Efficiency improvements tend not to create more leisure time, because one person working eight hours a day remains more efficient than two people each working four hours a day. The reasons behind this are communication and training overheads. If AI leads to a world where people don't really have to know anything to do their jobs - just provide high-level judgements that LLMs seem farther away from than they are from accuracy, or if they could somehow keep the human beings out of meetings, the forces keeping labor concentrated could abate.
On the other hand, if AI accuracy limitations drive the labor demand even further towards expertise, and if making tasks higher-level raises the communication requirements rather than somehow reducing them, the preference for having a few people work 80-hour weeks while twice as many people remain unemployed will become even stronger.
I'm surprised that someone has pointed this out because it is 100% on point.
The moment your economy is productive enough that it is not necessary to hire every single person, employers start firing people rather than cut the working hours evenly across the population, because each employee represents a fixed cost. Hence you either work full time or you don't work at all.
This is peak utilization of efficiency in the USA. This is what corporations think is acceptable optimization/efficacy looks like if we give them the choice. And then the non-wage slave caste people will complain that this single mother didn't raise her children well enough and call her lazy/a bad parent:
> Efficiency improvements tend not to create more leisure time, because one person working eight hours a day remains more efficient than two people each working four hours a day.
Efficiency improvements tend not to create more leisure time because the gains all go to the people at the top.
Taking that at face value, what happened before 40 years ago? There was unimaginable growth in per-person GDP, so people could have plausibly kicked back and relaxed rather than toiling in factories.
If you're asking "what caused the change?" the answer is "Ronald Reagan".
If you're asking "why didn't the people before then work less?" the answer is "because productivity gains hadn't yet made that possible".
Just to expand on that a bit, in case it isn't fully clear:
There is a level of productivity per worker required to support everyone in the society. This level fluctuates some with the overall standard of living, but does not vary with total productivity, population, or GDP. Let's call this level P, for Parity. (And let's assume it's not "just barely enough to support everyone", but "enough to support them comfortably and reliably, with a decent buffer".)
Once the level of productivity per worker passes certain thresholds—multiples of P—the total amount of work required to maintain the society at the same level drops. More work produces surplus. That surplus can be then used in a variety of ways. One of those ways is by reducing the amount of work being done. So, for instance, if the productivity level reaches 2P, then every worker can work half the amount of time they were working before, and still be producing enough to fully provide for everyone. If it reaches 3P, then every worker can work 1/3 the amount of time, and so on.
If, instead, the surplus is captured by the wealthy, the workers don't see benefit, and inequality grows.
I don't know exactly when real-world productivity got enough higher than P that we could realistically start reducing worked hours, but it definitely did so at some point in the last several decades. I'd say that at this point, just as a rough estimate, we're probably somewhere between 1.5P and 2P, but that's not really my field of expertise. But because the wealthy have captured approximately all productivity gains above the level we were at in 1980, they have seen their wealth massively increase, while the rest of us have just been scraping by.
>If you're asking "why didn't the people before then work less?" the answer is "because productivity gains hadn't yet made that possible".
>[...] I don't know exactly when real-world productivity got enough higher than P that we could realistically start reducing worked hours, but it definitely did so at some point in the last several decades. [...]
GDP per capita has been growing exponentially for centuries[1]. Is there some arbitrary GDP per capita level where people should be expected to kick back and relax? Why makes your arbitrary line more or less correct than someone else's arbitrary line? Moreover people's revealed preferences show that for most people, that line hasn't been reached yet. Why should people's revealed preferences be overridden by whatever number you came up with?
>If, instead, the surplus is captured by the wealthy, the workers don't see benefit, and inequality grows.
Economic statistics shows it hasn't been captured. Even if you're some sort of marxist that thinks labor's share of GDP should be 100%, at the very most this means the point at which everyone can kick back and relax is delayed by 40%.
This isn't about GDP in a currency sense. It's about Creating Enough Stuff that people can thrive.
Like I said: there is a threshold of Stuff Created Per Person above which providing a comfortable life for every person is purely a distribution problem. For most of human history, we have not been above that threshold.
"Subsistence farming", for instance, is effectively defined by only being able to meet the much lower threshold of "enough that people can survive".
A post-scarcity society is, broadly speaking, defined by being able to produce enough for everyone to thrive with minimal work from anyone.
We are somewhere between the two, but we are reaching the point where we're closer to the latter than the former. Technological advancements have, for some time, ensured that we have enough food for everyone on earth (again, there's still a distribution problem; that part is nearly 100% about politics, not about scarcity). If the very wealthy had not captured all the productivity increases since 1980, I don't know what else we could have achieved, but it wouldn't have been small.
> Economic statistics shows it hasn't been captured.
Look at any graph of income growth by quintile that goes back to the middle of the 20th century or earlier, and you'll see it starts with some roughly parallel lines, and then one line that keeps going up at about the same slope, while the rest stay nearly flat.
The 4 day work week movement continues to move forward, and becomes more likely to succeed as workers gain more power as the prime working age population continues to shrink over decades due to structural demographics (total fertility rates below replacement in most of the world).
Well for one, no computers collating every persons value to the economy to buy and sell as and manipulate through targeted effort
Credit system didn’t exist, which conveniently grandfathered in all the Bloomberg, Trump, and other old money …obviously inheritance made them geniuses who deserved it
Basically Boomers came of age 40 years ago and needed something to do. So the 80-90 year olds of the day handed them all the power.
I mean come on. Do we really need to circumlocute the cause? Why the old rich people writing the rules and always winning is suspicious af still?
All of that might be true, but it doesn't address my core question: why didn't people kick back and relax 50 or even 100 years ago? If people 50, 100, or 200 years ago made the choice to keep working hours constant rather than convert productivity gains into leisure, why is the choice to keep working hours constant today suddenly caused by malign influence of the powers that be?
>The average person worked fewer hours through almost all of recorded history until the last few decades.
If you're claiming that people worked fewer hours during the industrial revolution than today, I'll need a citation for that. If you're referencing the claims made by "Original affluent society", that has problems around how working hours are counted.
The criticisms have the same lack of information problems.
So modern anecdotes make the most sense. I grew up in 80s dairyland. None of the farmers worked a 9-5 but they rarely worked 40+ too.
We worked much much less in rural-landia before the last few decades gave rise to "service and knowledge" work with no concrete goal but "make line go up".
Knowledge work comes along with zero concrete termination points. Programmers grinding code 80+ hours are not stopping to see if it's useful or just repetition.
Sure from a physics perspective they're doing "work". From a lived experience, to real needs of biology, economy perspective they're just sitting at a computer juicing their hormones, while exploiting farmers labor, who now works more hours as more knowledge workers contribute less real outputs essential to biology. Same with why carpentry and other trades services are so expensive; supply and demand. Millennials wanted 24/7 office jobs.
While I grew up in farmland I later went into EE, and have a good sense for who is moving the ball. It isn't software people. Their biology is addicted to a stupid loop of zero real productivity. Playing abstract snake games in one's head, to count initialized memory registers and recording their data is some basic bitch work marketed to the point of fostering delusion it's cutting in 2010-2020 to use 1960s style syntaxes to manage machines.
Efficiency improvements tend not to create more leisure time because the capacity for the economy to create "bullshit work" (e.g. work which is engaged in zero sum "wealth defense") is unbounded.
For efficiency improvements to create more leisure time or even let the bottom 50% retire early, society would need to be radically restructured so that its locus of control is not capital. This would probably only happen with violence.
This might happen one day, but for now, efficiency improvements get capitalized into the ponzi-esque stock market while the efficiency gains which could be realized at the bottom of the pyramid get "burned off" via inflation targeting of 2%. A quirky side effect of this is that the vast efficiency improvements we have seen have not even been allowed to prevent a pensions crisis.
The Economist, serving in its capacity as a dutiful servant to neoliberal capital, frets equally about AI inevitably causing us to "run out of jobs" as it does a demographic crisis inevitably causing us to "run out of workers" (https://archive.ph/6hgYq).
>This might happen one day, but for now, efficiency improvements get capitalized into the ponzi-esque stock market
That's a seductive narrative, but not backed up by data. The share of GDP that goes to labor is has held relatively stable in the past few decades. It has admittedly fallen, but it's on the order of a few percentage points, whereas the GDP per capita more than tripled.
How do you square these “growing wages” against the zeitgeist which says more and more people are living hand-to-mouth? At what point do these statistics get tossed out because they have little to no bearing on the lived experience of the great mass of people? At what point does one acknowledge that these statistics seem like they’re compiled simply to muddy the waters to make reality seem different than it is?
Famously, one can spin statistics any way they wish. Prove to us this isn’t the case here.
> How do you square these “growing wages” against the zeitgeist which says more and more people are living hand-to-mouth?
Both can be true if the distribution has changed, which seems to be the case:
"Over the last four decades, the income gap between more- and less-educated workers has grown significantly; the study finds that automation accounts for more than half of that increase."
Additionally, necessities like food and fuel seem to have jumped in price, outpacing wage growth over the past few years, while other important goods and services (housing, health care, higher education) have outpaced inflation for decades.
If they weighted nondiscretionary housing, education and healthcare in the inflation basket properly and de-emphasized discretionary electronics and clothes I doubt it would even be flat.
> For efficiency improvements to create more leisure time or even more wealth (for the bottom 50%), society would need to be radically restructured so that its locus of control is not capital.
And even then, even a small percentage of bad actors who are selfish would likely restructure it back due to technology enabling them to do so! The fact is, the elite at the top is a steady state that is hard to leave if you worship efficiency. Only the Amish and some native tribes have realized that the only way out is not to play.
This seems circular… capital has become such an important gatekeeper in society in the first place because most people aren’t that trustworthy, and will likely betray any trust or faith placed in them if put under enough pressure.
So society has to depend on something fully independent of any specific person or group and that cannot be changed without leaving behind a long papertrail.
In other words nobody dreams of becoming a banker/accountant/auditor/etc. at 5 year old. Probably even double entry bookkeeping would be superfluous if everyone were virtuous paragons.
> This seems circular… capital has become such an important gatekeeper in society in the first place because most people aren’t that trustworthy, and will likely betray any trust or faith placed in them if put under enough pressure.
True, but it's also mainly because we have a society that is based on technological innovation with rapid transportation/communication. I was only responding to the other reply that seemed to imply that there was an alternative restructuring that puts less capital at the top.
>capital has become such an important gatekeeper in society in the first place because most people aren’t that trustworthy
Not really. It became that way because wealth begets power.
The narrative it tells us is that we should be terrified to change this because nobody else except the owners of capital can be trusted as stewards of our economy and political system (particularly not "government").
On the other hand, if AI accuracy limitations drive the labor demand even further towards expertise, and if making tasks higher-level raises the communication requirements rather than somehow reducing them, the preference for having a few people work 80-hour weeks while twice as many people remain unemployed will become even stronger.