Nobel Prize Winning Economist Eric Maskin: In Highly Innovative Industries, It May Be Better To Scrap Patents
from the good-point dept
We've talked in the past about how Nobel Prize winning economist Eric Maskin has done research questioning the value of patents in software and in other industries. So it's not surprising -- though still fantastic -- that he's responded to the recent NY Times piece by Charles Duhigg and Steve Lohr that questions the value of our broken patent system. Maskin has sent a letter to the NYTimes arguing that the article did not go far enough:...the article doesn’t go far enough.His conclusion? In such industries (which, by the way, may go beyond just software), we should probably just get rid of patents altogether:
It argues that software patents may reduce innovation because they are too broad, vague and loose — criticisms that are well taken.
But if these were their only shortcomings, then simply tightening patent standards would solve everything. Unfortunately, the problem is deeper than that.
Specifically, in the software industry, progress is highly sequential: progress is typically made through a large number of small steps, each building on the previous ones. If one of those steps is patentable, then the patent holder can effectively block (or at least slow down) subsequent progress by setting high license fees.
...in an industry with highly sequential innovation, it may be better for society to scrap patents altogether than try to tighten them.This view is not new in economics circles, but it's good to see more people realize that the patent system may be doing exactly the opposite of what we're told it does.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: eric maskin, innovation, patents, sequential innovation, software patents
Reader Comments
Subscribe: RSS
View by: Time | Thread
Contrary to what most critics seem to think about me I do think there should be a protection system in place. The issue is how to make a balanced and fair one. The current one is clearly not.
[ link to this | view in thread ]
While I don't disagree with THIS opinion, that's all "economics" is.
http://exiledonline.com/the-nobel-prize-in-economics-there-is-no-nobel-prize-in-economics/
[ link to this | view in thread ]
Re:
The whole system probably needs to be ripped out and started again from scratch at this stage.
[ link to this | view in thread ]
Re: While I don't disagree with THIS opinion, that's all "economics" is.
[ link to this | view in thread ]
Re: Re: While I don't disagree with THIS opinion, that's all "economics" is.
[ link to this | view in thread ]
[ link to this | view in thread ]
Re: Re: While I don't disagree with THIS opinion, that's all "economics" is.
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
Bad patent
http://www.google.com/patents/US7028023
This patent illustrates just how badly the patent system is for software. Any programmer know that pointers are used to represent relationships between things.
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
Removing patents MAY also destroy true innovation. MAY.
[ link to this | view in thread ]
Re: Re:
But I also agree with you in saying that the time it takes to something fall out oiu copyright is way too big.
So, to me, AFTER a copyright reform, softwares should be protected under it.
[ link to this | view in thread ]
Re:
Perhaps, but the patent system is not supposed to be a protection system - it is supposed to be an incentive program for furthering progress. It should only be allowed to be used for that purpose.
I'm no constitutional scholar, but I don't think there is anything in the constitution that indicates our legislators have the power to create industry/business model/company protectionist systems simply for the sake of protecting them.
[ link to this | view in thread ]
Re: Re:
It makes a lot of sense to me (as a software developer) that I should have the ability to prevent people from making copies of software (or source code) that I wrote, but preventing someone from writing a program that does the same thing is lunacy.
[ link to this | view in thread ]
Re:
That depends on what the definition of 'is' is.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
Software is the only thing currently protected by copyright AND patents in the US.
[ link to this | view in thread ]
Re:
However, as I do not expect any government to agree with me on the first two, I'd be willing to accept severely reduced terms of such policies as a compromise: say one year.
Anyway, feel free to direct all of your critics' vitriol in my direction.
[ link to this | view in thread ]
Re:
Why? Any sort of 'protection' system is also a 'eliminate competition' system. Why should anyone deserve protection from competition, and how does that protection provide a net benefit to society?
[ link to this | view in thread ]
Re: Re: Re: Re:
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Bad patent
[ link to this | view in thread ]
Re:
Legal protection on innovation is unnecessary and counter-productive. The most innovative people will never have a problem coming up with something new to make money on, so it does not truly serve them. It can only serves to create property the corporations can own and cases the lawyers can charge billable hours for.
[ link to this | view in thread ]
Re: While I don't disagree with THIS opinion, that's all "economics" is.
[ link to this | view in thread ]
Re: Re: Re:
[ link to this | view in thread ]
Re: Re:
But I do think there is an appropriate protection mechanism available: copyright (as it was, not as it is). Protecting the expression of the principle, rather than the principle itself, seem like the right thing.
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
Re: Re: Re:
Why?
[ link to this | view in thread ]
Re: Re: While I don't disagree with THIS opinion, that's all "economics" is.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Bad patent
[ link to this | view in thread ]
Re: While I don't disagree with THIS opinion, that's all "economics" is.
However, the economic price - albeit not really a nobel price - is a sign of good scientific work.
[ link to this | view in thread ]
Re: Re:
Show a mathematician a statement like "x = x + 1", and he'll say that's absurd; there is no value X such that X is equal to X + 1. But a programmer recognizes that as a very simple and common operation: incrementing the value of a variable.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re:
I'm not sure there has actually been anything innovative with software EVER. All "new ideas" for software are just an expression of what is already known by the community at large.
Leave software with copyright, so source code can't be strait-up copied, but leave patents out of it.
Anyway, math and facts aren't supposed to be patentable and software is nothing but math and facts.
[ link to this | view in thread ]
Let me toss this out
One negative potential of dropping patents altogether is the possibility that a small company or impoverished inventor might come up with an idea and then a big company with huge amounts of resources will beat it/him to market. No protection for the little guys.
But let's say we reinvent the world economic system so that it is so decentralized there is no longer any competitive advantage to scale. As much as possible every household, every small company, and every small community can deal with their own needs without using the services of major corporations. With each small unit meeting its own needs, the corporate system declines in usefulness and there is no longer a fear that corporations will "steal" ideas and the idea creators won't be properly compensated.
Again, if you want to see where this kind of thinking is headed, follow along with what gets posted here.
P2P Foundation
I plug this organization so much that many of you might assume I have some connection to it. I have none. I'm just looking at new business and economic models and this has proven to be the best single source of info and discussion I have found.
[ link to this | view in thread ]
Re: Re: Re:
If you're in the auto industry, you don't design cars to then patent the design and sell it. You design a car because you intend to build it. This works well for industries whose main interaction with customers is a physical part.
If you're in the software industry, you do not write programs because you intend on representing the code onto pieces of jewelry that are then sellable and wearable, you write code so that it can be copied and used. Legal protection in such a case offers the tempting possibility that your code becomes a physical thing that can be controlled, but this is an illusion. Computer code (compiled or not) is not a physical item, it's virtual, it's a digital extension of human thought. Code is valueless when not in an executable form. This means that in software, the idea itself *is always* the same thing as execution of that idea.
Am I explaining the problem clearly?
[ link to this | view in thread ]
Re: Re: Re:
y = x + 1
then another operation
x = y
[ link to this | view in thread ]
Re: Re: Re: Re:
Maybe you can patent the whole final result of your project, not specific parts. IE: you can't patent the shaft and piston setup but you can patent specific optimizations to that setup in an specific model of engine. If another company wants to build that specific engine exactly as it is to be part of a bigger project then they'd need to pay some predetermined reasonable fee. I may be saying something absurd though, I'm not specialist.
Also there should be a mechanism to allow independent invention.
Software should not be patentable. If I build a software that does exactly what Autocad does then too bad, it's my right. Maybe you could have some sort of copyright, not patent but it goes beyond what I can analyze.
Again, I'm not an expert but I think it is possible. Maybe there is no silver bullet and each sector could be regulated accordingly. My concern is that there are ways to abuse and exploit creations commercially and fck up the one who created. How can you address that without locking up the market? I personally don't know the answer and that's why it needs further and careful discussion, not some congress clowns that never saw technology doing whatever.
[ link to this | view in thread ]
Re: Re: Re:
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
The equation is a simple iterative process that is not uncommon in math. A mathematician would put a lowered i+1 after the first x and a lowered i after the second x.
Your argument is basically based on semantics.
When it comes to patents, however, it is usually on something able to change the behaviour of another thing or avoiding certain effects. That seems to be close to what a virus scannner, firewall, browser, spreadsheet, writer etc. etc. does. Except for the innovation stunting and how new IT is, are there other good reasons not to have patents. When enough math is collected in a single place it could be called an invention. The use of state as in x+1=x1 for chars and x=2 x+1=3 for integers is the real argument here.
[ link to this | view in thread ]
Re: Re: Re:
No. A mathematician will merely point out that whatever is being represented by the two occurrences of the symbol "x" must hold different meanings within the syntax of the expression. A mathematician (or a Lisp programmer) might object to this choice of syntactic representation -- and instead propose that subscripts be used to denote that the "x" is representing a changed state (yes, Virginia, mathematics has a fully developed conception of "state") or that the first "x" be "hatted" or "roofed" with a circumflex (old school style) -- but a mathematician would find no absurdity in the expression being the statement of a mathematical function.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
Re: Re: Re:
When any program runs on a computer or other type of electronic device, it is doing a set of mathematical operations on a set of data that is represented by mathematical values. There may be levels of abstraction between what you're seeing and what is happening, but that's what happens at the machine code level.
Show a mathematician a statement like "x = x + 1", and he'll say that's absurd; there is no value X such that X is equal to X + 1. But a programmer recognizes that as a very simple and common operation: incrementing the value of a variable.
That's a snytax or abstraction issue. You might as well claim that either a British person or an American aren't speaking English because when they see 'boot' it means different things to each.
[ link to this | view in thread ]
Re: Re:
[ link to this | view in thread ]
here is my detailed explanation of how it should work
case 1. Two almost equally sized companies working on the same product(s). Seems fine. Let them compete on execution. The one to think of it first has a head start already, they don't need to be granted exclusive rights.
case 2. Smaller company tech is reverse engineered by large company and then out competed in the marketplace by the larger company with more resources. This case is the only one patents help with and should cover.
It seems like we do need to allow some sort of limited ownership of tech to prevent case 2 from stifling startups. So for patents to work as a benefit to society, they need to operate as follows.
1. Startup company files a patent to give them exclusive right to develop a new product they thought up first. Patent includes deliverables and development timelines. Company pays a yearly fee along with updates to the timelines. If timelines are missed by more than 1 year on each update, patent is invalidated. (This to make sure progress to market is actually being made.) Company is given 1-5 years exclusive market rights after the product is delivered (depending on the product) to recoup costs and establish a market.
2. An established company (one with at least one product on the market and revenue) can file a patent only to keep (1) from preventing them from developing a product. That is, since (1) allows startups to block an established company from entering a market for a period, if the established company thinks of an idea first they need a way to keep that from happening. Patents from an established company do not require a development timeline since they cannot be used to prevent anyone from developing the patented technology.
As far as I can tell, this scheme would fix the patents system to be beneficial to everyone that actually does something. (ie. not lawyers or politicians). There is room for the non-practicing inventor to file a patent (without development timeline) and then sell it within one year, either to a big company, or to a startup (that then needs to add a timeline to it and refile.) See any holes in it? Why wouldn't this work?
[ link to this | view in thread ]
It's totally sequential, never copyrighted, usually shared (always eventually shared), techniques are copied wholesale with credit (socially enforced), later they become standard, knowledge grows like lightning, people do it for glory and a steady job.
No need for copyright.
(Indeed, I used to copyright my articles in the sense that journals wanted that, but it seems irrelevant today.)
[ link to this | view in thread ]
Re: Re: Re:
But there are real conceptual differences between various mathematical sciences, such as:
pure math (static and, well, pure)
classical applied math (modeling)
statistics (epistemology)
theoretical physics (e.g. string theory has an Alice in Wonderland element)
programming (like you said, state counts. plenty of other fresh concepts too)
It isn't just a matter of what "is" is, i.e. what = means !
[ link to this | view in thread ]
Re: Let me toss this out
Eliminating IP protection is appealing to me if it is accompanied by significant flattening of ownership. Let everyone share in everything as much as possible. What's different than the communism days are digital networks, digital products, and increased productivity due to technology. That which couldn't easily and equitably shared in the past is more doable now. Of course, we'll still be fighting over land, water, minerals, and fossil fuels until we figure out how to reduce scarcities of those, but in other areas, maybe abundance can be achieved.
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
Re: Re: Bad patent
[ link to this | view in thread ]
Re: Re: Re:
You're incorrect. State is something that is integral to higher mathematics. So are concepts like conditional branching and looping.
[ link to this | view in thread ]
Re: Let me toss this out
Patents don't stop this. Patents only protect you against entities that can't outspend you. If a big company wants to rip off your idea, they can do so easily whether or not you have a patent.
[ link to this | view in thread ]
Re: Re: Let me toss this out
I'm a bit confused. If patents don't stop copying, then what's the problem with them? Let's ignore them and move on.
[ link to this | view in thread ]
Re: Re:
Is that why Microsoft is a billion plus company and Ubuntu is still not making enough to pay development?
It seems to me that there is no conclusive demonstration of anything, except that flailing around in the free software world flattens all of the grass in the field, which may make you think someone actually mowed the lawn.
[ link to this | view in thread ]
Re: Re: Re:
Proprietary/legacy databases and software solutions.
I'd even bet a lot of regulatory software is much older than 10 years (air traffic control, power stations...).
[ link to this | view in thread ]
Re: Re: Re:
Free software has its place, although it's unlikely to be a panacea.
[ link to this | view in thread ]
Re: Re: Re: Let me toss this out
[ link to this | view in thread ]
Re: Re: Re: Let me toss this out
[ link to this | view in thread ]
Re: Re: Re: Re: Let me toss this out
Then that is also the problem with trying to get rid of patents. If patents benefit big companies and keep out their competition (I don't dispute this), and the big companies are the ones with the money to lobby Washington, then it is unlikely they will use that money to ask for significant patent reform.
Calling for patent elimination or reform without addressing the bigger economic and political issues in the world probably isn't going to get us very far. And that's why I've never focused on IP issues per se. I'm more interested in seeing much bigger changes and would probably start elsewhere first. Trying to get IP laws changed when the money doesn't support it is going to be a challenge.
I think the only reason we've had so much discussion about copyright laws is that a big company (Google) sees copyright as an impediment to its own business. Google wants to post content without hassles and copyright gets in the way of that. Otherwise, I don't think Google's lobbying money would bother to get involved in this issue.
[ link to this | view in thread ]
Eric Maskin WIns Ignobel Prize for 2012!
It is also true that all code like all invention is made up of pre-existing elements combined in new ways to obtain new or more effective results.
What is absolutely false and has earned Eric his Ignobel (the anti-Nobel prize unfortunately fully negating any presumption of wisdom earned by his original Nobel) is that there is no novel software key to innovation and therefor there should be no software patents.
As an example let me offer the Blackmun-Tuckey Fast Fourier Transform and its related cousin the ultra-high speed Invert FFT spatial filter for N dimensional spaces. Both of the latter deserve patents. Only the more recent Invert has received one. Like any real algorithm either of these subroutines can be written in different orders and pieces so as to overcome copyrights. Both of the latter pieces of software are components that can be used (and hidden) in others programs and run 1000's of times faster then a conventional sine approach to a sequential construction of an FFT or Inverted Weighted FFT.
Good software patents enrich the community of resources much as new logic and analog ICs enrich the resources for the creation of whole classes and fields of electronic applications.
We should be trying to convince the USPTO to improcve the quality of software patents not throwing the baby out with the poorly examined badware.
[ link to this | view in thread ]
more dissembling by Masnick
[ link to this | view in thread ]
another biased article
As to the quality of patents; based on court rulings of the last several years, roughly half of all litigated patents are upheld in court. That's pretty balanced and suggests there is no problem with patent quality. Further, seldom do cases ever make it to trial as the parties settle out of court. The facts do not support the contention that there is a patent quality issue. Still, with almost half a million patent applications filed each year a few are bound to be issued that shouldn't. However, rarely are they ever an issue because you can't enforce them without money and you wont get the money unless you have a good patent. Keep in mind it costs the patent holder about as much in a patent suit as it does the accused infringer. Investors are not stupid. If they don't have confidence in your patent, they will not invest. It's that simple. Bad patents do not get funded.
When corporate America agrees to not use our inventions without consent, American inventors and small entities will agree to stop suing them.
[ link to this | view in thread ]
0.1%
Most programmers are not really trained as engineers. Because of this they are not able to to see the bigger picture that hardware and software can both be used to implement an invention and in many cases a software invention can be covered with hardware like claims. Before software became patentable, it was common to craft patent protection for such as a broad hardware claim and then argue doctrine of equivalents.
I started programming around 1964 at General Motors Institute on a timeshare system in Algol (precursor to Basic). I was in 8th grade at that time. The next year I learned Fortran on an IIBM 1130.
I used a minicomputer, Alpha 16 in 1972 to run an automated disco lighting system with a 12' x 60' sound modulated 105 channel dance floor. This was done with FFT. I started playing with the 4004 as soon as it was available.
By the mid eighties I was using 6500 and 6800 family micros in multiprocessor configurations to run real time control systems.
While my primary interest was hardware design, I quickly discovered that if I wanted to use microprocessors at that time that I had to write my own code.
The end result is that I have extensive analog and digital engineering experience and also programming from raw machine code to over a score of high level languages and hundreds of dialects.
The difference between that time and today is that specialization has left the software industry mostly devoid of people who have the kind of experience it takes to understand how software and hardware are often interchangeable and that inventions can be implemented both. It is ridiculous to argue that software should not be patentable.
Ronald J. Riley,
Affiliations:
President - www.PIAUSA.org - RJR at PIAUSA.org
Executive Director - www.InventorEd.org - RJR at InvEd.org
Senior Fellow - www.PatentPolicy.org
President - Alliance for American Innovation
Caretaker of Intellectual Property Creators on behalf of deceased founder Paul Heckel
Michigan * Washington, DC
Direct 810-597-0194 * 202-318-1595 - 9 am to 8 pm EST.
[ link to this | view in thread ]
Re: Re: Re: Re:
My point was the almost irrelevance for the length of copyright on software if 37 or 130 years. Windows XP is perfect for my point.
[ link to this | view in thread ]
Re: Re: Re:
[ link to this | view in thread ]
Re: 0.1%
[ link to this | view in thread ]