Getting from Here to There: A Post-Labor Odyssey

Some thoughts about the transition period between now and a Fully Automated Luxury Space Communism future.

Previous Work

This blog post is a follow-on for my Post-Labor Economics and Economic Agency series. For some context:

1. Economic Agency: A Key Principle: In a near-future economy where AI and automation make human labor increasingly obsolete, we face potential societal collapse unless we fundamentally restructure how people participate in the economy. The solution requires a three-pronged approach combining universal basic income, decentralized private ownership, and stakeholder capitalism to ensure everyone maintains their economic agency even in a post-labor world.

2. A Post-Labor Economics Manifesto: A comprehensive new framework called Post-Labor Economics proposes replacing outdated neoliberal policies with principles focused on decentralization, radical transparency, and human economic agency. By pushing decisions to the local level, requiring organizational transparency, prioritizing automation, and ensuring economic power remains competitive, this system promises to deliver higher GDP while distributing benefits more equitably and preparing society for an automated future.

3. What do I mean by “Post-Labor Economics”? As AI and robotics become superior to humans at virtually every task, we face an “economic agency paradox” where increased automation creates a death spiral of deflation and collapsed consumer demand. The solution may lie in transforming everyone from laborers into AI-assisted investors through universal asset tokenization, allowing humans to maintain economic agency by directing resources rather than providing labor in our automated future.

If you’re not up to speed, there’s just a couple of key terms you need to be familiar with:

1. “Better, faster, cheaper, safer” This is my mantra; when machines (AI, robots, etc) supersede humans on these four metrics for any given job, it becomes economically inevitable that machines will replace humans. Same reason that internet replaced messengers, tractors replaced manual laborers, and so on.

2. “Frontier of Automation” A term used by Anton Korinek to describe the total sphere of machine capabilities. Specifically, this is the level of task complexity that machines are capable of performing, encompassing both mechanical (physical world) and cognitive (abstract reasoning) skills. Soon, machines will be able to handle more complexity than most humans, meaning that no job remains safe from automation.

3. “Economic Agency Paradox” Technology is deflationary, however, with too much automation, we also lose jobs. So while goods and services become increasingly affordable, we paradoxically lose the ability to pay for those cheap goods and services because we have no jobs! Therefore, we’ve lost economic agency in this future. Even if we have UBI, we still have less authority over our economic destinies because we can’t work.

• lump of labor fallacy

• creative destruction

• labor dislocation/substitution

• NEETs and total employment

The Article

Based on the critical reception of my blog posts and videos on the subject of Post-Labor Economics, most people have bought into the scheme. The response has now, more or less, been: “Okay, Dave, we’re onboard with better, faster, cheaper, safer and we’ve even drank the Economic Agency kool-aid. But now what? You seem to be glossing over the transition period!”

I have personally said “It’s probably going to get worse before it gets better” and while that sounds ominous, I need to say that it’s probably not going to become catastrophic (e.g. Civil War bad) before it starts to get better. First, there are a few caveats that it would be irresponsible of me not to cover.

Lump of Labor Fallacy

Way back during the days of the First Industrial Revolution (1IR) a group of people called the Luddites sabotaged farm machinery, thinking that machines were taking a finite amount of work from them. Their reasoning was intuitive—yet still very wrong—that if these newfangled machines could come in and do their job in a quarter of the time, they’d all be sent home early and, as they were payed hourly wages, either they’d take home a quarter of their pay or three quarters of them would get fired.

Today this probably seems a bit quaint and silly. We invented automated switchboards and while that did destroy telephone operator jobs, no one is moaning or whining about bringing those old jobs back. The intuition from simple farmers two centuries ago was that there was simply a fixed amount of work to do, then everyone would go home. I mean, think about it from their perspective: if you were running a farm, you needed farm hands to help with jobs and chores, and once the work was done for the day, everyone went home. This was a very narrow view of the economy.

It is true that machines would eventually destroy their entire way of life. So, while they were sort of wrong, they weren’t entirely wrong.

Emails and printers destroyed typist pools too. Instead of Xerox, you’d have entire rooms full of typists like these making copies of documents. (AI generated image by the way)

This leads to the next critical concept:

Creative Destruction

This term, as many economic terms do, sounds a bit like an oxymoron. But if that sounds bad, just try on the original German pronunciation: “schöpferische Zerstörung” (I am not even going to try).

We’ve already seen a couple of examples: tractors and industrialized agriculture destroyed 18th century peasant lifestyles. Automated telephone routing destroyed telephone operators. Xerox destroyed typists. Here are a few more examples that are contemporary with the origin of this term creative destruction (1942, Schumpeter, Capitalism, Socialism and Democracy)

• The railroad destroyed the stagecoach business

• The automobile destroyed the railroad’s dominance, and horses

• Electric light replaced gas lighting

• Electronic calculators destroyed the mechanical calculator industry

The rule of thumb, according to economists, is technology always creates new jobs!

Hence, creative destruction. Out with the old, in with the new. Capitalists and neoliberals really get a hardon for this kind of thing. They see labor substitution as a sign of progress. Labor substitution, by the way, is a very euphemistic way of saying “Sorry, Jim, you’ve been mechanized out of a job.”

Mechanical drawing drafting pool before CAD software (not AI generated)

Right now you might be thinking, “Okay, Dave, not only have you glossed over the transition period but you’re also glossing over creative destruction! What if this new tech creates more jobs?”

It’s entirely possible. After all, just explaining the job of software developer to a Luddite in 1810 would have been nearly impossible. You’d have to walk them through the invention of electricity, mechanical computation, the Internet, and then describe how typing a bunch of symbols into this newfangled machine made it do a lot of math, which was somehow more valuable than growing food??? You’d have sounded like a crazy person and locked up in an insane asylum just for that!

From this perspective, it’s easy to conceive “Well, if the past is anything to go by, then perhaps this whole Post-Labor Economics thing is a bunch of hogwash.”

So what’s different this time?

The Meaning Economy

It took me a loooong time to figure this one out. I remember having long arguments with ChatGPT, which was artificially constrained by OpenAI’s training schemes, to insist that AI could never be more creative or intelligent than humans, that human emotional intelligence and intuition would always be superior to machines, which is intellectually disingenuous at best, and was insanely frustrating to work with. Even when I asked ChatGPT “For the sake of argument, assume that machines will be better, faster, cheaper, and safer than humans at all economically relevant tasks!” (using Sam Altman’s own words even!) and it would just default to a boilerplate list of principles such as life-long learning.

Just shoot me already. I don’t want Sam Altman’s future where “life-long learning” is the best answer he’s got for me.

I had a breakthrough when I realized we were looking strictly at the supply side of the labor economy. Without getting too lost in the weeds about labor market theory:

The labor supply represents the total workforce capacity available in a market, including both the number of potential working hours and the skills, expertise, and capabilities those workers possess. This encompasses not just how many people are willing to work at various wage levels, but also their qualifications, experience levels, and specialized knowledge that determines their productivity and market value. The labor supply thus represents both the quantitative dimensions (available hours, number of workers) and qualitative aspects (skill distribution, education levels, expertise) that together determine what human capital is available to employers at different compensation levels.

Well, if my theory is correct, and AI + robots are about to destroy the entire supply side of labor—basically creating hyper-abundance in intellectual and physical labor—then maybe it was time to look at the demand side of the labor equation?

When I made this suggestion, the entire conversation with ChatGPT shifted dramatically. Claude didn’t exist yet, so I have no idea what the conversation would have been like otherwise.

A new question emerged in these conversations?

In what conditions will humans always be willing to pay a premium for human labor?

I understand if this question seems a bit arcane. It goes like this: assume that machines do, indeed, become better, faster, cheaper, and safer than human labor. And I don’t mean by narrow margins of 10% or 20%. In some cases, you’re looking at efficiency gains of 1000x (100,000%) which means that humans cannot hold a candle to the machine. Evidence is mounting that this is happening. OpenAI’s o1-preview did about $8,000 worth of legal briefing work for $3.00. ChatGPT 4 is already better than most doctors, getting diagnostic accuracy of 90% where the average doctor is only 74% in some tests. Oh, and it costs just a few cents per diagnosis with ChatGPT.

So why “premium”? What do I mean by that?

Think of it this way: soon you might have a robot working in your house that has the equivalent of every PhD on the planet, and its only operating cost is the electricity it uses to charge its batteries and maybe a monthly subscription fee for cloud compute. But you’re looking at maybe $30 per month for the cloud service (let’s just say $100 for price gouging purposes) plus maybe $5 per day in electricity. For $250 per month, you have a robot that can do literally every job you could ever dream of doing, plus cooking and cleaning, plus raising your children, plus yard work, plus renovating your entire bathroom. In what circumstances would you STILL hire a human to do anything at all??

I mean, think about it: a machine with infinite patience, infinite skills, practically free to run compared to the value generated, and oh what if it also looks like Scarlett Johansson and is willing to have sex with you?

The list of jobs that you’d be willing to pay a huge premium for the privilege of human labor is getting pretty narrow. Even education might be right out, especially since the robots would already be equipped with all knowledge and infinite patience.

Here are a few kinds of jobs that might stick around:

1. Sentimental Jobs: This includes stuff like news anchors, YouTube creators, fashionistas, and philosophers. Humans generally prefer to get meaning from other humans.

2. Creative and Performance: Artists, musicians, performers. Same reason that we watch humans play chess even though computers are already better. Machines just aren’t that interesting to us.

3. Athletes and Celebrities: Same story as above; even if robots can beat the tar out of each other like rock’em sock’em robots, it’s just not as interesting. People love seeing Zendaya on the red carpet.

4. Experiences and Embodiment: I would not want robots to sit for me at a psychedelic retreat. Some things just really require that je ne sais quoi of being human.

5. Statutory Jobs: Jobs that, for whatever reason, are legally required to be done by a human. Think POTUS and judges and maybe even doctors (for a while). Even still, maybe it would be better if robots ran things… I would prefer to get pulled over by a robot cop than a dumb gun-toting power-hungry redneck with a Thin Blue Line sticker on his pickup truck and a Punisher wallet.

These are just a few ideas, but they all orbit around a first-principles view of humans, in other words, if we look at human biology and neurology, in which cases is having a human present just so rewarding that we’ll still pay for it?

If CNN replaced their news anchors with robots, do you think that would hurt or help their ratings? If you wanted to go to your favorite band’s concert, but it was just a holographic recording, would you go again?

No, humans are hardwired to value connection with other humans. Parasocial relationships are extremely valuable. This is what I call the meaning economy.

But can everyone join the meaning economy? If machines take over the vast majority of economic activity, does that mean we’ve all got to become performers and musicians and content creators? That sounds somewhat bleak.

The Transition Period

What will be the first signs that the AI “Jobpocalypse” is coming? Personally, rather than looking at headlines about layoffs, I would prefer to look at aggregate data.

NEETs (Not seeking Employment, Education, or Training) would be expected to rise sharply if the machines start coming for our jobs. For now, we seem fine.

The first metric is the NEET. These are people who have permanently checked out of the labor force. Either they can’t find a job or just don’t care anymore. You can see this number is starting to tick back up, but it’s still below Pandemic and Great Recession levels, so we’re probably okay for now.

The other prime indicator I’d look at is total labor force participation rate, or “population-employment ratio” which is an indicator of how many people, in total, are working.

The Great Lockdown briefly saw us plunge to just over 51% of the population working!

So far, so good. There are no signs of critical job loss, and I will be the first concede that, if I’m wrong, we may never see these numbers go pear shaped.

The reason that I look at an employment ratio is because it automatically adjusts for population growth (or decline). If you just look at absolute employment numbers, it is trending up (albeit more slowly than population) simply due to births and immigration.

Even so, in absolute terms, it almost looks like employment has plateaued!

If we have hit the final high-water mark of human employment, it could be all downhill from here.

That’s the “good news”—at least it’s good if you’re all for maintaining the status quo. Personally, I’d like to see us go ahead and automate away the labor economy ASAP. But I will concede that more time to prepare and adapt is not a bad thing.

Stop-Gap Measures: UBI

Let’s imagine that, over the next 5 years, for round numbers let’s just say 2030, unemployment starts to creep up, labor force participation rate starts to drift down, and total jobs never gets above 162 million (it’s currently hovering at 161 and some change).

There is one rule of thumb to remember here (and it’s why Trump got reelected!)

As economic agency decreases political anger and willpower to change things increases.

Four years of Democrats and Biden failed to move the needle for voters. Sure, the economy is doing fine, but we’re less hopeful and less certain than ever. So it’s time to try something else. Maybe the guy who acts like a human wrecking ball is the way to go?

This is the beauty of democracy. Some people say “humans aren’t wise enough to use democracy properly!” (as if they get to decide what “proper use” is for democracy). You might even say voting is mostly about vibes. But what if it is?

As we saw last time around with Trump, he was willing to try something new, like the stimulus checks. Now, many people will say that the stimulus checks caused inflation… but I’m sorry they are not causing inflation 5 years later. There’s actually increasing evidence that price gouging is a big cause of inflation. Look at the McDonald’s scandal. Shrinkflation and other such terms are being bandied about now.

Anyways, if we do start to see huge jobs loss, that would be highly deflationary to the economy. Less work means less consumer demand. So maybe we’d need stimulus checks to shore up inflation? That’s assuming that inflation is a good thing (MMT, or Modern Monetary Theory says it is)

Trump’s stimulus checks proved that the government can, indeed, just give out a bunch of people money. No middle-men and no corruption. Unlike the PPP loans, which were incredibly inefficient and did not go to consumers, stimulus checks were basically a prototype for UBI. If you have a bank account, you just wake up and once a week you get a deposit for $500 or something like that. No bureaucratic overhead. No waiting in lines. Just get your money, pay your bills, buy your groceries, and the economy keeps humming along.

I usually get nice comments like this. Sometimes I get nasty comments like “You’re never getting UBI you’re a delusional fuckwad!” (actually they are much worse than that, but I’m paraphrasing)

Stimulus checks were UBI, guys.

But UBI is no way to live. Sure, it’s better than nothing and many people would be happy with that. Stress-free even if it’s low-income. You could move out to lower cost areas and just garden, or whatever you want to do with all that free time.

Banking on Cognitive Surplus

More than a few people think that UBI, or Elon Musk’s UHI (Universal High Income) is a bad idea. They don’t want humans to get bored and lazy and entitled. We need some pressure, some fear of failure, and certainly fear of homelessness and hunger, to keep us motivated!

At least, that’s the Protestant value system. It’s also fire-breathing totally-unhinged neoliberalism.

“If workers are more insecure, that's very healthy for the society, because if workers are insecure they won't ask for wages, they won't strike, and they won’t push for benefits.” ~ Alan Greenspan, 1997, Chairman of the Fed, 1987 to 2006

So there you have it. Worker insecurity is state policy!

Now, let’s imagine that you lose your job and you can’t find any work due to machine labor just being so much cheaper. You have to sell your expensive suburban house because, well, you can’t afford it anymore and who cares since you’re no longer working in the city?

I call this the urban exodus, which is already happening from San Francisco and other high-dollar areas like Austin.

So you move out to the countryside where houses are 50% cheaper and now you’ve got some land, plus a bunch of free time. Maybe you plant a garden, get a goat, and life is good now, right? Perhaps you even finally pick up some trades you’ve been letting rust; finally learn to play that banjo or make fine furniture, or perhaps even restore old Mustangs?

If you’re like me, you use your cognitive surplus and spare time to write blog posts on the internet.

By the way, this is what I’m referring to.

So, how can define “cognitive surplus” in the context of I-just-got-replaced-by-a-robot and I’m-on-UBI-now-what?

Cognitive surplus is the collective intellectual, creative, and productive capacity that emerges when humans are freed from survival-focused labor. It represents the natural human drive to create meaning and value even when basic needs are already met. Rather than becoming idle when freed from wage labor, people typically redirect their mental energy and time toward personally meaningful pursuits—whether artistic, technical, social, or entrepreneurial—that often generate unexpected forms of social and economic value. This surplus manifests as everything from community gardens that feed neighborhoods, to restored antiques that preserve cultural heritage, to new art forms and innovations that emerge from passionate hobbyists with time to experiment and create.

Boredom needs an outlet.

My neighbors at my old house on the outskirts of Raleigh were an elderly Chinese couple. They’d immigrated from mainland China and, whatever they did, they were doing very well for themselves. One day, as I was cleaning up my yard, the man came over to the fence with a bunch of potatoes clutched in his hands. He didn’t speak English but his smile and gesture was enough to say “Here! Look at all the potatoes I grew! Here have some!” They were the purple Asian sweet potatoes that you can find anywhere.

There are non-monetary ways to add economic value to your community, and I suspect a lot of people will be happy to do engage in this lifestyle.

Poor Consumers, Bad Consumers

However, I suspect the powers-that-be will realize that keeping everyone on a subsistence wage that locks them into agrarian lifestyles is not really good for GDP. If you stop buying from Amazon and using credit cards, that makes you a bad consumer and we just can’t have that, now can we?

So what about Elon’s UHI? Universal High Income?

Theoretically, this is possible. If machines cause deflationary pressures in the factors we’re imagining, 1000x or more in some cases, then it’s entirely possible that even $2000 per month of UBI will go a lot further than it does today.

I recently wrote about solar sovereignty, a hypothetical future where, due to ongoing trends and the demands of AI, we’ll soon be generating so much solar energy that electricity will be practically free during the day.

So let’s break it down, what could this deflationary future look like?

1. Monthly power bill: $5

2. Healthcare: $10, plus labs and prescriptions

3. Rent: $500 (due to urban exodus)

4. Internet: $80 (not much change)

5. Food: Variable (you can just trade with neighbors)

“High income” might not be necessary, or it might cause inflation. Either way, the brains at the thinktanks will come up with policy suggestions to keep America (and other nations) “productive” whatever that happens to mean in the future.

Our very definition of “productivity” might change drastically. After all, if you no longer measure human labor as a significant component of GDP, then what do you measure? What are you optimizing the economy around? Neoliberalism optimizes the economy around “free market primacy” and economic agency for corporations. But if the fundamental social contract of wages-for-labor breaks down, who cares anymore?

I have already written about how I believe we need an investment-based future to shore up economic agency and labor participation. After all, if we’re all sitting around with our big brains and nothing else to do, why not allow the human hivemind to find more ways to add economic value?

But this is leading into the previous blog post, so we can call it a wrap today: