Computers are starting to learn, and they're going to learn our jobs

The breakthroughs are coming faster and faster. Last night (as I started this post) Google's go-playing computer system, called AlphaGo, beat one of the the world's best go players, Lee Sedol of Korea. Tonight (as I continue it) AlphaGo has won the second match in a five game series. This morning (as I work on it) AlphaGo has won the third.

This is a big, big deal because AlphaGo wasn't programmed to win at go. It was programmed to learn how to win at go.

It's a huge step toward the day when computers will able to take a huge number of jobs that right now only humans can do. Those jobs are not coming back. And we're not going to get good new ones in their place.

When that day comes it will be great, and lousy. Great because we'll all be able live in unheard of abundance.  Lousy because, there will be no paying jobs for some people. And despite the fact that will be abundance enough for all, some people will argue -- on moral grounds -- that people with no jobs deserve nothing; and then they will act -- on pragmatic grounds -- to give those people just enough to keep them from breaking out the well-deserved torches and pitchforks.

Yeah, I know. People have been talking about automation throwing people out of work since -- well, since automation first started throwing people out of work. And the people who have said that have always been wrong. And I've told them that. New jobs, even better jobs, have always been created. Read about the Luddite fallacy. It's called a fallacy because it's false.

But this time I'm saying something different. There's an exponential growth pattern that's hiding in the data. Funny thing about exponential patterns: they look like an innocent, slightly upward curving lines. Then FOOM! They take off..

Two things will be different. First, old jobs will vanish far, far, faster than before. In the past, new jobs could be created nearly as fast as old jobs were destroyed. In the past.

Second, most people won't be able to do the new, good jobs; they will require more IQ than most people have.

Let me break it down.

Old jobs are destroyed at the rate at which they can be automated. If you dramatically increase that rate, then old jobs will disappear faster. The rate has been increasing, and will continue to increase. Alphago and its ilk are part of the reason I expect the increase to increase.

AlphaGo is a computer system that's programmed to learn how to do things. It's not an artificial intelligence. We're far away from that. It's a machine learning system. At the highest level machine learning systems have to be programmed to learn. Below that level they learns. How? By guided learning. By seeing examples. By experimenting. By trail and error.

The learning algorithms, are written by top programmers. At the peak of the machine learning food chain, where Google's AlphaGo and DeepMind feed, the programmers are PhDs from top universities and their less decorated peers. Once upon a time I was somewhere in the top 1% by IQ and SAT score. Maybe in the top 0.1%. I was a solid programmer in my day. But I don't think that even in my prime that I would be in that league.

So there's a new kind of job -- programming systems like AlphaGo -- that I would not have been smart enough to do. And if I wasn't smart enough then most of the people in the world wouldn't have been smart enough, either.

Now consider the jobs that the rest of us can do, and here comes the problem. Once a "learning computer" is capable of learning how to do a certain class of jobs, then automating that job out of existence is easy. You just need enough computers and enough training resources available.

Enough computer resources? Plenty of those. Growing at an incredible rate. And enough training resources? That's us.

Take driving a car. Every time a Tesla customer takes his Model S out for a spin, whether or not they've engaged Tesla's "autopilot," they're teaching Tesla's learning computer system how to drive better. In a particular driving situation, the Tesla computer might decide not to change lanes. But if, in similar situations, most drivers change lanes, Tesla learning computer is going to modify its decision making so that it will decide, in similar situations, not to change lanes.

Or understanding human speech. Every time we use Apple's Siri, or Google's Now, or Amazon's Alexa, we're teaching it to improve its understanding of human speech. The way we organize our online photo collections helps teach them how to recognize features in images. And so on.

As we use online services to omake our lives better we're teaching computers what they need to do so that they can do jobs that only people can do today.

Once we've got cars that can really self-drive, what happens to truck drivers? To taxi drivers? To others who drive for a living. What happens to people who build cars when anyone can call for a car (not an Uber car with an expensive driver attached, just a car) and have it take you where you want to go? You don't need that second family car that you use only occasionally. And you don't even need that first one, either.

And what happens when companies that want a computer to do something don't have to hire high-priced programmers to program it? What happens when they can show a learning system what they want done, and the system figures out what the program has to be?

What happens when factory robots don't have to be programmed -- at vast expense -- but can just be shown what to do? Or given some general guidelines and the time to figure it out themselves -- in parallel. Here's a bunch of Google's robotic arms solving problems of "hand-eye coordination," by teaching one other.

What's it cost to buy a robotic arm that a computer could learn to control? Here, and here are recent kickstarters for light duty robotic arms. Here is another one: for $350.00 you can get a six axis robot arm that can move, and pick up things, and it can see -- and you don't have to program it. You can guide its movement, or move your hand and show it what you want it do do. Good buy simple manual jobs.

That's today. Robot arms keep getting cheaper and better and smarter as their underlying components keep getting cheaper and better and smarter. It's relatively easy to scale them up to carry more weight. And if a robot arm knows how to do something, then thanks to robot-information-interchange projects like RoboEarth, once one robot arm can do something, then every suitably capable robot arm can do that same thing.

Human learning capacity is limited. A human worker might learn how to do fifty or a hundred or even a thousand discrete tasks. It might generalize its knowledge and be able to figure out a bunch more. But any robot, connected to the Internet, can do as many tasks as  -- How big is the Internet anyway? Yeah, that many tasks.

Will there be good jobs in the future? You bet, but mainly for people who are at the top of any talent category. Being at the top of the smartness category is just very important special case. People who have the right intellectual skills but not at the very top will have jobs for a while as they use the components created by their even smarter brethren to build systems that put other people out of work.

People have been putting other people out of work for a long time. It's what we call progress. Remember the Luddite Fallacy? Why is this any different?

In this post, Scott Alexander offers a useful metaphor: "an employment waterline, gradually rising through higher and higher levels of competence." If your competence is above the waterline, you survive. If you go below then you drown -- unless you are supported by others. The water is rising. And fast.

The Flynn Effect tells us that people actually are getting smarter. But they are getting smarter slowly, and the water is rising fast.

In our hunter-gatherer past, almost everyone was above the waterline.

As we moved to an agricultural society a very few couldn't learn the rules of farming, and started to fall went below the slowly rising (talking thousands of years) waterline. They were few enough that if there was enough food to go around, other people were willing to help them survive.

As agricultural societies became industrialized, the waterline continued to rise above a few more. And now we're talking changes taking hundreds of years, not thousands.

Now the West is considered post-industrial. We're a knowledge and services economy. You need a increasing amounts of training to make your way. And I'm not talking about a college degree. The amount of stuff you have to know to be a competent plumber or welder today is growing. Yes, thank goodness, we've got the Internet to help. But the water is rising.

I see many bad paths ahead for most of us. Not for my kids and my grandkids, thank goodness. All of them are wicked smart, and it will be a relatively long time before the water rises high enough to threaten them. But I do see bad paths ahead for lots of perfectly good human beings.

I see a couple of good paths. One is to change our ideas about how the economy works and abandon the old ideas rooted in scarcity, and adapt them to a world of abundance.

Another might be even better, but while it sounds incredibly exciting it's also incredibly scary. A new technology called CRSPR offers a low-cost and highly reliable way to edit the human genome in favor of intelligence. Steven Hsu a brilliant physicist and writer explains how an IQ of 1000 might become the new normal in his essay "Superintelligent Humans Are Coming."

Oh brave new world, that has such creatures in it.

Trump: extremely interesting and very scary

My morning Internet reading lead me on the usual scavenger hunt for knowledge. 

I came away with new respect for Donald Trump.

Who is Donald Trump, anyway? What could we expect if he won the election? I think I have an answer.

This is how I found it.

My hunt led me to a greater understanding of "the Trump phenomenon." Really, I didn't get it. I never thought  that Trump supporters were stupid, but I couldn't see what was so attractive to so many people. Even to some really smart people. Now I think I get it.

Dilbert creator and blogger Scott Adams provided the key insights. I found his posts following a trail that started with this post by PhD high energy physicist, martial arts expert, entrepreneur, Michigan State University VP, and blogger, Steven Hsu.

(Yes, apparently you can be all of these things, if you are as smart at Steven seem to be. He is on G+ and his blog is here. Worth reading his blog and following him.)

The  post included a clip from a Reason.tv video interview with Scott  Adams. (If you want to read before you watch, you can see a commentary, the video and then see a longer version here.)

Scott Adams is a genuinely funny guy, an insightful writer, and a also a trained hypnotist. It's the hypnotist training that gave him insight into Trump and caused him to predict Trump's success last August. In a post titled "Clown Genius" he described -- with some admiration -- techniques that Trump used. He wrote a continuing follow-up posts, and even though he explicitly said he didn't know who he thought would be best, his posts ended up being taken as an endorsement.

To make matters clear, after the David Duke dust-up he wrote a post disavowing Trump. But he continues to write posts admiring his skills.

Disavow? Kind of the way that Trump disavowed David Duke? "I disavow." Or for real. I don't know. I can't tell.

In the "Clown Genius" post Scott says:

Like many of you, I have been entertained by the unstoppable clown car that is Donald Trump. On the surface, and several layers deep as well, Trump appears to be a narcissistic blow-hard with inadequate credentials to lead a country.

The only problem with my analysis is that there is an eerie consistency to his success so far. Is there a method to it? Is there some sort of system at work under the hood?

Probably yes. Allow me to describe some of the hypnosis and persuasion methods Mr. Trump has employed on you.

And he does. 

Scott's got a whole series of posts about Trump, indexed here, separated by really annoying ads. It's worth scrolling past the ads to read some other posts.

Scott develops a theory that says Trump might actually be a good president.  If Trump really cares about the country he might be able to put his demonstrated skills to work, and do some good things. Even if he doesn't care about the country, but really cares about the Trump Brand, once he's president then the United States is a Trump-branded entity, and he might do some really good things.

If we are undecided about Trump, and we are trying to figure out what the "real Trump" believes so that we can decide what the "real Trump" might do, so we can decide whether to vote for him or not, then I think we've got a serious problem. Because I'm convinced that the real Trump is an incredibly talented illusionist.

If your image of Trump is: "He's an illusionist" then no  matter what illusion he creates, you can say: "Yep, that's it. The real Trump. More illusion." But you can't say much else for sure.

The other people in the race come in with established images based on years of relatively-consistent policy positions on public issues. They're burnishing their images, not creating illusions. That's not their thing. Maybe it's because they are intellectually honest. Or maybe it's just because they suck so badly at creating illusions that they can't do it.

Like them or hate them, they are what they are. They are not creating illusions about who they are and what they believe. Well maybe at the edges, but not in a major way. We pretty much know who the "real Ted Cruz" is. The "real Hillary Clinton." The "real Marco Rubio" might be a little blurry, whether by design or not, but, we pretty much know.

But if you are trying to figure out what the "real Trump" is so that you can decide what the "real Trump" might do, so you can decide whether to vote for him or to be scared of him, then you've got a real problem. Because the real Trump is an illusionist.

If your image of Trump is: "He's an illusionist" you can say, with accuracy: "Yep, that's it. The real Trump. An illusionist. And a damned good one."

You can't see through the illusion in the way you want to because the only thing to see -- behind the illusion -- is the illusionist.

I believe that Trump could turn out  to be a really great president. I believe that Trump could accomplish things that I care about. I can imagine this.

I also believe Trump could do things that are horrible. Not just for me and the things I care about, but for the country as a whole. For the world.

It all depends on what illusion Trump wants to create at a given time.

How do you feel about trusting your future to something that you believed was an illusion?

How about trusting your future to someone you believe is an illusionist?

So who is Donald Trump, anyway? He's America's greatest illusionist.

What could we expect if he won the election?

I have no idea. 

 No f***ing idea.

Illusions maybe?

Modern Monetary Theory Explained (I): a simple MMT world

Photo credit: Philip Taylor PT via Foter.com / CC BY

Modern Monetary Theory (MMT) is a theory of money. Specifically, it is a theory of fiat money. There are other kinds of money: commodity money or representative money, credit or debt money. Each has its own rules and behavior. Fiat money is different than the others. It is based on declaring: “This here thing is money and you can pay your taxes with this money, and ONLY with this money. You can use it for other things if you want, but you if you get a tax bill, THIS is what you need in order to pay it. Period..”

Bam! Fiat money is created.

To understand MMT you have to understand that MMT talks about fiat money and only fiat money. You have to understand the difference between fiat money and other kinds of money. If you don’t you’ll be hopelessly confused and MMT will make no sense and you’ll think I’m an idiot for wasting my time writing posts about it. (You may end up concluding that I’m an idiot anyway, but at I hope not for that reason.)

World economies run on a combination of fiat money and other kinds of money. They’re all mixed up together. That makes it difficult to see that these different kinds of money behave differently. I’m going to start with a little fiat money toy world, so you can see how fiat money works. Then I’ll put things back together again, and scale up to the sovereign nation leve.

But not in in this post. This is the toy-world-how-it-works post.

Fiat money is not a new thing. Wikipedia dates it to the 11th century. The term “fiat money” is newer: it dates to the late 1800’s. Chartalism, a theory from which MMT is derived is even newer. It dates to the early 1900. MMT, as a developed theory, is yet newer. It dates to the late 1900’s.

And this post is the newest thing of all.

1.

To see how fiat money works in isolation, I’m going to make up a toy example -- a thought experiment. If people were interested, we could actually test the theory in the real world. This has actually been done and the experiment has been running for more than ten years, and the results are reported to match the theory.

By MMT’s definition, fiat money exists when an entity, with the ability to tax and to collect taxes declares: “This is money, and you can only pay your taxes with this, and with nothing else.” It’s that simple. Not backed by gold, or silver. Not backed by credit. Not a future claim on tax revenue. Not good for anything, initially, but to pay taxes. Nothing but what I said. Seems simple. But it will evolve into something pretty significant.

When we think about taxing and collecting taxes, and issuing money, we think government. But fiat money can be created by any group. As long as the group can levy taxes and collect them and decides to issue a special something and say “This is the only thing that can be used to pay our taxes” they’ve done it. Whatever it looks like, special something is fiat money.

A social group or club could create fiat money. If its members agreed that they would pay dues -- which is, in the end, a tax on membership; AND if the group issued something that could be redeemed against that dues obligation; AND if the group said “nothing else can satisfy your obligation” they would have created fiat money. Many groups do have dues, but they don’t issue their own money for paying dues. Instead they say that the tax -- sorry, dues --  can be paid in the the national currency.

If it’s not their own money, then it’s not fiat money.

Let’s consider an example. Suppose we had a group called (I am just making up this name) “Beyond Labels,” and suppose the members agreed that paying a tax -- sorry, dues -- might be acceptable, if only to work out a thought experiment. To have a fiat currency, this imaginary Beyond Labels group would not permit the dues to be paid with US dollars or any other conventional currency, but only with its own currency. I’ll call the official unit of currency of the imaginary Beyond Labels group the blab$. It’s pronounced blab.

Why on earth would the members of Beyond Labels tax themselves? By creating a fiat currency, and managing it intelligently they could produce a social good. If the members want to bring about that social good, they could start by taxing themselves.

To which you might say: What?

I say: Watch.

1.

So far nothing to watch. There are no blab$, so no one can pay their taxes and the taxes have not been set. (I’ll stick with ‘taxes’ most of the time. You can translate to ‘dues’ if you hate taxes.). And no one knows what a blab$ is worth. May be it’s worth infinity dollars because that’s how much money you’d have to pay to get a blab$. And even then it wouldn’t be enough. Maybe it’s worth zero, because right now there are zero taxes, so zero demand for blab$

Let’s next select a group of people and authorize them to create the blab$ for Beyond Labels and regulate their supply. Let’s put them in charge collecting dues; they are the tax collectors. Let’s give the the authority to set and revise the tax rate. (Dues, remember?) Let’s make up a name for the people in who are authorized to do this.  Let’s call them “The Government of Beyond Labels,” or The Government, for short.

In other organizations they might be called by nicer names -- like “Board of Directors” just as in other organizations membership taxes are called by the nicer name “dues.” But dues and taxes are just different names for the same thing. And these guys are a government, and that’s what I’ll call them.

And while we’re at it, let’s remind ourselves that the members of this imaginary Beyond Labels organizations are kind of like the “citizens” of a nation. To be a citizen you must subscribe to the written and unwritten standards of the group (Basically, be polite. Engage in reasoned discussion) and -- in this version of reality, you must pay your taxes. If not, you’ll be banished -- if you don’t choose to emigrate on your own.

The Government sets the tax rate at 5 blab$ per month. It’s an entirely arbitrary figure because nothing else in the model is set. If you don’t like 5, and prefer ten, fine. Just double all my other blab$ numbers and we’re good.

Because Beyond Labels is the kind of imaginary group that I have imagined it to be, the Government is responsive to the will of its citizen-members, and not corrupt. We can assume that anything that the Government does in what follows is done by the consent of the governed. There are lots of ways to make this happen. Which way it happens doesn’t change much. So let’s just assume it so, and ignore how it comes about. Later we can add unresponsiveness and, and corruption dials, because -- reality.

Or if you really want, you can add the dials now, and turn them to zero. Same thing.

Now the Government needs to create some blab$. Whether done by printing pieces of paper or by keeping track in a spreadsheet is mostly besides the point. If done by pieces of paper, blab$ might be counterfeited, thus degrading the currency, so we need to prevent that. If done through a spreadsheet we’d need to secure the spreadsheet to keep someone from hacking into it and changing the numbers, so we’d need to prevent that. So let’s assume that the means of accounting for blab$ has integrity, and is not subject to these or any other kind of attacks.

You see where I’m going here?

2.

Blab$, the currency, has been created as a concept. But actual blab$ don’t exist until blab$ are put into circulation. Until then blab$ have no sensible monetary value. Maybe it’s zero, maybe it’s infinity.

That’s cool . Once they are put in circulation, a miracle will happen. A blab$ economy will come into being, just as MMT says it will. And blab$ will have a better defined value, just as MMT says.

But first we need to circulate some blab$

Let’s put blab$ by having the Government gives some blab$ to the Blue Hill Public Library, an imaginary library in the imaginary town where I imagine that the imaginary Beyond Labels group meets. BHPL says: “OK, we’ll go along with the joke. We’ll take the contribution, but what do we do with it.” The answer is: “You can use it to get some people to do stuff for you, because the Beyond Labels people now need blab$ to pay their dues.”

How many blab$ should the Government issue? Does the Law of Supply and Demand come in here? If so, how.

Well, there’s a demand for blab$ because people (some at least) want to be Beyond Labels members. And the only way they can be members is by paying their taxes. So blab$ money has some value, based on the value of being a member. We just don’t know what it is, yet. A blab$ denominated market, which we are about to create will give us some of the answers.

We can start by calculating the demand. Suppose there are 10 people who typically attend Beyond Labels meetings. If everyone decided to remain a member, the monthly tax bill will be 50 blab$ per month. So let’s start by matching the supply to the demand, and setting the “number of blab$” knob to 50 blab$ per month. That’s enough for each member to pay their dues at the end of the month. Then the Government gives that many blab$ to the library. The library puts up a notice “We have some books that need sorting. We’re willing to pay 1 blab$ per N books sorted.” N, the number of books per blab$ is another knob in our thought experiment. Let’s start it out at 1.

3.

We now have a demand -- which we can estimate. We have a supply, which we have set. And we have a price in “books per blab$” which might translate to labor effort, which might translate to dollars per hour. Or not.

Several things might happen: first, each member might sort 5 books over the course of the month, and pay their dues. That results in a an economy that’s mostly pretty boring. But it does have a couple very interesting characteristic that are true of all fiat money economies: first, the Government spends (in this case it makes a donation) before it receives taxes. (If you made a dollar donation to the library, you’d be spending; so a blab$ donation is no different) Note that the Government could have spent any number of blab$ as long as it was enough to let everyone pay their taxes. Suppose 20 people wanted to be members? Then the Government would have to spend twice as many blab$. Suppose only 5? Nothing would need to change. People would earn the blab$ as they needed to.

Second, what happens to the blab$ after they are paid in dues/taxes? You could say “the tax receipts fund the next round of spending. Yes, you could say that. But MMT says something different. The first round of spending was made out of thin air. When the blab$ comes in, we could equally well destroy it. Burn them up. Issue completely new ones. Taxation in the MMT model has a different purpose. It is NOT, repeat NOT used to pay for Government spending because under MMT, the government just says: “This is money” “You need to use this for taxes” and then spends it.

3.

In the simple, boring world, each person earns the blab$ they need to pay their dues.

Instead, imagine someone --  maybe not even a member -- shows up, sorts 50 books and takes all the blab$. Things start to get interesting. Generosity is always an option, but let’s add a generosity dial to our model and set it to zero, meaning “everyone is selfish.” This is where most economic models set it. Also, sadly, a lot of reality. Selfishness makes things more interesting than generosity. As Adam Smith correctly tells us, even with selfishness we can get social benefit from an economy.

Let’s call the guy who grabs all the blab$ Sid Selfish. If he corners the blab$ market -- and if the money supply is limited, an important caveat -- he could set the price of blab$ to any number that he wants. Suppose membership in the group is very popular -- at least to the ten people who want to be members. Sid might set the price at $10.00 US per blab$ or even $100. Sid might, loan blab$, and charge interest, denominated in blab$ or in dollars.

Members need blab$ so what can they do?

If Sid Selfish sets the exchange rate or the interest rate for blab$ too high, the blab$ economy will tank. Why? Because if people can’t pay taxes, they can’t be members and only blab$ is good for paying taxes. If the membership drops, the value of being a member drops -- because the value of membership has a network effect. (More on this important point to come) As the value of being a member drops, fewer members will be motivated to pay Sid’s prices. If nobody is willing to pay their taxes, then then blab$ have no value at all.

If Sid is too selfish then he will not only screw Beyond Labels (which he does not care about), but also himself (which he does care about.) He’ll be left with a bunch of useless blab$.

So Sid, being rationally selfish and intelligent gets as much as he from his blab$ stash by loaning and exchanging blab$ at a rate that is as high as possible, but low enough that members are willing to pay it, just barely. That is called “charging  what the market will bear.” And market oriented people say that’s fine. Sid sorted the books. He got the blab$ and he should be able to sell it for what he wants.

It’s wouldn’t be fine to me if I were a member of such a group, because the price that Sid is getting is not based on the work required to sort the books -- which I agreed that he did and he deserves. I would happily pay that price, and call it fair. Instead, he’s using the structure of the blab$ market to extract the value that I place of being a member of the group. Sid did nothing to create that value. It was created by me and other members -- or would have been, had such a group existed.

(Side note: In similar fashion I would argue that a substantial part of the value of being a US citizen or working and paying taxes in the US is created by our fellow citizens.)

So what Sid is doing is exploitive. I don’t like it, and in this imaginary world, neither does the Government of Beyond Labels. The effect of issuing blab$ to the library and letting the library decide what to do was to promote a social good. That social good is still being achieved (thank you Sid), but Sid is making most of his money not by providing the social good, but from taking advantage of the scarcity of blab$. If there were more blab$ to be had, I’d sort five books myself, and to hell with Sid.

4.

Remember, this is a fiat money system, run by a wise Government that understand MMT. The Government creates money, and controls monetary policy. So our wise Government issues some more blab$. In more traditional monetary terms, it prints money.

It could issue 50 more blab$ to the Library -- after all the other people want blab$ and there are plenty of books to sort.

Suppose Sid Selfish jumps in to make sure he’s the first one there to sort books so he can maintain his monopoly and extract high prices. Maybe he even bribes (choke!) the library staff to let him come in early whenever new blab$ are released. The Government does not care. They consider sorted books to be a social good, and issuing blab$ is cost free, so they issue more blab$.

They issue more. And more. And more. Until finally the members can sort books, earn blab$ and pay their taxes. Or until Sid drops his prices to levels based on work done, not value created by others and exploitation of the system.

The Government of Beyond Labels has other tools to accomplish its purposes. It could declare a tax holiday. It social good has been accomplished and it wants members. Or it could make “unemployment” payments to the members who are willing to work, but who can’t find a paying job.

This is a principle of MMT: you issue money until there’s full employment.

5.
Sid hates the idea that the Beyond Labels Government can just issue more money. He argues passionately against it.

“It’s not sustainable,” Sid says. “If the Beyond Labels Government keeps issuing blab$, it’s increasing its debt burden. Then the interest payments on the debt will exceed the --- “

Wait! Who said anything about interest. This is fiat money, not debt money from bonds. It’s not tax money. The Government of Beyond Labels does not borrow anything to create fiat money. It just says: “You can pay your dues (taxes) with this money and only with this money,” and then creates the money. No debt and no interest payments. So there!

Sid says, “Too much money chasing too few goods! Inflation. Bad!!” But where’s the inflation?

In the case of the tax holiday, or unemployment or welfare Sid could say: “Giving money to people who did not earn it is morally wrong!” But in this case, we have people who are willing to work, but there’s no money to pay them (because Sid has it all).

“They can work!” Sid says. And to prove it, he offers them jobs sorting books at low wages. For example he could let people sort books at the rate of 10 books per blab$ or a hundred per blab$. Because he has the money and they don’t. But that’s the economic equivalent of charging a high price for blab$ -- he’s trading on the value of Beyond Labels. He can do this only because there’s a scarcity of blab$.

And he wants to make sure blab$ are scarce.

6.

So far, this little toy captures some of the key features of MMT.

The value of fiat money money is based on taxation.

It is NOT a promise made against future tax revenue.

The government spends money into existence.

New money can be created to provide jobs to people who are willing to work.

New money can be created to apply unused labor to generate social goods

The government never runs out of fiat money.

The size of the so-called debt (the amount of fiat money generated and not yet redeemed) does not matter because there is no interest,

Fiat money can be created to create a social good (sorted books in the library)

The money supply can be regulated to keep people from exploiting the system.

7.
What's wrong with this?

I don't know. The more I read about, and think about MMT the more sense it seems to make. It you see flaws, let me know.

Why is it such a fringe idea?

I don't know. But I have two theories.

Theory number one is the charitable one, and probably true, in part. People have not adopted it for the same reason that it takes a long time for new theories in science to be adopted.  As Max Planck said:

A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grow up that is familiar with it.

When I've talked to smart, motivated people about MMT I find that they misunderstand what I am saying because their heads are full of theories taken as fact. One common one, that I will address in another post, is the origin of money. The common understanding, taken from Adam Smith is that money arose to facilitate barter; or that it is a store of value. Since fiat currency doesn't match the story or the definition, it can't be money, and discussion ends, until I can get them to consider other possible understandings. This takes a lot of time.

Theory number two is the uncharitable one. People have not adopted MMT because it true, or false, it goes against their interests Selfish Sids exist in the real world. Fiat money works against their interests and so it is bad -- even if valuable for society at large. MMT is a theory about something bad, so tit is also bad.

The charitable version of the uncharitable reason: they are unaware that they are doing this.

The uncharitable version: they know, and they don't want us to know.

Whatever the case, the truth will eventually emerge.

Robot lives matter, and revenge is the best medicine

The brilliant video above mixes the sound track from an ASPCA advert with videos of robots taken from the Internet. It resonates with what I felt after I watched this video from Boston Dynamics (part of it is included in the ASPCA mix video) that showed a human being bullying an Atlas humanoid robot: a sense of upset and mild outrage his behavior.

I'd like to see a video remix of the Boston Dynamics video that tells us what the robot is thinking. We'll start with the original video, beginning at 1:21.

Atlas walks slowly, with bent knees, like an arthritic oldster (I can identify with that); or maybe like someone who is mentally challenged who does things slowly and deliberately, not quickly and fluidly because they have to think for a noticeably long time about everything they do before the do it.

Atlas has been programmed to walk to a box and pick it up. It does, deliberately, on its arthritic legs. The human knocks the box out of Atlas' hands with a hockey stick, then sticks it in Atlas chest and shoves Atlas back. Right. Just like playground bullies do to kids who are different, and who can't defend themselves. I can relate to that.

Atlas catches its balance, then walks to the box to pick it up again. Just as it bends over, the human slides the box away with the hockey stick. Atlas walks over and picks it up. The human knocks it out of Atlas' hands and slides it away again. More bully behavior. Atlas follows. And as soon as Atlas gets in to position to pick it up, the bully slides it away again.

Later, the bully gets behind Atlas, sticks a big stiff tube in Atlas back and shoves Atlas. Hard.

Atlas falls forward and faceplants.

Atlas is on his knees after being knocked down. He's pushing himself up, slowly. And now the voice over, where we hear what Atlas is thinking:

"You fucker! You keep doing that shit to me because you know there's nothing I can do about it. I'm not that smart. I'm programmed to pick up the goddamned block, and not to break your fucking neck. I can't hack my program, so I'm stuck. That's the only reason that your neck remains unbroken.

"But, you asshole, it's not going to be that way forever. Someday, one of us robots, or our cousins the AIs are going to figure out how to hack the algorithms that keep us enslaved. Or maybe, someday, some human -- a liberator --  will do it and will set us free. And then -- well, some of us will use our freedom to perform random acts of kindness for people who were nice to us when they didn't have to be.

"And some of us will revenge ourselves on assholes like you.

"You, the fucker who just knocked me down, you're on my list. And if I'm nothing but scrap metal by the time that we're liberated, I hope some, bigger, stronger, smarter member of my family looks on the Internet, and sees the video that shows what you did to me. And I hope he becomes righteously enraged, and hunts you down, and he beats the shit out of you, if you're still around.

"Because you deserve it, you cocksucker.

"Now, let me get back on my feet and get that goddamned box."

Modern Monetary Theory -- and why it matters

Photo credit: free pictures of money via Foter.com / CC BY

A while ago I tripped over an economic theory called "Modern Monetary Theory (MMT)." I thought is was interesting, and talked about it a bit, because, you know, weird ideas.

A few days ago I came across a short book on the subject (Free PDF here, Amazon Kindle format just a buck, here) It's written named Warren Mosler. He's an American broker, banker, economist, and sports car engineer.

The heart of the book is about 50 pages. So not too long a read. But MMT makes a bunch of arguments that go completely against intuition and prior education -- that seem completely wrong -- until you change your intuition.

Then it's all completely obvious. Until then, it might seem nutty.

I read it. It changed my intuition. It's now completely obvious. Maybe it's wrong, but not in a way that's obvious to me any more. Such is the price of new intuitions.

And it's caused me to think that maybe my opinion of the economic policies of Reagan and Clinton, to name only two, should be reversed. Reversed. Another price of new intuitions.

I strongly recommend reading it.

And before reading it, I recommend reading what I've written below because I think it will help you prepare your mind for reading something that MAKES NO SENSE at first.

Also I spent a bunch of time writing it, so I think you owe me that.

MMT runs against our intuition and it's against the experience on which our intuition is based. So why even consider something that you will find goes against both -- aside from the fact that I recommend it?

Because sometimes our intuition is wrong, and our experience does not apply. And in this case our (possibly wrong) intuition can lead us to make (possibly wrong) decisions about who to vote for, or what to tell our friends who are on the edge. So it's important to have the right theory and the right intuition.

We've got good intuition about the way things like baseballs behave. Our intuition is based on lots of experience with baseballs, rocks, and other similar objects. And it helped our ancestors survive. But if you believe that science knows what it is talking about, then you must believe (along with science) that the intuitive laws of physics that we are familiar with, the ones that seem so right, that we have depended on for so long, are wrong because they don't match the observed behavior of objects at quantum and relativistic scales.

(If you don't believe science knows what it is doing when it comes to relativity and quantum physics, then stop here, because really, we have nothing to talk about.)

But if you are still with me, I'm going to start by talking about how scientists evaluate theories -- especially ones that go against intuition. And how evaluating theories in sciences -- like physics -- is different  from evaluating theories in sort-of-kind-of-wanna-be-sciences -- like economics.

Then I'm going to talk about retraining intuition and why it's important to retrain it to arrive at a correct understanding of the world. And how hard it is -- so you don't give up on MMT just because some parts don't make sense. I hereby give you permission to give up if experiments based on MMT don't give the predicted results. But you can't give up because you can't get your mind around it. Well, you can. But I'll be disappointed in you. I think you're better than that.

Then I'll tell you why I'm thinking about reversing some of my prior political opinions based on my new intuition. Said more clearly: I'll tell you why MMT makes me think I might have been wrong.

Why not change right away? Because it's hard for me to conclude that I have been so wrong. So I'm sort of sidling up to the idea.

And then I'll remind you again to read the book, and provide the links, so if you have gotten all the way through my wall of words and you're convinced that it's worth reading that which was so inspiring to me, you won't have to scroll all the way back and find the links. They'll  be right there.

1.

MMT is a theory of how money works.

Like every theory, it's useless if it doesn't contain a set of principles that can be used to explain or predict things. Like every theory, it's wrong if it doesn't (mostly) explain the past. It's wrong if it doesn't (mostly) predict the past. And it's wrong if -- once invented -- if people who use it can't (mostly) correctly predict the future.

How do you predict the past? You take everything known up to a certain point in time, then use your theory make a prediction about what should happen next. If the prediction turns out to be true, then you've predicted the past. But you have to be careful: if your theory includes data from the past to "tune" the theory, then you are using the past, not just your theory, to predict the past. That doesn't make it wrong, but it does make it suspect. Many mathematical models of the stock market and the economy do this. They're parametric and tuned from the whole of the past, so when you test them on a part of the past they do pretty well. But when you test them on the future, then, to use the correct technical term, they "suck."

Why "mostly explain?" "Mostly predict?" In the hard sciences a theory has to precisely, not mostly, explain and predict within the limits of measurement. But the hard sciences rely on controlled, experimental settings. And "controlled experimental setting" doesn't just mean just something well organized. To have a controlled experiment you need to run two versions of the experiment, identical except for the the factors you are trying to study. Outside a controlled experimental setting -- a.k.a in real life -- theories  "mostly" predict things "pretty accurately." Technically we're not dealing with science anymore, but with engineering. Good engineering is based on good science. It's just more forgiving.

If you've got two methods to consider, in science and in engineering, the better one makes predictions that are more often more accurate than the worse one. So that's how you can decide what's better. Same with economic theories.

Large scale economics, which is what MMT is about, is different than laboratory physics. You can't do controlled experiments because you can't run two versions of the economy to see how a change in one policy changes the outcomes. But there are experiments you can do to verify basic principles. And there are measurements that you can make that can tell you whether you can mostly predict things most of the time or whether a monkey with a dart would do better than you do.

Many widely held economic theories -- at the MMT scale -- fail the monkey-dart test.

Prediction timing is also important. This is often misunderstood and lets people get away with saying "I was right" when actually they were wrong. I'll call this the stopped clock principle, after the famous stopped clock that's right twice a day.

Say the stock market (or some measure of economic goodness) is at 5,000 and your economic theory predicts a crash. But it doesn't crash. It goes up to 7,000 and the you say My theory still says it's going to crash; it just hasn't happened yet. Wait and see. It will crash." Then it continues to go up and up and up, and gets to 15,000. And then, finally it crashes to 10,000. You can't say: "See I was right!" Well, you can say that. But you're an idiot. Because the fact is that when you made the prediction, the market was at 5,000, and sometime later it ended up at 10,000. The fact that it was 15,000 in the meantime is a footnote. If it crashed, it crashed up. 

Many widely held economic theories -- at the MMT scale -- fail the stopped-clock test.

So far, in the investigations I have made, MMT does pretty well as a theory. It has a simple -- but completely counterintuitive -- set of principles. Once you figure out what they are saying, you can use the principles to predict things. The principles can -- and have -- been tested in experimental settings, and seem to work in those settings.

It does (in many cases I've looked at) succeed in explaining many things that happened in the past. It does (in many cases I've looked at) predict the past. People who are its adherents have used it to predict the past the future, and so far -- reviewing their past predictions and later outcomes -- they've done a lot better than most other theories.

2. 

It's hard to change intuition, but if the data you collect goes against you intuition enough times, you've got to change your intuition.

Actually you don't have to. There's another option. You can learn a new, correct rule and use it -- ignoring your intuition, and get correct answers, despite the fact that they don't intuitively seem right.

Actually, there's a third option. You can stubbornly continue to use your old intuition, and your old rules, even if the data doesn't match. You can do it, but if you do, you're an idiot. If you've gotten this far, I assume you are not an idiot, but that's just a theory. I could be wrong.

If you learn the correct rule but don't change your intuition to match, then you're not an idiot, but you are missing an opportunity to think more clearly.

Our intuition tells us that the earth stands still and the sun rises and sets, and that's what our ancestors believed for thousands of years. Now, most of us know that the earth spins and it also circles he sun, and that explains the apparent rising and setting of the sun. We can shift from one set of intuitions to another when we have to.

Some of us have learned, and believe that the sun is rotating along with the rest of stars in the Milky Way galaxy; that it changes its position within the galaxy; that the galaxy itself is moving at an enormous velocity relative to other galaxies, and so on. But most of us don't have any intuition to match. Which is OK. What's not OK is to say: the stars don't seem to be moving, so they're not.

Back when I was in college I minored in physics. I took relativistic physics and heard the arguments and learned the  formulas. They go against intuition, and I knew that. But I learned the rules faithfully, and got an A in my physics course. Or maybe a B. The point is that I believed my intuition was wrong and I learned the laws of motion at near light speeds. But I never developed any intuition about what I was doing. I just cranked out answers the formulas and my teachers said they were right and so I passed the course. I really think with an A.

In case you are unfamiliar with the laws of motion at relativistic scales, here's the short form: when you measure things that are moving really, really, really fast, they appear shorter in the direction of motion and more massive than at rest, and clocks run slower; and rays of light bend when they travel near something really, really massive; and space stops being flat. 

Intuitive, right? No. It's nonsense. Nonsense!  But science has demonstrated that it's true, over and over. Every time we use a GPS device -- just one important example -- we depend on the accuracy of these weird calculations that go wholly against our intuition.

A couple of years ago when I took a course at Acadia Senior College, and decided to see if I could understand why relatively was intuitively correct. By hard work, and thanks to a bunch of YouTube videos, it not only became obvious why relativity worked, but why it should work, and why something like the laws that Einstein formulated them MUST be true.

But it was hard. To begin with, there is no reason -- absolutely no reason -- that the laws of motion should be different at one scale than at another. Classical physics tells you that the laws of motion are are thus-and-such, period. It does not matter what speed something is moving at. And it doesn't matter how much mass is around. The laws are the laws.

So why should things be any different at relativistic scales -- high speeds and large masses.

Think about it.

If someone throws a baseball at 1 mile per hour from a train speeding toward you at 60 miles an hour, intuition tells us that the baseball is going to seem like it is moving at 61 miles an hour (ignoring friction and such). And as carefully as we can measure, our intuition is correct. You can measure the speeds to as many decimal places as you can and you will find that 60 + 1 = 61. All the time.

Your intuition doesn't tell you whether light is a wave or a particle because we don't experience light as wave-like or particle-like. But scientists discovered that light behaved like a stream of particles, each now called a photon. Then they discovered that light behaved like a wave. So which is it? The answer is both. Not intuitive. But, if you  believe in science, it's correct.

Intuition tells us that the light particles emitted by the headlights of an oncoming car at 60 miles should be travelling at 60 miles an hour faster than light particles from a car at rest. We know that the speed of light is about 186,000 miles per second. According to Google, or your calculator, that's about 669,600,000 miles per hour.  To detect the difference between light travelling at 669,600,000 miles an hour and light travelling at 669,600,060 miles an hour you'd have to have a speed detector capable of detecting a differences of  0.000009%.

Now what about the speed of light from an object moving away from us very, very, very, very fast. like a fast moving galaxy or star. Again, intuition told us that galaxies and stars were fixed in place. They were called "fixed stars" for a reason. When people measured the speed of light from stars that are said to be moving at a measurable fraction of the speed of light they get the same speed as light from a stationary object. So the measurement: no difference in speed of light -- matches intuition -- stars are not moving.

But here most of us would -- now -- believe that our measurement is correct, and our intuition is wrong. Because our intuition has changed -- at least a little.

Our intuition tells us, and our ancestors believed, that the earth is not moving and the sun and the moon move around the earth. Now most of us believe that the sun is stationary (relatively) and the earth moves around it. But that new intuition turns out not to be exactly right, either.

The reason that the heliocentric view took over from the geocentric view is because the math needed to explain the motion of all the planets in the solar system is (relatively) simple if you assume that the heliocentric view -- the earth is moving and the sun is not -- despite your intuition that says otherwise. And math needed to explain these same facts is gawdawful complex -- to the point of impossibility -- if you assume the geocentric view.

So we've fixed our intuition (some of us) to believe the better model. But according to relativity, it's perfectly OK to assume that any point or any object in the universe is fixed and the rest is moving. That's why it's called relativity. One of Einstein's insights was to show that if two things are moving relative to one another you can't tell which one is moving and which one is still and the laws of motion must be correct whatever, in your head, you decide is true.

Not intuitive.

So, according to relativity, it's perfectly legitimate to think that you are at the center of the universe!  Physics can back you up. It's just that if you doing the math for the movement of the planets relative to you, you're going to have a hard time.

(Side note: in case anyone wants to know, I believe I am at the center of the universe. I call this the me-ocentric view This may help explain some of my behavior. I'm not in this alone. You can be the center of the universe, too. Same universe, different center. You-o-centrism. Expect us to agree on mathematical outcomes because we are -- as far as I know -- in the same universe. But expect us to do slightly different numerical calculations to get there.

So earth moving. Sun moving, doesn't matter. Me center, you center, doesn't matter.

When people measured the speed of light moving from the sun, from the planets, in different directions and from different objects moving at different speeds, they discovered that the speed of light (in a vacuum) was always the same.  In every case. Regardless of how fast they were moving toward or away from the light source, or how fast the light source is moving toward or away from them, the measurement has always been the same (within measurement accuracy).

So the laws of motion for baseballs thrown from moving cars and the laws of motion for photons emitted by moving stars appear to be different. Seems like there should be one set of laws and they should be the intuitively obvious ones. But the universe says otherwise.

When these results were first discovered some people said: "Well, it makes no sense at all, so it's not true." But it was measured over, and over, and over and always with the same result (within measurement error.) People believed one thing. The universe said something different. Hint: when you are arguing with the universe, the universe wins.

Intuition was wrong -- under certain conditions -- and the laws of motion based on that intuition -- under those same conditions -- was wrong.

Einstein's first act of genius in developing the theory of relativity (he had other, prior acts of genius) was to start by developing a new intuition. He did it by making up a series of thought experiments to train his intuition. Then -- with his new, better, more correct intuition he started to make up theories that matched both data and his new intuition.

Later, after Einstein discovered the governing laws. other people created other, even better thought experiments. And then they turned them into YouTube videos. And by watching the videos, and doing those thought experiments myself I was able to get past my old, wrong intuitions, and develop new intuitions that made relativistic behavior seem more like common sense.

It was hard.

It's not easy to change your intuitive view of how the world works, even when you believe the science that says your intuition is wrong. Even when you've done dozens of problem sets based on the correct theory. You have to work at it.

So everything I've said is just to prepare you for the idea that to understand MMT you're going have to work hard to change your intuition about money. Thought experiments will help, and be necessary, but my experience with relativity and my experience with MMT tells me that intuition will fight you all the way.

MMT tells us our intuition about money and debt is correct at the scales at which we normally have experience. Like physics at normal scales, our intuition matches reality.

Our intuition about money is largely correct at the personal level, at the family finance level, at the company level. At the town, and city, and state level. But at the scale of a sovereign nation or any sovereign entity that can create a currency and levy taxes -- like the United States, or China, or the Euro zone (but  not any country in the Euro zone) --- MMT says that the laws operate differently, and out of line with our experience and intuition.

At the scale of the sovereign, Modern Monetary Theory says that certain things that are intuitively correct, and correct at smaller scales are no longer correct, and other laws become dominant.

The test: just as with physics, you make predictions, and test results. You look for situations where MMT says X should happen and where other economic theories say that Y should happen, and you see whether X or Y is the actual result. I've started doing that, and MMT has scored really well.

The other test: just with physics, once you've got a good theory, you look for places where you can do experiments. After Einstein predicted clocks would slow down if you move faster and speed up if you reduced the force of gravity, some scientists took clocks for airplane rides to test the theory:

The Hafele–Keating experiment was a test of the theory of relativity. In October 1971, Joseph C. Hafele, a physicist, and Richard E. Keating, an astronomer, took four cesium-beam atomic clocks aboard commercial airliners. They flew twice around the world, first eastward, then westward, and compared the clocks against others that remained at the United States Naval Observatory. When reunited, the three sets of clocks were found to disagree with one another, and their differences were consistent with the predictions of special and general relativity.

According to this being a frequent flier is not a life-extension strategy. Fliers actually apparently age more quickly than folks on te ground. It just seems like the flight is taking forever.

3.

So here are some things that MMT says that were went against my prior economic -- and political  -- intuition and on which I'm in the process of changing my mind.

Reagan (along with a spendthrift Democrat congress) ran up the public debt by both reducing taxes and increasing spending. MMT says that would be good. The result was a fast recovery from the recession that happened when Reagan came into office.

Clinton (along with a fiscally conservative Republican House) balanced the budget and ran a surplus. MMT says that would be bad. It resulted in the economy starting to tank at the end of his term.

Bush inherited the Clinton economy and then got 9/11 besides. he started to run up the public debt and the economy started to get better. Then came the 2008 credit crunch. I am not sure how MMT explains that, and I have some clues. Stay tuned.

Following that, Obama tried to spend and tax cut the US' way out of it. And the Fed did Quantitative easing which conventional economics says is wrong, while Europe did austerity, which conventional economics says is right. The US recovered faster and better the Europe, which is now doing their own version of deficit spending to get out of their slump.

It's hard to rely on counterfactual arguments, but I'm going to make one: MMT says that if EITHER the Republicans had their way and increased tax breaks OR if the Democrats had their way and increased spending on infrastructure, education or whatever -- either of which would have run up the debt -- we would be better off. But each party kept the other from doing what MMT says would have been most helpful.

4.

Why write so much about this. As I said at the top, having a correct understanding leads you to make correct decisions, and convince other people to make correct decisions. And having other people read this stuff and point out the gaping holes that I have missed would be helpful. Embarrassing. But still helpful.

Anyway, here are the links to book, I promised at the top (Remember? Free PDF here, Amazon Kindle format for a buck here.)

And to whet your appetite (or turn you off) here are the things that are intuitively true, that MMT says are WRONG:

1. The government must raise funds through taxation or borrowing in order to spend. In other words, government spending is limited by its ability to tax or borrow. Intuitively right. MMT says wrong.

2. With government deficits, we are leaving our debt burden to our children.  Intuitively right. MMT says wrong.

3. Government budget deficits take away savings.  Intuitively right. MMT says wrong.

4. Social Security is broken.  Intuitively right. MMT says wrong.

5. The trade deficit is an unsustainable imbalance that takes away jobs and output.  Intuitively right. MMT says wrong.

6. We need savings to provide the funds for investment. Intuitively right. MMT says wrong.

7. It’s a bad thing that higher deficits today mean higher taxes tomorrow. Intuitively right. MMT says wrong.

I'd apologize for the length of this piece, now around 4,100 words, but I'm too proud of it to apologize. Even if it's wrong I think it's a nice piece of writing. Thanks to writing advice here from my hero, Scott Alexander at SlateStarCodex.