Saturday, August 7, 2010

Intellectual works aren't scarce -- just like money

If you listen to Stephan Kinsella or his acolytes, you're probably well familiar with the argument that "Intellectual property rights should not exist" because "intellectual works aren't scarce", though this is often confusingly shortened to "IP isn't scarce". Here's Kinsella's latest compilation of the anti-IP arguments, that being one of them. (Which led to a very lengthy discussion.)

Well, I've been reading Jaron Lanier's recent book, You Are Not a Gadget, which gives a good reply (p. 102):

It is a common assertion that if you copy a digital music file, you haven't destroyed the original, so nothing was stolen. The same thing could be said if you hacked into a bank and just added money to your online account. ... The problem in each case is not that you stole from a specific person but that you undermined the artificial scarcities that allow the economy to function. In the same way, creative expression on the internet will benefit from a social contract that imposes a modest degree of artificial scarcity on information. [bold added]


I've made a similar point before: Money is information -- specifically, the relative amount that the world (believes that it) owes you. When money is stolen from you, then you can certainly force yourself to think of it in terms of

-a physical item being removed from your possession, or of
-a server having "unauthorized use".

But what's really important is the editing of that information: where before, the world thought it had a remaining balance against you of $X, now it thinks that the thieves are owed that $X. This problem persists even after you are given compensatory paper or the bank gets standard compensation for trespassing, and it's what people care about.

MMORPGs (World of Warcraft, Everquest, etc.) have already assimilated this lesson. In such online games, your money really is nothing but a database entry. It doesn't correspond, even in principle, to a physical object, just the knowledge of a relationship.

Lanier's alternate suggestion, following Ted Nelson, is that we could instead simply have an automated system that charges for each time a given intellectual work is accessed. People could "pirate" these (already freely-accessible) works by only using versions stored outside of where there access would be recorded, just as they do today when pirating works. But so long as the public regards this as wrong, and wrong for the same reason as counterfeiting, they would run into the same problem as counterfeiters. And the relatively low cost with which the works could be accessed under such a system would remove most of the sympathy for them.

I note that one particular snag of this is that people will not want to have to think of the costs each time they want to look at a book again. However, if everyone paid a fixed amount each year, and their choice of what to access only determined which fraction of that payment went to each creator, then choosing to view anything would cost nothing on the margin, further eroding any incentive to pirate.

24 comments:

Taylor Conant said...

Hi Silas,

I need to read and re-read this post and some of the links to try to get "on top" of the IP debate but I am curious what you think of the following situation (and apologies if it's glaringly obvious/really stupid):

Say I buy a music CD and then rip the files to MP3. I don't distribute the MP3s to anyone, it's just easier for me to use my own music that I purchased on the CD, in digital format. I prefer keeping it with my other music on an iPod, or listening to it in a mass library on my computer, etc.

Now, I decide to raise a few bucks by selling some of my old CDs. But, I still have the MP3 copies of the music that I made. Selling the CDs allow me to "have my cake and eat it too" so to speak. Should I destroy my MP3 copies at the time I sell the CDs?

Not trying to play gotcha here, really just curious what you think of a situation like that.

Silas Barta said...

Thanks for your comment, Taylor.

I would have to say that you were buying the usage rights of the song when you bought the CD, and that this was the understanding of the conditions of the purchase on both sides, so they are "attached" to the CD. This means you're obligated to destroy your copies on sale of the CD, which is really sale of the rights.

Is this hard to enforce without most people accepting that such rights should be respected? Yes. But as I keep emphasizing, you can say the same thing about regular, physical property rights. (And keep in mind, under the system I envision, it doesn't cost you anything at the margin to use an intellectual work; the only reason to disconnect your viewing from the view records that determines compensation is because you want to deprive the world of which works are enjoyed, which would be really frowned upon.)

I'm writing up a longer post on IP that you should be interested in, as it discusses the sources of my intuitions on the topic and (what I believe is) the right way to think about these topics.

Taylor Conant said...

Silas,

I will look for that post (maybe it's your latest? Haven't checked yet) but in the meantime, if the rights you're purchasing are the use rights, then if you accidentally lost or broke/damaged the CD, could you sell the MP3s to someone else and then destroy your copies?

Silas Barta said...

This post was the one I had in mind, which right now is the most recent.

And the answer to your next question is yes, though again the workability of such a system depends on how well people keep up these norms.

Stephan Kinsella said...

"It is a common assertion that if you copy a digital music file, you haven't destroyed the original, so nothing was stolen. The same thing could be said if you hacked into a bank and just added money to your online account."

No, the same thing could not be said. In the latter case it is a violation of contract, and also an act of trespass: use of the bank's property without its permission or consent. (See Why Spam is Trespass.)

When you use information to guide your actions, the use of the information does not do either one--not necessarily, anyway.

So easy to knock these things down.

Silas Barta said...

@Stephan_Kinsella: That argument doesn't work for the reasons I gave in the post:

This problem persists even after you are given compensatory paper or the bank gets standard compensation for trespassing, and it's what people care about.

If someone hacks into a bank's servers, what compensation are they owed? Why does it make a difference if they change the account balance numbers, rather than just fiddle with some settings?

If someone trespasses on the server to "steal money", why is the bank owed more compensation than if they did some other kind of trespass? Once you (reasonably) agree that the compensation depends on how much "money" they "stole" (i.e. "information" they "edited"), then you agree that that is the real crime, not the server trespass per se.

And if you agree with that, then you agree that there has to be some crime going on above and beyond mere trespassing.

Stephan Kinsella said...

"If someone hacks into a bank's servers, what compensation are they owed? Why does it make a difference if they change the account balance numbers, rather than just fiddle with some settings?"

For the same reason it's worse to restrain a woman for 5 minutes than to rape her for five minutes.

"If someone trespasses on the server to "steal money", why is the bank owed more compensation than if they did some other kind of trespass? Once you (reasonably) agree that the compensation depends on how much "money" they "stole" (i.e. "information" they "edited"), then you agree that that is the real crime, not the server trespass per se."

No. This is confused. It's what you get when you try to reinvent the wheel like engineers often do. Once trespass occurs, then there are damages. If you punch someone in the face and do no physical damage, then you owe some kind of restitution. But if the same punch breaks their nose, requiring surgery, and causes them to lose an acting gig, then there are consequential damages there that are owed that are not in the former case. The fact that different types of harm can flow to the victim does not mean there are property rights in values. I explain why b the way the victim should have a vareity of options to choose in the section The Victim's Options in Punishment and Proportionality: The Estoppel Approach, p. 65 et seq.

"And if you agree with that, then you agree that there has to be some crime going on above and beyond mere trespassing."

Wrong. It just goes to the extent of damage the trespass did. This is really not hard.

Silas Barta said...

@Stephan_Kinsella: You're not addressing the fundamental argument, and neither are your papers.

Under your theory of property, people own the physical integrity of the property, *not* its value. (And you remind us of this quite frequently.) So if someone can restore the property to its original physical form, that (plus some modifier independent of the property's value) suffices as compensation.

Therefore, *if* you could restore some model's nose back to its original condition after punching her in the nose, then it definitely wouldn't matter how much money she makes by having an intact nose for purposes of compensation.

But even that concedes too much: by your very own theory of property and contracts, it simply doesn't matter what the victim "intended" to do with the property, for purposes of determining the maximum compensation they are entitled to, and it is for this reason that (IIRC) you dismiss arguments grounded on "detrimental reliance".

And in the case of the bank, they can restore the physical condition of the server: you can change the data stored so that it reflects the original account values. But if the hacker has already transfer the money through multiple banks and then (irreversibly) "spent" the "money" (again, this is actually editing the information the money represents!), the victim bank would not consider this sufficient compensation, because the rest of the world not accept that the bank has that money, so such restoring of the data would be pointless.

This remains true even if the bank were allowed to "do the same thing" right back to the hacker, by editing data on *his* servers without his consent, which is the extent of what your estoppel theory permits the bank to do.

The bottom line is that you can't recognize the full crime that has been committed by only thinking in terms of physical integrity of physical property -- you have to recognize a kind of "information right" to fully appreciate the transgression as occurred, which your theory of property can't do. Ergo, it fails.

Stephan Kinsella said...

Silas, you are reasoning like an engineer-dilettante. This is silly.

Silas Barta said...

@Stephan_Kinsella: If it's silly, identify the silliness, including where the argument goes wrong, or at least recognize that it's not quite so silly, if you would.

(lol @ captch: ruffle)

Stephan Kinsella said...

Silas, it's not worth it. You have to educate yourself.

Silas Barta said...

No, you have to educate yourself. You're the one who can't reconcile your EM and IP positions, not me.

nfactor13 said...

Regarding the altering of bank records, I would think it's a case of fraud (in addition to trespass), and that it's the fraud aspect that measures the damages, not the trespass aspect.

Let's suppose that I deposit two ounces of gold (or anything else, it doesn't matter) with you, and you write out a note that says, "I, Silas, owe Nathan 2 ounces of gold upon his request," or something to that effect. We both sign, and you put it in your safe, along with other similar notes.

Later, I hire someone to open your safe and add a "0" next to the "2," thus changing the note to read, "I, Silas, owe Nathan 20 ounces of gold upon his request."

Now, you don't have a good enough memory to know what every note in your safe says, so when I come to you and ask to receive my deposit back, I know you are going to rely on the note that I've altered and will give me 20 ounces instead of 2. That's fraud, isn't it? And isn't it the fraud that's the main determiner of the harm done? Perhaps I'm missing part of the argument.

Silas Barta said...

@nfactor13: I think it makes a big difference whether it's money or gold.

Money (in its capacity as money) is purely information. It's the knowledge about the net debt owed to each person. So, while the act you described is fraud, you have to recognize that undoing the fraud requires undoing the change to public knowledge. It can't be phrased purely in terms of whether someone transfers possession of a physical good -- it's only the change in information that people care about.

So even changing the records of the wronged party won't matter, because no one will recognize the validity of such a record without a corresponding re-deposit of money -- i.e. an accepted editing of a different part of the publicly held information represented by the money.

nfactor13 said...

But the information is only valuable as it ultimately relates to the transfer of a real asset at some point, correct? Or is that the part I'm not getting?

The money, in whatever form, is intended to convey a claim on assets. And if the trespasser/hacker in your example is falsifying a record to indicate a claim, and then uses that claim to convince someone else to transfer assets of some kind in return for that falsified claim, then it seems as strange to call that falsified record "money" as it would to call a fake painting a "Picasso" in the same way you would refer to a real Picasso painting. The fact that someone could be fooled into believing that it was actually money wouldn't necessarily make it so.

I could be missing something very simple here, but just trying to understand your point better.

Silas Barta said...

But the information is only valuable as it ultimately relates to the transfer of a real asset at some point, correct? Or is that the part I'm not getting?

Yes, that's the part I'm criticizing -- money need not correspond to any physical asset, and in its capacity as money, it doesn't. For example, what real asset does in-game money in a MMORPG correspond to? Note that this in-game money has an exchange rate against real dollars, and by extension, gold.

nfactor13 said...

"Yes, that's the part I'm criticizing -- money need not correspond to any physical asset, and in its capacity as money, it doesn't."

I guess to me, that wouldn't count as money then. Or to be more precise, it 'shouldn't' count as money. To the extent that something used as 'money' does not correspond to any real asset, it involves some kind of distortion or fraud, even if people treat it as money. Of course, if people agree to the distortion ahead of time and it's all above board, that's fine with me. But I wouldn't use the term 'money' to describe it.

"For example, what real asset does in-game money in a MMORPG correspond to? Note that this in-game money has an exchange rate against real dollars, and by extension, gold."

As someone who's played WoW and made a few thousand in gold at the auction house, I speak with some authority. *clears throat* :-)

What gold in a MMORPG generally corresponds to is time. (Of course, you're referring to an extremely manipulated economy, one that doesn't allow any competition, and one that is constantly adjusted through various fees and money sinks.) What people buy MMORPG gold for is to save time that they would have had to spend acquiring that gold in-game. They're paying for another person's time investment.

But you're making a logical error, specifically, the converse error. P->Q; Q; therefore P. "Money is a claim on real assets. MMORPG gold can be traded for real assets. MMORPG gold is money."

The fact that people believe it has an exchange value for real goods does not make it money, though, unless you're saying that anything that has exchange value is money, in which case, there's no real point to the term as far as I can tell. A paperclip is money because I could trade it for a real good.
http://en.wikipedia.org/wiki/One_red_paperclip

Silas Barta said...

@nfactor13: So what you're saying is, that something that is widely recognized as a medium of exchange, and therefore liquid in a certain context, doesn't count as money? Isn't that the very definition of money?

Keep in mind that:

1) There's the liquidity criterion, too, not just (max achievable sale) value. A red paperclip isn't very liquid, while WoW gold is, within a certain society.

2) The only asset that you said WoW gold corresponds to is time -- but that's not physical.

Now, you may have reason to think that people *should* only use money that is redeemable for a real asset. But that doesn't mean things that fail to meet that criterion "aren't money".

Also, when you claimed I was committing a fallacy, I don't actually endorse the P claim ("Money is a claim on real assets"). Rather, I claim that money is the the information about the net debt owed to each person (which need not be satisfied with real assets -- WoW grind time works too!).

nfactor13 said...

"So what you're saying is, that something that is widely recognized as a medium of exchange, and therefore liquid in a certain context, doesn't count as money? Isn't that the very definition of money?"

I should amend what I've said, or at least clarify, that it's not for me to go around telling people what they can or can't call 'money.' What I should have said is that I don't think the term means very much if we use it so loosely as to include things that are not real assets or in no way represent a claim on real assets. Certainly people can call things that don't fit those criteria 'money.'

Perhaps this veers too metaphysical, but I had thought that time was a part of the physical universe on certain views. I understand physicists have their disagreements, but I'm pretty sure time still shows up in the equations. :-) I don't keep up on all the latest developments, though.

But let's set that part aside for now. I don't think that's a key part of the disagreement. (I could be wrong, of course.)

Perhaps you could put it in these terms then. If there is no money in the world, and then I decide to use money for the first time, how might I use it in a way that it is not itself an asset, or that it does not represent a claim on other assets? Or more positively, if money = information, explain what information the money would represent that would be distinct from (or more than) a claim on assets.

Silas Barta said...

But let's set that part aside for now. ... Perhaps you could put it in these terms then. If there is no money in the world, and then I decide to use money for the first time, how might I use it in a way that it is not itself an asset, or that it does not represent a claim on other assets? Or more positively, if money = information, explain what information the money would represent that would be distinct from (or more than) a claim on assets.

That question actually goes back to an important issue for Austrian economists, specifically Mises's regression theorem, which says (IIRC) anything that becomes money must have been originally valued for something other than money.

But regardless of that claim's validity, I think that your argument proves that money is a claim on assets differentiated from information. In the case of WoW, for example (and I may be getting this wrong; I'm basing this off of EverQuest), WoW gold was originally valued because it let you buy things from NPCs. So people would accept it as payment, knowing that they could at least use it to buy from NPCs.

As time progressed, people's valuation of how much e.g. time they could save from having better equipment caused them to be willing to spend US Dollars to get the WoW gold, and then to get the equipment. Over time, its use for buying from NPCs may become irrelevant, and people simply regard it as a unit of account that can be redeemed for goods -- effectively containing the information of the net debt owed to each person.

So that gold has become money without anyone warehousing an asset. Also, there's the money we currently used, which aren't backed by

The other topic's exchange is a bit lengthier, so I won't get to it till tomorrow.

Greg said...

I find it a little annoying that Kinsella starts with the insults the moment he starts losing the argument.

Anyway, to the discussion at the bottom of this:

I don't think you can have "money," as it is always understood, unless you have the money understood as information---some sort of intangible promise of value. It doesn't seem to me that copying paper money is wrong because you're copying the physical asset, but because the Information (that corresponds to some universally held "value") is being copied---this is what we consider Wrong.

As you said, it doesn't seem like it needs to correspond to any physical asset at all. The physical asset is just a placeholder. In MMO's, the little numbers on the screen are just placeholders. Same for my bank account. I mean, even if the bank information were tied to real physical gold somehow, the gold would be pretty valueless unless it represented that sort of intangible information of "value." So, if only I could somehow Copy gold (through some magic process), people wouldn't be mad because the gold was copied but that the Information---the value---was being copied.

This obviously rubs us the wrong way because we do not believe in copying value but only in Creating it. If I suddenly had a million dollars in my bank account (even if it were just a computing error), people would say things like "You don't Deserve that money! You didn't Earn it!" "Deserve" and "earn" just seem like synonyms for something like "value." And they are both intangible incidentally. (Do people even overstretch this idea to justify things like wealth redistribution, I wonder? That is, the claim that people born rich don't Deserve their wealth because they didn't really Earn it?)

As you pointed out, the trespass isn't really the issue, and I have no idea why Kinsella mentions it. Even if I stole Your gold to copy it in my magic gold machine, it's not the trespass people would be mad about. Though that trespass should require restitution (I could just return the original gold and provide appropriate damages like any other trespass---but the copy would be Mine following the anti-IP logic), the world would be angry because by copying value I'd be making everything else less valuable. Not only would that not be fair or just, it would harm everyone else, as my new value would diminish their value. The fact that there's this dispute and lessening of their value, as you have pointed out, points to something like scarcity---which anti-IP'ers don't claim exists with information.

I guess I'm not saying anything new here; I was just annoyed by that trespass red herring.

Silas Barta said...

Thanks for the comment, Greg. That's an excellent way of putting it, and I'm glad to know that this position isn't as controversial as I thought.

Stephan Kinsella said...

No.

gwern said...

> However, if everyone paid a fixed amount each year, and their choice of what to access only determined which fraction of that payment went to each creator, then choosing to view anything would cost nothing on the margin, further eroding any incentive to pirate.

Have you ever seen Wei Dai's proposal? http://weidai.com/measuring.txt