).
If this improvement in living standards is despite, or because of, each business trying to maximise revenue, is basically the entire disagreement between Adam Smith and Karl Marx; though it's worth also noting that neither liked rent-seeking, and if you have a completely efficient market then nobody can extract any profits because competition drives margins to zero… but also since then we've had research show that complex markets aren't and can't be efficient: https://arxiv.org/abs/1002.2284
There are better definitions.
> Karl Marx
You said Voldemort's name! Now he will surely haunt us.
In any case, your objective function dramatically changes when you own things for a living rather than work for a living. Many people would call someone with a 1MM retirement portfolio rich, but 4% returns on a 1MM portfolio is $40k/yr and probably isn't going to replace a job, not for someone who managed to amass 1MM in the first place. However, 4% returns on a 10MM portfolio is $400k/yr, and it might. 4% on $100MM is $4MM/yr, and it probably will. So the composition of incentives switches from worker-like to owner-like somewhere around $10MM net worth, and this is worth noting.
If you ever saw old timey movies from the 50s or 60s where "millionaire" was said with reverence, note that $1MM in 60s dollars is $10MM in 2024 dollars. They were aware of the class divide back then in a way that people generally are not today.
I'm a linguistic descriptivist, not a prescriptivist. This is how I think most use the term, even if there's a more useful term of art in economics or in politics.
> You said Voldemort's name! Now he will surely haunt us
One might even say that a spectre is haunting Europe… ;P
> So the composition of incentives switches from worker-like to owner-like somewhere around $10MM net worth, and this is worth noting.
In a sense, sure. This is also close to what I'm told is the actuarial value of a human life in the USA, and that's not a mere coincidence.
But also, at that level you most likely only got the money by having a very different objective function to begin with, as it comes from starting a successful business. (Other options are "inherit" and "win big on lottery" and both groups have differently bad reputations to the kind of bad reputation that successful businesses people get; I can't possibly comment how accurate any of these reputations are).
But I also see the dark side. The economic notion of "value" being maximized is actually "wealth weighted value," a tiny but enormously consequential difference. Creating value isn't about doing what people want, it's about doing what rich people want, and rich people mostly want to get paid for being rich. Competition is supposed to keep this in check, but we police competition with the vigor and enthusiasm of a decaffeinated sloth, because that's what the people who own the economy want. When suits brag openly about anticompetitive strategy -- network effects, platform effects, last-mile and two-sided-market dynamics -- and regulators snooze more loudly every day, it's pretty clear that the system is captured.
Not that tech is unique, in fact we are pretty good all things considered. We still provide more value for less money every year. Many sectors have been doing the opposite for decades. Real estate is probably the stinkiest sector, where the incumbents are perpetually voting for dysfunction to pump the value of their own properties. Health care has similarly been squeezing its own training pipeline for the explicit, stated-out-loud purpose of pumping salaries, and now that boomers are aging this is going to get spectacularly ripe. None of this is wise stewardship, but it follows directly from incentives and power. Anyone who wants to claim that capitalism produces systematically good incentives had better have a really good answer for this cartel shit, because it rules the economy.
It's not just the economy, though. Capital interests are in the driver's seat in government as well. The extent to which tax policy has been pushed away from the capital gains loop and towards the wage-labor loop is almost comical. We brought women into the workforce without "meeting in the middle" on aggregate hours so now the average couple isn't allowed to raise a baby which shockingly produced a fertility crisis. Our trade policy has been re-geared so heavily to pump assets and dump exports that Alexander Hamilton is spinning in his grave next to the 30 year old opioid addict who is in his own grave because said policy changes sent his job overseas so that rich people could benefit from the higher stock returns.
In terms of the capital/labor pendulum, we've actually been here before, after the industrial revolution failed to net trickle down. I don't think we need glorious *ist revolution, but I do think we need another Roosevelt.
Indeed; if there were no dark side, Marx would have had nothing to write about.
The difference between "money" and "value" can also be found in the difference between what is measurable and what is important is noted in Goodhart's law, which was originally specifically about two different measures of money and why putting pressure on them specifically for economic reasons was not having the anticipated result: https://en.wikipedia.org/wiki/Goodhart%27s_law
I hope that the successor to Smith and Marx will account both for Goodhart and for Nash equilibria/game theory, both of which were developed and formalised over a century after the Communist Manifesto and around two centuries after The Wealth of Nations.
That's nonsense. It's not just businessmen that compete for consumer's money. Consumers compete for businessmen's money as well.
Suppose a widget-making enterprise made 3% profit on it's capital. What's the sensible thing a widget maker should do? Obviously, fire his employees, sell or scuttle his machines, turn everything into cash and turn the cash into 4.313% treasury bonds.
And the destruction of widget-making machines and enterprises shall continue until widgets sell at a margin that justifies their existence. And if nobody can afford widgets at the that price, there shall be no more widgets.
I find it amazing how people can talk about capitalism, without any understanding of the concept of capital, it's formation and destruction.
I wrote "if", and then provided a citation that the clause is false.
> Suppose a widget-making enterprise made 3% profit on it's capital. What's the sensible thing a widget maker should do? Obviously, fire his employees, sell or scuttle his machines, turn everything into cash and turn the cash into 4.313% treasury bonds.
This seems to be even more of a flawed toy model as the words of mine which your complaining about.
Even assuming that's the same period so these are equivalent (3% per sale but you make a widget in 4 hours and sell them just as fast is much more than 4.313% per year): Why does a treasury bond earn interest in the first place? How does the government cover this cost?
The usual answer to the latter is "the bond is to raise money for a thing the government wants to do, in order to boost the economic output of the country, thus future taxes pay for it". If they set the interest rate so high that everyone laid off all their workers, this would not be effective, so a government would not do it.
You also suggest they may sell their machines: to whom? If everyone is doing this, there is no market to sell to.
--
Now consider: an investor has investment opportunities of 2% and 3% return. The economic choice is 3%, you are a widget maker promising this, they invest in your business.
Your competitor knows your profit margin and your costs, they know that if they can make widgets at the same cost then they can charge 3% and get a share of the same customers… except that both producers have fixed costs as well as per-widget costs, and the market size is not in their power to change.
The competitor's options:
1) Continue to compete at the same profit margin
2) Undercut your prices: if all the customers are perfectly rational spheres in a vacuum, then they will immediately switch supplier even with an infinitesimal price reduction, significantly increasing their sales at the cost of yours
3) Increase their prices in order to promise a higher return on investment to their investors — but they know that if they do this, homo-economicus customers will all reject their widgets in favour of yours, and that their business will therefore earn nothing, and they know that their investors know this too
4) Leave the market entirely, potentially allowing you to become a monopoly and raise your profit margin (to maximise their options within the supply/demand curve limits)
Option 2 is the game theoretic choice on any given round, but it's not a one-shot game and you as the original widget company get to respond before running out of money. The whole thing is symmetric, so you tend towards no profit even though you'd like to collude because it's a prisoner's dilemma payoff matrix.
This is also symmetric with regard to different industries, so switching from widgets to gadgets in order to boost return on investment merely changes the players and not the, ah, game.
What slows this process down includes, but is not limited to (because I'm not an economist):
1) Customers aren't perfect spheres of economic rationality
2) People don't know all their price options
3) People are lazy
4) Your costs aren't likely to be exactly the same as your competitor's costs
These[0] reasons are why markets aren't efficient.
[0] and I would hope many other things, because otherwise this topic is simpler than I expected
Most people are afraid of “next-level” AI, where no operator is required. If trucks can be moved without anyone at the wheel, truck driving is a job that won’t exist anymore. As simple as that.
Whether trucks can be 100% autonomous remains to see.
A truly 100% autonomous vehicle requires a much higher level of intelligence than a self driving vehicle with a driver able to take the wheel when necessary.
Take the case when some work is happening on the road and workers make signs with their hands to tell you to go this or that way. But on the same road there is also someone who is dressed up as a policeman cause it’s Halloween, and he’s waving at some friends.
It's possible someone might figure out a way to create a training loop using a multi-modal LLM to generate synthetic training data based on the situation you just explained and then updating the driving model by training on this new data until its performance improves on the task.
I imagine we will reach a place where fully autonomous vehicles will pull off to the side of the road in certain weather. Which I wish we could force for humans, but seems infeasible to implement.
Seems likely to me; we built EV's before we built the infrastructure to support them and conversely we created in infrastructure for petroleum vehicles before they entered mass production at the largest of scale in the period (1950's) after the fact.
I don't see how autonomous vehicles are not going to become a reality. Perhaps not in my lifetime, but, absolutely likely and possible.
I don't doubt that, but the timeframe is unknown. 5 years? 10 years? Within our lifetime?
So what we have is the ability to input data, but not yet a delivery system and retrieval system that can fit on say, a small chip, or light array, or other small systems.
It's a giant part of reason we'll see diminishing returns with data being applied in classical material approaches. New materials (currently being workd on, like graphite and others) will be needed to harness the compute power to enable large scale data capabilities at increasingly smaller and smaller levels (already a well known issue to be resolved).
Similar in physical approach but different in application would be TinyLM.
Timeline? Not sure but we created ION drives over 30 years ago. Seems to me we're limited not by the science, but the material needs to continue technological advancement. Seems to me autonomous driving is within reality in under 25 years. If I had to put a guestimate on it.
A lot of this breathless talk surrounding this turn of AI is so uncomfortably reminiscent of what I’ve seen before in the mainstream the last turns around the 1970’s and 1980’s, and the potential failure mode might not be so different: solving the last 5-10% is tantalizingly close but remains stubbornly out of reach of calls for the “more cowbell” of each era or call to action by the sales legions (currently cowbells look like NVIDIA boards and various counts of AI models be it tokens or what have you), and the last 5-10% is the necessary advance to cross the chasm.
I love and use the tech myself every hour, but it has deep gaps I don’t see being resolved even incrementally between versions or competitors.
As technology becomes more intelligent, the human comparative advantage shrinks. But what is likely to happen? Don’t look to the past. Look at the fundamentals of the system dynamics.
Consider an analogy. The differential equation that governs blackbody radiation depends on the difference between the object and the environment. Similarly, a person’s earning capability is proportional to the difference between their capability and the next best option. As the next best option gets better and better, the human differential advantage craters.
Max Tegmark uses the metaphor of a rising ocean to signify the ever greater capabilities of intelligent machines.
AI isn’t just coming for low-skilled jobs. It isn’t just supplementing knowledge work. AI will take more and more intellectually demanding jobs. This arguably doesn’t even require AGI. This just requires industry deploying capable AI’s that can do enough tasks to replace humans one industry at a time.
So people that have an ownership stake in these AIs will do fine.
My take: The others should demand a universal basic income. How do we find ways to allow people to live good lives? People that weren’t lucky enough to be born into the right places in society don’t deserve to be displaced by AI.
If you read the AI literature, I’m not alone in what I’ve said above.
Granted, I’m also not offering particular time frames and not covering all future scenarios. But what I’m saying is more considered than the usual default ‘reasoning’ you hear people say.
You're right to be skeptical. Don't believe them. Demand a greater slice of the created value for yourself, and until you get it (and it manages to have a positive impact on your finances - see inflation), tell these people to shove their whitepapers where the sun doesn't shine.
Note that in this analysis, if you are a software engineer, you are not the "middle class" as most signs point to. Rather you are a member of an elite that the middle class must wrest power from.
The identity of the middle class is another question. You might assume that it's made of the people you know, but more likely it's underpaid workers from developing countries.
Healthcare should be free for all. UBI. And of course continued education provided to train workforce into something else.
We have people and companies with so much power that it rivals governments, and in the US we also have a large segment of the population who has been conditioned to believe that the government is bad and having less of it will make it better, but it seems to me like better government that actually supports the needs/desires of the populace at large is required to make our society better for everyone.
Yes, AI will lower the barrier for entry level positions (IMO a good thing; more chances for smart creative people to get in the game who may have been failed by more traditional systems). But it will also allow the creation of projects by much smaller teams, who no longer need to stress the details and have a focus on making a thing just to try it out.
Individuals will be able to get much more experimental and iterate way faster; if the amount of investment to build an MVP is less than the investment to research and assess all that crap that usually has to go into a commercial product (market fit, funding, marketing, sales, etc.) then you can just make the damn thing and see if the pasta sticks to the wall. This could be a really great environment for testing out weird ideas, some of which might be completely novel and break the mold in ways that are simply not feasible if you have to manage all that other stuff.
I'm also excited to see what new stuff pops into the OSS space as a result of devs-with-jobs being able to hack out a proof of concept in a couple hours and smacking some stuff together over weekends.
No. Empirically, has it?
It will do the opposite. It removes trust, therefore making existing powers more powerful. Ain’t going to democratize anything but cybercrime.
It hasn't really been around long enough for an empirical analysis of effects. It's also incredibly hard to measure effects with the systems themselves being in such flux.
For what it's worth, I think categorical statements of any stripe are a little premature. Even the EA folks speak in probabilistic margins.
Yes. Is that not obvious? It’s already a major problem in educational settings. I’ve seen video of people using it for job interviews.
Have you been handed any code yet that was written partially or wholly by gpt? If not, you will soon, and then you’ll get it.
I knew a very comfortably well-off lady who made a fine point of how she never used the automatic machines in the supermarket, because she wanted to ensure that those poor cashiers didn't lose their jobs.
How noble!
Generally speaking, this is an oxymoron.
Sorry, I've become very pessimistic these days. If something can be used to permanantly reduce FTEs, it will be done.
How long can this state of affairs stay afloat? Does a business require the ability to exploit workers in order to succeed and be useful?
> The unique opportunity that AI offers humanity is to push back against the process started by computerization
Yes. More automation and computers will clearly solve the problems created by automation and computers.
I can't find where he probably says something like "outsourcing of these jobs will also improve middle class outcomes" but I assume it's in there through some kind of weasel-wording.
https://kottke.org/21/04/ted-chiang-fears-of-technology-are-...
So long as the dominant theology in this country is extractive capitalism, nothing that tilts the cost balance towards capital and away from labor will be good for the middle class or the average person.
Actual title of PDF: "Applying AI to Rebuild Middle Class Jobs"
Let's look at history. The industrial revolution was predicted to create copious leasure time for the working class. How did that turn out?
The steam engine, the cotton gin, the automobile, the computer...
Every single technological advance that was predicted to improve the lives of the working class due to massive increases in productivity, was used instead to almost exclusivly bennefit ownership.
This is exactly what will happen with current LLM tech (please don't call the bullshit generators "AI")
If the tech cannot be deployed so as to exclusively bennefit ownership, it won't be deployed at all...