…and I will probably be quite lonely in that belief, which I am confident is true.
People who state that automation will replace human labor entirely tend to be stock pumpers for the California tech firms who have become a little too proud for their britches, or are otherwise beguiled by said stock-pumpers. As a sometimes stock-pumper (all within the law), I am cynical about flimflam world-changing prospectuses, especially when they come from California, where they probably put LSD in the water supply.
It will take a long time to make this argument to my satisfaction, so this is the tentative outline for my position.
- Value is subjective.
- Humans set the prices for labor, because machines have no independent judgment or desire. A machine by itself generates no demand.
- Prognostications about artificial intelligence are entirely premature.
- Machine learning is not artificial intelligence.
- The people who conflate machine learning with artificial intelligence except in a metaphorical way generally have jack-all to do with actually working in machine learning.
- If machine learning did not require massive human input all the time, companies like Google would be entirely autonomous, instead of requiring an enormous staff of people and contractors to tweak the outputs of their machine learning algorithms constantly.
- The human labor force to achieve this is actually quite massive. It’s just not entirely contained within the walls of the Mountain View campus.
- Similarly, most of the humans involved in building iPhones do not work out of Cupertino. They’re in China. And it’s a lot of manual labor, some involving machines, some not.
- Anyone who has worked on these projects can tell you at length how much is automated, and how much isn’t, and how any given improvement algorithm is often trumped by large numbers of semi-retarded people looking to exploit the predictable algorithmic outputs for fun and profit.
- Again, the ‘machine learning’ revolution in Silicon Valley is much, much more labor intensive than it appears on the surface of it — competition in manipulating all the algorithmic outputs also creates competitive races that suck up more and more labor and budget for companies throughout the economy.
- Because value is subjective, there are essentially as many potential jobs as there are feasible human desires.
- These jobs can take many strange forms.
- Ex. Today, we have ‘dog bakeries’ in upscale neighborhoods which serve fancy treats to dogs.
- Much of what makes up an economy is not based on ‘needs’ but ‘desires’ which are often arbitrary and in continual flux.
- Beanie Babies are a hot item until they aren’t
- Open-toed heels are hot one year, then it’s knee-high boots.
- Netbooks are the next big thing, and then it’s tablets.
- None of these things are ‘needs’ — they are all more like ‘whims.’
- In the past, the nobility and middle class employed servants to dress them in the morning… or even to just watch them get dressed.
- That jobs may be demeaning, hold low social status, or are only feasible and sub-minimum wage or indenture levels do not make them no longer valid social roles.
- These jobs can take many strange forms.
- Humans set the prices for labor, because machines have no independent judgment or desire. A machine by itself generates no demand.
- Automation has limits.
- You’ll notice that products produced by mass manufacture are of inferior quality to artisan products.
- Artisan production requires long training periods and cannot scale to mass production.
- The New Deal and related labor legislation going back to the abolition of slavery has interfered in the creation and enforcement of apprentice contracts which facilitate artisan training.
- Our legal system tends to favor mass production and penalizes artisanal production
- ex. the FDA will send military raids against your raw dairy farm, but may even subsidize a mass-production dairy farm
- The DOE will send agents to shoot you if you want to put together a startup power plant.
- The states and the Feds tend to oppose alternative electricity distribution systems which would be more amenable to decentralized industrial production.
- You are forbidden from starting a factory without going through lots of red tape, harming the ability of small-run, more artisan production to happen.
- If all the good and beautiful progressive people believe in something, it is probably a lie.
- This thesis is popular because it provides an excuse for substantial unemployment throughout the Western world.
- The unemployment is in large part caused by the Western tax-regulatory regime.
- The good and beautiful people do not want to reform the regulatory regime…
- So they invent a scapegoat called ‘automation’ which has been routinely used since even before the invention of the assembly line.
- This is conflated with science fiction visions of hostile artificial intelligence.
- The boogieman of artificial intelligence, which is nowhere near reaching maturity (especially as limits to computer hardware enhancement are appearing), is used to justify welfare programs
- Degree-bearing bums who want a handout are gleeful for an intellectual-sounding acronym to use to beg for a handout with (‘UBI’)
This is the skeleton of the argument that I want to flesh out. It’s also cribbed from the first book project which I temporarily iced built off of this post.
None of this meas that automation is not a major actor driving market behavior. It is just that dogmatic statements that it replaces all forms of human labor are intensely speculative and should be held up to withering scrutiny. My position is not Krugman-esque “internet has no more impact than the fax machine” drivel.
Further, if you disagree, that is fine: I don’t have to convince you, and I don’t even necessarily want to. I just have to profit from your shoddy thinking about technology.
Underlying much of these prognostications about automation is the labor theory of value. To the extent that they believe in that theory is the extent to which they believe that automation alone ‘destroys jobs.’ There are many more factors at play in what goes into the market for human labor.
What is needed is political confrontation. People frightened of direct confrontation with prefer to write up exculpatory narratives about why they don’t have to pick up a spear and point it in the right direction. People tend to be bamboozled easily by technocratic jargon about technology that has absolutely nothing to do with how those technologies really work.
There is a great deal of public panic about technology that ought to be smothered, rather than encouraged. Understanding the limits of both existing technology and foreseeable is not something that most people are willing or eager to do. Panic or surrender to ‘inevitable processes’ is both easy and effeminate.
Instead, it’s better to form a more accurate perception of labor as it relates to technology, human society, and the laws that have been erected to impact what sorts of labor arrangements are legal.
As one of the pioneers in programming and programming languages, Alan Perlis, put it a few decades ago: “If you can imagine a society in which the computer-robot is the only menial, you can imagine anything.”
I do enjoy your writing, but this seems like a lot of words that could’ve been saved by simply saying, “Read Mises, retards.”
Libertarianism’s failure in other arenas doesn’t change the fundamental facts of economic reality.
No, it doesn’t.
But telling people that they’re retards doesn’t compel action, even if it’s true. It actually takes a lot of effort to get people to read, period, even people who are more than intellectually capable of understanding the material.
Even professors with authority who have a class of students tested as the highest-performing in the country have trouble getting their students to read anything assigned to them. I have much less authority than even such a professor.
It’s also harder to make this case because a lot of people with status and titles huff and puff a lot of nonsense over this particular topic.
So, I have to spend inordinate effort establishing the argument. This took me roughly 1h to write this morning, but more time to think about in bits and pieces.
I used to work in a factory on a production line and I guarantee you machines cannot do the things we did. When something goes wrong – and it did, a lot – the machines would have to be intelligent to fix the problem. And that’s never going to happen.
I tend to agree with you about the all the automation doom-saying being a mostly lot of hot air, but there is a very real potential for even basic 21st century automation technology to seriously disrupt a lot of lowish-skill job sectors. For example, and this is firmly on the mundane end of the spectrum of possibility, some agricultural jobs (crop planting, soil conditioning, harvesting) could be replaced with industrial robots, and clerical/retail jobs could be replaced with machine kiosks. Both of these are already happening.
The restructuring and labor reassignment processes associated with these economic changes make ordinary people apprehensive, and elites gleeful. The ordinary man asks: What will I have to learn do tomorrow in order to support myself and my family? Will I still make enough money to live the way I live now? Will I have to relocate? In contrast, the elite man asks: How much will I save in overhead costs? How slick would it be to be able to interact my computer like in “Star Trek” (or perhaps, “Her”)? Technological changes affect different groups in different ways. The ordinary man is swept up by the current, the elite man sails any direction he chooses on his jet-propelled hydrofoil.
However, the whole process is slow enough that everyone generally can recognize the trend and then act in anticipation of the coming changes. Ultimately there is no big shock to the economy, just a gradual transformation. And this is, as you touched upon, not really any different an animal than previous eras. Things change, life goes on.
Sure. I would not tell my kid to go be a truck driver or a journalist. I would encourage the kid to screw around with building and scripting robots.
As you already know, automation is wiping out low skill jobs. The few that remain will be going to newly “legalized” illegal immigrants. This will drop kick what remains of the Black lower-middle class right into the growing underclass. Black liberals (99% of Blacks) are insane for buying into the rich White Liberal agenda of flooding the country with Third-Worlders in the name of “Social Justice”. They are being totally conned. They are firmly sealing their own fate and that of their children.
This seems to be the common argument in the Sailersphere. I disagree on some small points. I probably have too expand on them.
Convincing points. Only disagreement would be that all demand is based on “whims.” This is a feature of postmodern decline. Sure, all people have whims, but it completely spiraled out of control in the last few decades.
Also, I’m new to the twitter, but recently found saw a post by “reactionary CEO” that’s pertinent: “1965 Engineer: I’m working on a rocket to put a man on the moon. 2015 Engineer: I trick people into clicking on ads.” This seems to be the direction of automation, which gives me the impression of a net negative.
UBI may be inevitable to delay Planet of the Apes style uprising. Its likely cheaper than thousands of riot cops/FEMA camps/billions of Xantax bottles. When people have nothing to do and no guidelines they just go bananas.
Religion and morality have been incapable of keeping up with technology.
>Convincing points. Only disagreement would be that all demand is based on “whims.” This is a feature of postmodern decline. Sure, all people have whims, but it completely spiraled out of control in the last few decades.
That is just me rephrasing the subjective theory of value for rhetorical effect.
I also disagree that it is postmodern. Most people used to live on thin gruel, packed into small huts. In Latin America and elsewhere, you can still see peasants who live 6 people to a mud hut without a chimney with a floor coated in guinea pig droppings. They die young from the long term effects of smoke inhalation.
>Also, I’m new to the twitter, but recently found saw a post by “reactionary CEO” that’s pertinent: “1965 Engineer: I’m working on a rocket to put a man on the moon. 2015 Engineer: I trick people into clicking on ads.” This seems to be the direction of automation, which gives me the impression of a net negative.
I don’t think this was his intention, and I don’t think you’re interpreting what’s going on entirely correctly.
Keep in mind also that ad-driven technology is in the midst of a serious speculative frenzy right now, with p/e ratios far beyond what is historically appropriate or justified by the fundamentals of the companies in question.
During the late 1990s bubble, the ‘new’ advertising technologies were actually rather inferior relative to the older technologies.
Adwords is and was a legitimate innovation that meaningfully managed to control the click fraud problem. Facebook, Pinterest, blah blah blah, are all still in a speculative phase, along with many other ad-driven firms, and could be blown out of the water by a crash.
There are enormous technical problems within the advertising-technology industry right now, similar but at a more pervasive scale, to the ones that brought down the former technology bubble. This is not exactly concealed — the problems have been written up in the NY Times, Adweek, Ad Age, and every other publication of note.
So indeed, it may be rather worse than our friend thinks. Many engineers may discover that their careers for the last 10 years or so have actually delivered almost no value to the economy, and have destroyed a great deal of value.
That is what ‘malinvestment’ means in the concrete.
>UBI may be inevitable to delay Planet of the Apes style uprising. Its likely cheaper than thousands of riot cops/FEMA camps/billions of Xantax bottles. When people have nothing to do and no guidelines they just go bananas.
GWB2 already made access to Xanax a human right with Medicare Part D. UBI was essentially the original socialist plan for social security, from back in the 19th century. It grew from that. UBI would not stay UBI for long.
>Religion and morality have been incapable of keeping up with technology.
I don’t think this view is correct. It is the consensus understanding of history, though, so you will find a lot of people who will agree with you on that point.
The “consensus understanding” seems to be that religion became obsolete like an old iPhone, just as the concept of labor as a virtue becomes obsolete for the unemployed class as menial labor is replaced with corporate bots. The mentality of obsolescence also lends itself to completely discarding things perceived as outdated, which include not just techne or practices that may be inefficient but persist due to tradition, such as winemaking styles or shoemaking, but it also tends to render “obsolete” entire ways of viewing the world. Interesting how about a week after a “three parent” child was concocted in a lab there was a “three way” gay marriage.
Regardless of where it came from, I simply mentioned UBI as a practical form of urban pacification for a liberal government that has few other options with hordes who are unemployable and see themselves as entitled.
If automation is all based on bots manufacturing/selling beanie babies 2.0 to UBI recipients with Amazon accounts, then that would be another example of how counterproductive and morally destructive it can be.
Also, AI does not have to be “sentient” to be a concern, machine intelligence is just as dangerous. For example, stock market flash crashes are likely a presage to what governments and money-changers will do to have their way.
I’m not arguing against all automation, but the attitude that anything invented is “good” doesn’t make any sense.
Just a hedge for the below I generally enjoy most of what I read and if I post with a slight disagreement it is not because I am not a fan or something. I quite enjoyed “The Way of Men” and the majority of your posts.
The people who are saying automation might eat all the jobs are just manufacturing this centuries new waves of hysteria and probably are not worth paying any attention to.
I’m interested in why you think California tech firms are too big for their britches ,most of us are surprised about what we even did to earn such specific dislike. People who live here have less opinions on tech firms than people elsewhere it seems.
Andreesan says to be a tech executive you need to be in the top 25% of at least two fields or top 1% in one. I’ve probably been studying 12 hours a day for more than a year and I’m afraid if I don’t make the most dazzling technological improvement this century people will hate me for no reason other than that I’m “One of those tech people”.
Price per earnings for technology companies is pretty important but most of the earnings for ventures come in the middle of technology companies lifetimes and not the start like is usual for traditional companies. There is a huge risk for a “This time is different?” sort of argument but I’ve seen Thiel make the same argument in his recent book. It takes a while to develop a technology and then get it out of the door.
The rocket vs ads argument is ridiculous of course.
>I’m interested in why you think California tech firms are too big for their britches ,most of us are surprised about what we even did to earn such specific dislike. People who live here have less opinions on tech firms than people elsewhere it seems.
Generally I like techies, have written in their defense, have enjoyed working with them, and have derived almost all of my income since I began working from the wonderful innovations created by California and Seattle tech firms.
Their political ideas are almost entirely pernicious, insane, and left-wing. I am often sympathetic to what Peter says.
>The people who are saying automation might eat all the jobs are just manufacturing this centuries new waves of hysteria and probably are not worth paying any attention to.
Marc has done a lot to promote this line of argument. Not to take him to task for this more than is warranted.
They are generally great when speaking on their area of expertise. When they speak outside their area of knowledge, they tend to make basic mistakes, but because of their wealth & influence, few are able to point it out effectively.
I agree with you. The most horrendous job destruction in human history was probably the automation of farming. We went from >50% of the workforce to much less than 2%. Computers did away with secretaries. Appliances did away with housekeepers and maids. Already few work in manufacturing anymore. We probably already swallowed the largest chunk of automation.
There is a demographic issue at play in the job market. For instance, my family could use childcare help, but a high proportion of the ‘available’ workforce would be positively detrimental to our children. What proportion of the population can’t be trusted even to work in retail? Libertarians and most economists like to assume that labor supply can be cleared at the right price, but what if due to social factors that price is negative?
NALALT, but it’s an accurate generalization of people like Bryan Caplan. People are not equal, and what moral values the workforce holds also matters.
Admins are still important to handle e-mails. A secretary (admin assistant) can handle 100s of e-mails and phone calls a day either for one person or a team.
Obviously when farming went away, it was mostly replaced by farming, mercantile labor, finance, office work, etc. etc. — much larger displacement than what computers did.
Henry, I agree with your points. I’ve been researching this topic for a graduate economics research paper. I would love some more ammo from free-market intellectuals, but most of the interest in this topic comes from lefties (for the reasons you touched on). Please share any other resources that may be able to inform my paper.