Sunday, June 15, 2008

IP and a Tentative Solution Involving Ghosts

Or perhaps BitTorrent.

After a particularly intelligent - if party-line-toeing - essay on the subject of software piracy, my dander is once again up. While I've previously laid out the basic issue, but in that piece my motivation was to be consistent, not to go for social engineering.

The basic fact is that IP law doesn't in the least reflect the reality of IP.

Shamus' essay doesn't either, instead going for the moral angle, instead of analyzing the facts first.

For instance,

"It’s like sneaking into a movie. Sure, it’s not “hurting” anyone - nobody becomes poorer by virtue of your viewing of the movie - and you are not depriving anyone else of the product. (We must assume the theater is infinite in size and all the seats offer the same view for this analogy to work.) But most people recognize that sneaking in is still wrong."

But it's not like that at all. It's more like Coke projecting a movie onto the moon, and then trying to restrict binocular sales so that no one can see it without paying.

Competition will always drive the price down to the risk-adjusted capital cost - input plus profit - and the input cost of most IP, including all digital data, is essentially zero.

How much does it cost you to transfer a movie across a high-bandwidth connection? That cost - something people do for free every day - is the actual input cost.

It is, as the economists would say, not scarce. Data is like air.

"Piracy is theft, and pirates are thieves, plain and simple. Downloading a movie off of the Internet is the same as taking a DVD off a store shelf without paying for it. In 2005, MPAA studios lost $2.3 billion worldwide to Internet piracy alone."

The basic point is that people have worked on it and deserve to be paid. But is it so? If I made a huge factory to produce air, I would certainly have done a lot of work, and perhaps I might do so if there was a law that supposedly guaranteed me a return, but would I actually deserve to be paid?

No. I am not providing anything of value. Similarly, a studio that produces a game or movie cannot justify charging more than a few cents for the data, and this is the basic reason piracy is rampant.

Unless my proposal is sound.

The first component is that you cannot restrict data sharing based on any moral premise. Because data can be copied essentially for free, it doesn't make any sense, and as we can see, it does not work in practice.

Therefore, I look at this as an engineering problem. How do I produce actual incentives to restrict data sharing? How do I produce a system where, if a studio and a pirate both have the IP, which is exactly the scenario we have for nearly every release, how do I make it so that the studio can recoup their R&D costs? How do I stop the pirate from automatically out competing the studio?

My answer to these questions is to adapt my view of IP to match my view of physical property. If I own a physical thing, like my computer, I may sell it to other people. Thus, if I own a piece of IP, like a game, I should be able to sell it as well.

Here's where you can see it's important to look at the practicalities. When I sell my computer, I cannot keep using my computer. Conversely, if I sell a game, I can keep using the game. In fact, I can sell the game an infinite number of times. Again, this is the basic problem that makes IP so very different than real property, and the problem underlying piracy.

So, if I adapt property law to allow the studio and the pirate to sell the game, the pirate has a specific incentive to restrict distribution - every time they reproduce the IP for free, they are forgoing earnings and letting their customer in on earning opportunities.

Here's where BitTorrent comes in. BitTorrent could partner with PayPal or similar, and straightforwardly implement a metering system. This would also solve the leeching problem - how often would you seed if you were actually getting paid for it?

Also, the reason the studio is allowing pirates to obtain the game is that they cannot stop them at present. Most likely the pirates are obtaining review copies or have a person inside passing them the data. These leaks are highly pluggable, but there simply isn't enough incentive to do so yet, as it only takes one pirate willing to buy a copy to undo all their hard work.

So that's the system in abstract. Now let's look at how it might work out more concretely, in two steps.

I start my analysis as soon as the studio has produce a piece of IP, for example, a movie.

They have a huge market advantage - they have the only copy in existence, and they can set any price on copies that they want. Supply is the minimum possible, which means that the price set by the iron law will be enormous. Maximal, in fact.

Unless this proposal does not actually work, or there's a non-equivalent proposal that solves the problem, this price is, and always has been, the break-even price on IP. Attempting to make more money than this is quite risky, and at present results in excessive piracy. I think, however, that this price should cover the budget of most movies. If not, it's impossible to produce a movie without risking bankruptcy.

The only problem is that the buyer cannot turn around and sell it for more - they were the person who valued the IP the most. There isn't anyone who will pay more. However, they can sell it to multiple people, for example three people at 40%, making 20% profit.

These second-tier buyers will, because they can, turn around and also attempt to sell the IP. Obviously they can't sell it for more than 40%, because they'll simply be out competed by the first tier. They must sell it for less. And so, as supply grows, the price drops.

So now I'm going to tentatively imagine an actual market.

Let's say the movie is Hellboy II. Guillermo del Toro completes his movie, costing (for the sake of easy math) $10 million.

Del Toro knows that if he attempts to sell it at $30, a typical DVD price, all he'll be doing is immediately jumping to tier N prices, and attempting to compete with Nth tier sellers. This will be, as piracy has taught us, disastrous.

Instead, del Toro goes to a dedicated distributor. This distributor is either a middleman or a big pipe distributor. I'll assume it's a middleman, because the big pipes are the same in either case.

The middleman pays del Toro between $15 and $60 million for the privilege of being first buyer, based on expected sales. Del Toro essentially washes his hands of the issue at this point, and goes home happy. Obviously there is risk here - but that is normal business. This proposal does not magically allow all the market value to be captured by the primary seller. If del Toro wants to be safe, he can sell the first one on the open market to the highest bidder, and make sure he doesn't produce movies costing more than this.

The middleman turns around and sells the IP to several distribution giants - corporations with huge internet pipes, for example one for each country. I'll assume del Toro sold the movie for $20 million. The middleman must sell for less than this, or the pipes can go direct to del Toro and get the same price. As above, they sell to three or more big pipes at perhaps $8 million. The pipes sign a contract not to release until a certain date, as is usual.

Then, on this date, they open the floodgates. Even if a few pirates have obtained copies, they will not be able to materially affect the huge data floods of the pipes. Similarly, if the pirates attempt to capture high-bid customers, the big pipes can do it faster and more efficiently.

The big pipes set any price they want, of course, so presumably sell first to luxury clients in an attempt to segregate the market, selling for perhaps $60. These are the first clients that actually want to see Hellboy. They will not necessarily make up their investment financially - the reason they're also allowed to sell Hellboy is simply to prevent them from flooding the market with free copies.

Once copies stemming from the $60 buyers reaches some critical mass, the big pipes will decrease their prices, continuing to reap vast profits. However, now they are competing with peer-to-peer networks, which are increasing supply and satisfying demand, rapidly pushing the price down.

I'm going to stop here to more fully explain how the PayBitTorrentPal program might work. It's fairly simple. The seeder sets an overall price, which is then converted to a per-bit price. Each leech pays the bit price for every bit they download from the seeder, paid into a PayPal like account.

However, like any free market, there will be a variety of so-called ask prices. Therefore, the leech will use an implementation of smart contracts. The leech will set a maximum price they are willing to pay, and PayBitTorrentPal will download accordingly. The outgoing pipes of the lowest askers will be filled first, as will the incoming pipes of the highest bidders.

The smart contract will also allow the seeders to dynamically adjust their prices based on bids without having to monitor PayBitTorrentPal.

Because each tier of buyers can resell the product, the higher tiers can ask much higher prices, knowing that the lower tiers can make up most of it, which allows the price structure I have described.

However, eventually, and in fact rather quickly due to exponential growth, the market will be flooded with copies of Hellboy selling for one or two dollars a shot. At this point, it's likely that someone will start distributing for free.

This is the natural cutoff point for release into the public domain. At this point copies will circulate as they do now, but legally.

The beauty of this proposal is threefold. First, it can implemented entirely within current property law. Second, it could potentially turn a profit for you, yes, you personally. Finally, it takes advantage of peer networks instead of trying to fight them.

Tweaks, corrections, and counter-proposals are all welcome. IP is clearly broken. Have you thought seriously about what we should do instead?


Anonymous said...

Did You Say “Intellectual Property”? It's a Seductive Mirage

I can't see that the technical side of your proposal prevents piracy, if a disingenous person obtains a copy of someone elses work and distrbutes it for free, they can still do this regardless of any payment system tied to bittorrent.

Bittorrent is not centralized, anyone can host their own tracker to distribute data, independant of what anyone else is doing.

Alrenous said...

Indeed, it does not prevent this.

However, our current situation is also not preventing this. My solution at least puts all the incentives in the right direction. If someone wants to give up legitimate revenue, there's little further incentives of any kind can do anything about it. I would expect such nose-biting/face-spiting people to be a tiny minority, not capable of strongly influencing the revenue stream.

Unknown said...

Why would any of the producers be inclined to use this instead of the traditional form? They would make less money this way, wouldn't they? Or at least, that's the way it seems...

Anonymous said...

One major problem: Quite a few pirates are simply not motivated by money - they try to compete with each other to see who can get their files downloaded most, as a form of social currency. Plus, The altruists will immediately drop the price down to zero, within a day if not minutes. There would not be enough time to make back all the money invested.

Alrenous said...

Ask Geek

Because piracy is bankrupting the PC gaming business, for instance. They have a chance of making more money the traditional way, but practically speaking they're actually going to make more money this way.

Major Anonymous

You assert that the price will be driven down to zero.

I will hereby use your accepted technique and assert that you haven't read my entire proposal, because I already addressed that.

Regardless, the price is already being driven down to zero. My proposal does exactly no harm even if I'm wrong.