User:Three-quarter-ten/Deconflation

Deconflation

My brain has been bugging me about some conflations that need deconflating. The strange thing is that the brain can glimpse a simplicity that is difficult to convey below. Another person could look at the below and think that nothing simple was being explored. But that's just a specious appearance because communicating is challenging.

An object diagram would help. Of course, I am not a programmer. But the idea is relevant.

Begin:

Definitions

The term "old-definition_jobs" is used below to deconflate from "mirror-image_jobs". The word "jobs" alone, to date, means a concept that is labeled herein "old-definition_jobs".

Some premises

Premise 1 is that (automation may leave us without enough old-definition_jobs).

Premise 2 is that (there is a limit to the potential of how much work will be needed to be done).

Premise 2 is eternally false, I am confident. (This agrees with Martin Ford's statement that "it may well be infinite", regarding the amount of work that can be needed or desired). I am confident that it is indeed infinite.

You can declare that premise 2 is also called the (lump of labor premise) (alternate name). (Regardless of other ways that people may have defined the lump of labour fallacy, or could define it, let that one declaration stand for a moment, just for the sake of deconflation efforts.)

Premise 3 = (premise 1) because (premise 2): (automation may leave us without enough old-definition_jobs) because (there is a limit to the potential of how much work will be needed to be done).

Premise 3 is eternally false, I am confident.

Premise 3 is the classic econ 101 notion of the (overall Luddite argument). (Regardless of whether it accurately sums up what actual Luddites were thinking in 1811, it is the classic econ 101 notion of what they thought.)

Deconflation 1

Premise 1 and premise 3 get conflated in people's cognition in that both of them are labeled as "what the Luddites thought". But it is important to deconflate them, for reasons I will try to make clear.

So you can't alternate-name premise 1 as (Luddite premise) AND alternate-name premise 3 as (Luddite premise). They are different objects, so the name (Luddite premise) is poor for communication because ambiguous. Avoid using it here for the moment.

Some more premises, and deconflation 2

Premise 4 is (structural unemployment [SU]), which can be short-form-named (the mismatch premise). It is that unemployment can result simply from a mismatch between what employers need and what employees are able to provide. The main instance is that if employers need (high talent + particular knowledge), then most humans are not employable, because they lack those (and cannot acquire them through any *practical, affordable* means) (for example, you cannot simply teach someone talent).

Deconflation 2 is that, although the real spectrum is continuous rather than quantum, it seems practically necessary to dichotomize SU into acute SU and chronic SU, because it affects whether it makes any sense to assert that it is controversial to say that technological unemployment [TU]), which is (unemployment caused by technological change), can be defined as an instance of SU.

It is a mere truism to say that TU can be an instance of locoregional acute SU (and thus, to assert this cannot be controversial).

Why? Because all it takes for that to be true is that 1 factory upgrades its equipment with increased automation, and a production line that required 8 workers to produce 1200 widgets per day now requires 4 workers to produce 1500 widgets per day. This produces locoregional acute SU in the form of 4 laid-off workers in one geographic area who are temporarily out of work and whose next job will not involve the particular knowledge or skill of how to run the particular outdated equipment that got replaced.

Premise 5 is that (technological unemployment [TU]), which is (unemployment caused by technological change), can be defined as an instance of SU. That is only one of the ways that one could define it. But let it stand for the moment.

Premise 6 is that although it is uncontroversial that TU is an instance of locoregional acute SU, it *is* controversial to say that TU is becoming or is about to become an instance of systemic chronic SU.

Why? Because the latter has been the big argument about TU, about Luddism, about whether they are the same object, etc, for two centuries (1810s-2010s).

However, it does not seem logical that it could continue to be considered controversial for many more decades, because software, networking, and robotics, together constituting automation, are now pushing so hard in so many spots, dislocating so many pieces of so many business processes in so many industries, that many of the old-definition_jobs that get created require too much talent and expertise for employers to find people to fill them. This is already being noticed but seems likely to increase.

This strikes me as analogous to Lord Kelvin saying in the 1890s or 1902 that heavier-than-air aviation was a pipe dream. It was the end of the era in which the formerly conventional wisdom was credible. But he did not know that at the time; he was still peevishly disdaining the stupidity of questioning the conventional wisdom (which was soon to become no_longer_conventional).

Deconflation 3

The following is imperfect in terms of match from words to objects, but bear with it:

The Luddites thought the (overall Luddite argument), which was that (automation may leave us without enough old-definition_jobs) because (lump of labor premise) = true. (This is eternally false, because (lump of labor premise) = eternally false. [I am confident of those statements.])

whereas, in contrast,

New-market thinkers think that (automation may leave us without enough old-dfinition_jobs) because (mismatch premise). This can easily be true without invoking anything about (lump of labor premise).

New-market thinkers cannot accurately be called (neo-Luddites), because they are not new instances of Luddites. The (overall Luddite argument) is eternally false, and it is not what new-market thinkers are arguing.

Premise 1 can be true in some environments, simply because TU is a form of SU. This is the first thing that gets you in trouble with people who conflate premise 1 with premise 3, because premise 3 can never be true. And both get called (Luddite premise), conflatingly. So people think you are arguing that premise 3 could be true, but you're not.

Reformulate to another iteration closer to pithy natural language summation:

The Luddites thought that automation may leave us without enough old-definition_jobs because the lump of labor premise was true. (This is eternally false, because the lump of labor premise is eternally false.)

whereas, in contrast,

New-market thinkers think that automation may leave us without enough old-definition_jobs because we are approaching the constraint that most humans aren't talented enough to do the things that will exist as old-definition_jobs for humans in the future; and to do enough education and retraining to negate that will not be feasible. (This can easily be true without invoking anything about the lump of labor premise.)

Corollaries

Classic econ 101 is right that "new old-definition_jobs can always be created." But we will reach a time when the ones that are created will require amazing, scarce talent. And most people will be unable to supply that.

A separate branch under premise 2, but not to be digressed into until after the main circuit is followed

Simultaneously, it is important to understand that, in addition to the work being infinite, machines will be able to do most of it. (Martin Ford said the same thing, too, in the same paragraph.)

This is a pinch point where people will want to dispute the argument. As Martin Ford said in the same paragraph, you get accused of recommitting the lump of labor fallacy all over again by saying "machines will be able to do most of it". But that accusal is, in itself, a logical fallacy, based on conflation of objects (premises). You must deconflate what you mean by (labor). If by (labor) you mean (human_labor), then you are not using the word sense that M.F. was using. If by (labor) you mean (work getting done by an agent), where (agent) = a human, a machine, or a combination of those (a human using discrete machines, or a human-machine blend, or a human-machine blend using discrete machines), then you are using the word sense in which there can never be any "lump" of labor (labor is infinite).

The implication of (even though there is no "lump" of labor [no limit to it], machines can do most of it) is that you can create a large amount of wealth with a small amount of human labor. This latter concept is not science fiction; we have already seen a trend over decades where human labor becomes less central to value creation, and capital more so. If that trend flows far enough, it creates the condition for something in essence equaling a postscarcity economy, although without the utopian polyannaish drug-trip-fantasy element that today often accompanies that phrase. Which is yet another conflation that does a disservice to people trying to communicate and figure things out based solely on clear-eyed logic that never forgets the cruel lessons of the real world.

The way I would encapsulate the latter theme is that new-market mechanisms are an intermediate stepping stone (or a bridge, in a related but modified metaphor) that can get you from today's economy to a postscarcity economy. They start you out with a minority portion of the latter and allow a chance for it to grow into a fuller fraction. People skilled in chemical synthesis are familiar with that concept as it applies to progressive distillations and protecting groups. (Metaphor-mixture whiplash yet? It doesn’t hurt for well-conditioned athletes. Simply recognize that all analogies provide touchstones for touch-and-go, and we soon go, because they all have limits. We don't stay; we touch glasses and move on.)

But a lot of people won't want to hear any of this, because it sounds too much like a failure waiting to happen. All of us rational, responsible people here in the 21st century have learned the lessons of the 20th century; we know damn well, from real-world datasets of hard-earned experience, that there ain't no such thing as a free lunch. Consider the general theme of "doing something half-decent today and working slowly over decades toward something even better." One of the instances of that general theme is the American Dream. But another instance was when Leninists talked about "building communism" to arrive at something glorious eventually. We all know that the latter failed. And anyone who doesn't understand what you're trying to talk about but takes a sniff and suspects something like a "building communism" instance is going to immediately shut down, devalue, and dismiss. They're not listening anymore. Stop talking. Go away. What the hell? Do we need to get cops and soldiers out here to protect us from crazies and pinkos? To get you to stop talking about dangerous nonsense?

That's why, to me, it seems so important to understand, and to reassure others that you well appreciate, that this bridge has nothing to do with utopian fantasies. It is only a practical mechanism to avoid massive SU that strangles an economy. It is not an attempt to find paradise; it is merely an attempt to avoid an abyss. And what the new-market thinking reaches toward is a realistic thing that maintains incentivization (work for pay) and that has no illusion of "luxurious wealthiness for everyone". It has no illusion of "luxurious wealthiness for everyone" at all, let alone "luxurious wealthiness for everyone, for free". Really, all it is, is a way for "most average people" around the world to have what was considered middle-class comfort in the U.S. during the 1950s through 1990s. Not luxurious wealth, just middle-class comfort. Now, you could spout off blue-sky theories about how maybe it could eventually lead to a full postscarcity economy, which could imply "luxurious wealthiness for everyone"; but to me that seems an irrelevant distraction and a flamebait. That's not a conversation we're trying to have, and having it is just a waste of time and cognitive bandwidth during an era when you can't even get most people to deconflate premise 1 from premise 3, or Luddism from new-market ideas, or locoregional acute SU from systemic chronic SU, or old-definition_jobs from mirror-image_jobs, or mirror-image_jobs from old-definition make-work (spoons instead of shovels), or labor meaning human_labor from labor meaning work_done_by_an_agent. There is so much deconflation to be done, challenged by the discrepancies between the object diagram and natural language as she is usually spoken; and this fact is where the "mirror image" name came from in the first place. Forget utopias or allegations of possible utopias. It's irrelevant during this era. What are we going to do to employ real-world average people inside the next 3 decades when manufacturing, packaging lines, truck driving, taxi driving, package delivery, shelf stocking, legal discovery, transcription and translation, radiology, omics, and a hundred other categories of work quickly go from 5% automated to 50% automated (or from 1% to 90%, depending on instance; or whatever)? How to avoid a spike of SU, in other words? And how are we going to talk about the relevant topics, given the word sense conflation problems? As Ryan Avent paraphrased what various people have been saying in the past few years, "Maybe we need a new system for talking about how these things work."

Speaking of dichotomization

The distinction between labor and capital

parallels

the distinction between humans and machines;

but as the latter two get combined in endless permutations, you may find that you have a continuous variable that you are attempting to dichotomize.

Which means you are also facing a continuous vs dichotomized variable instance with the former two.

I am pretty sure I have heard this theme before, but I don't remember specifically where, and today I re-derived it and appreciated it anew.

BTW, the word "cyborg" is another devalue-and-dismiss trigger, because your listener is listening to you talk about the real world in the context of up-and-coming timeframes, and now you just used a science fiction word, so now you're an idiot that needs to be ignored. But the problem with that is that I think IT-augmented humans are coming up pretty soon. Prosthetics + implants + robotics + neuroscience + augmented reality + powered exoskeletons (all of which are decently advanced even now, compared to 20 years ago, and are only going to advance more over the decades) = augmented people, sooner rather than later. Yes, of course, it's going to start small. No, of course, they're not going to be full-fledged Terminators anytime soon. But I'd be surprised if your kid isn't competing with an augmented human for a job interview within 30 years. Uh oh, now we have to compete for jobs with that, too? It's a good thing you know better than I do about all this TU stuff being bullshit, right?