Henry Dampier

On the outer right side of history

by 22 Comments

I Don’t Believe In the ‘Automation Eats All Jobs’ Thesis

…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.

  1. 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.
  2. 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.
  3. 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.

by 18 Comments

The Ludic Fallacy

Nassim Taleb introduced the term ‘ludic fallacy’ in Black Swan. You may think that you know the contents of the book from the press coverage even if you haven’t read it, but my suggestion would be to read it even if you think you know what it’s about — the coverage tends to be misleading.

If you have the book, you can flip to chapter nine and review it before you read the rest of this post.

The ludic fallacy is mistaking a model, especially a model of human behavior, for the real world. It’s partly intended as a caution against using game theory, economic theories resting on faulty assumptions of humanity (‘homo economicus’), and other artificial environments.

Taleb contrasts two characters: Fat Tony and Dr. John.

Basically, Fat Tony always looks for an angle and never expects fair competition. Dr. John expects the world to conform to the models that he has picked up through his studies, and anticipates that the world is fair, much like a game of chess in which both sides start out with almost perfect symmetry.

Dr. John thinks that his plan will work. Tony expects to be punched in the face and instead trusts more in his ad hoc assessments and empirical experience.

Taleb writes “A nerd is simply someone who thinks exceedingly inside the box.” America elevates these nerds into high positions because America is a place that produces boxes upon boxes upon boxes upon standardized boxes, and does its best to mold its territory into a series of rationalized, predictable boxes. Yet nature dislikes being confined to such boxes and strains against them, despite the nerd’s attempts to eliminate unpredictability.

Even in highly managed environments, like the casino Taleb uses in this chapter, can be struck by difficult-to-predict catastrophes — the largest loss incurred by the casino in question, of $100 million, happened when Roy of ‘Siegfried & Roy’ was mauled by his white tiger, despite decades of placid behavior by the great cat. The casino was robust against card-counters and cheaters, but vulnerable against a shock from an unforeseen angle.

The desire to make the world predictable can blind people to reality:

It is why we Platonify, liking known schemas and well-organized knowledge — to the point of blindness to reality. It is why we fall for the problem of induction, why we confirm. It is why those who “study” and fare well in school have a tendency to be suckers for the ludic fallacy.

Taleb suggests that you denarrate — disconnect from media, including blogs, in part to train yourself to spot “the difference between the sensational and the empirical.”

He expands on the concept in Antifragile, particularly in Chapter 16, when he compares the ‘ecological and the ludic.’ Football is a game. War is not a game. But people often try to draw lessons from football to apply to domains like war and business. Similarly, business is governed by laws, but war usually operates under an entirely different set of laws which may or may not be enforceable.

From p. 241:

It is not well advertised that there is no evidence that abilities in chess lead to better reasoning off the chessboard— even those who play blind chess games with an entire cohort can’t remember things outside the board better than a regular person.

And later:

Provided we have the right type of rigor, we need randomness, mess, adventures, uncertainty, self-discovery, near-traumatic episodes, all these things that make life worth living, compared to the structured, fake, and ineffective life of an empty-suit CEO with a preset schedule and an alarm clock.

The way to gain an advantage over nerds is to just take advantage of their predictability.

If you know the model that they’re operating under, you can generally predict how they will blunder, and then position yourself to gain an advantage as they stumble. They can’t handle volatility because it breaks their mental models, which makes them panic. A sucker punch sends them scrambling to try to make their world predictable, solid, and box-like again.

The progressive project primarily works by conflating their models for the underlying world. Their priests believe that by changing the models, they can change the underlying reality. Because so many of the people who most fervently believe in the writ of progress live almost entirely in a model world rather than the real one (hence the Starbucks full of people staring into screens), they tend to be oblivious to changes in the natural world.

So, the focus ought to be on what can be done to de-Platonify life, rather than launching endless new counter-models against the ones who insist that everything can be rationalized.

by 13 Comments

Information As Controlled Substance

The FCC is assuming regulatory control over the internet amid worries from across the political right that this will lead to politicized censorship.

Considering that every other medium that the FCC has assumed control over has come under politicized censorship, that result would not be that surprising. I’ve already expressed my opinion about whether or not I think this is something to worry about, but I’ll expound on that some more here.

While the current move is purportedly about ‘fast lanes,’ the more important detail is the expansion of the agency’s authority. This will make it much easier for new regulations to go through which impact content — like the reviled ‘SOPA’ and ‘PIPA’ which astro-turf activists paid for by the tech firms believed that they had delayed.

The hilarious part of this is that by attaching the FCC’s expansion of authority to the long-beloved-of-nerds net neutrality issue is that foundations like the EFF were happy to support it, despite it inevitably leading to the same sorts of regulations that it has agitated against for years.

We don’t know how quickly the FCC will proceed in regulating internet content. We don’t even really know the details of what the current plans are. My guess is that the regulatory administration will take a gradual approach. The Republicans will do what they can to serve the interests of the ISPs, but won’t care all that much about the content side of things.

“Regulate as a utility” is a phrase that makes leftists very happy, because it means political jobs for the orcs, trolls, and goblins of their coalition. They use this particular phrase with banking, also — especially commercial banking.

If you work in or around tech, and you are at all familiar with what working in commercial banking is like, the thought of that sort of Soviet business culture should make you want to raise the black flag and cut throats.

This will probably be bad for the corporations that elected to lobby for it — particularly companies like Google that will contract and die without a growing web economy. They are shooting themselves in the balls by handing over the keys to the FCC.

The unregulated internet is the smartest thing that Robert Rubin and Bill Clinton put together during their administration. It even created a new class of billionaire supporters for Democrats. Blowing that up is more stupid and self-destructive than it is conniving and Machiavellian. Give me well-executed conniving any day over looming Zimbabwe-ism.

On the content censorship side of things, it just means that accurate information will become a controlled substance, and most of what will be sold by licensed providers will be either completely inaccurate or misleading. So, not that much different from the way it is today, but with less legal competition and more illegal competition.

Information is much more important to the ordinary functioning of society than drugs are, so what you would expect to see is for people to start hunting for more real-talk, because it’ll be more obvious than ever that the information they’re getting is not accurate.

In anticipation of this, it’s time to get started on building informal and formal networks which interpret FedNet as damage and route around it.

The solution to the problem will be some part human and some part technical. It wouldn’t be surprising to see more regulation of computer hardware, also, as tensions between  nations regarding snooping devices on consumer hardware are becoming more acute.

So the solutions will have to be partly hardware, also, since it’s possible that we’re going to see more locked-down devices which break rapidly and get filled with lots of spying cruft — this should be easy to see just extrapolating from current trends. Your iPhone will only connect to FedNet, and it’ll break every few years. It’ll be a gradual change that most people won’t notice.

On net, FedNet will be terrible, and it wouldn’t be surprising to see the international nature of the internet become impeded by its establishment. The plus side is that the general crisis which it precipitates will also bring with it a lot of opportunities, especially as the informational infrastructure that society has come to rely upon breaks down.

What can’t be stopped is the regulatory move. What can be done is to adapt effectively to the new legal regime.

Top Posts & Pages