Brink Lindsey and Daniel Takash write,
Eliminate patents for software and business methods. One crucial requirement for any workable system of property rights is the ability to define boundaries. The expansion of patent law in recent decades to include software and business methods runs afoul of this requirement. These patents currently make up a substantial fraction of all patents granted, and instead of serving to encourage innovation they have created a legal minefield that innovators now have to cross.
This proposal sounds sensible, as do the others that they offer.
When I think about this issue, I think in terms of two questions about the originator (who claims intellectual property in our current system) and the follower (who is prevented from using the originator’s work).
1. How much effort did the originator have to go through? In the case of pharmaceuticals, these days the effort required is usually a lot. In the case of new business methods, more of the effort goes into the execution than into the conception.
2. How much effort did the follower not have to undertake because of what the first inventor did? Again, the case of a pharmaceutical, this is a lot. In the case of a new business method (Amazon’s “one-click ordering” is a widely-used example), I would say not so much.
I guess the ideal for me would be if the follower had to compensate the originator according to the answers to these two questions.
One of the advantages of the intellectual property scheme as currently existing is that it actually captures all ‘useful inventions’ including any category of invention never yet imagined. It doesn’t, on the other hand, cover works of art, basic discoveries of science, or trademarks (the disaster of copyright is another discussion having to do more with length/longevity than scope). I don’t see a reason why patents shouldn’t continue to cover many kinds of software (not art or literature embodied in software, which would be copyright) – however, perhaps the ‘reduction to practice’ should be more tightly constrained. Business practices seem to be something uniquely tricky and I’m not quite sure how to consider them.
I’m not sure how to consider business practices either but I’d argue that it is very hard to distinguish between truly innovative and easily copyable software/services and business models/practices. What seems to be missing is a good model that can distinguish between novel innovation and good execution in a novel environment. The current system seems to overemphasize the first-mover and conflates the innovation process with the patent filing process.
I think intellectual property is very real and worth protecting but it is unclear to me that patents represent anything other than tactical weapons that large corporations use to wage or deter legal warfare. Perhaps patent law is another example where the gaming of a mature system has negated its original value proposition.
Open source is a common practice to defeat the patenters. Make the innovation a cut and paste piece of software. Then there is no patent issues, it becomes prior art. Open software industry has been quite successful, customers do not need fear patents, they can always say they use existing software.
I think that oversimplifies things.
For example, if IBM released an Apache-licensed implementation Y of patented algorithm X, users of Y and works derived from Y get patent rights from the Apache license. But users of a 3rd-party GNU-licensed (or, alternatively, public domain) implementation of algorithm X still need to license the patent for X.
And of course just because I found something on the internet (or on a scrap of paper in the street) does not mean my use of it is unencumbered by patents.
The idea is to open source it before the patents are filed.
The patent filer has w worthless patent, an alternative implementation is already public. Mathematicians often do this, the actual intellectual property is in the theory, any implementation can be derived from the proof.
I think tests (1) and (2) are fine in theory. But.
Say Alphabet tries 2983 different machine learning algorithms and finds algorithm number 1984 has very good performance and broad applicability compared to others. They patent the algorithm, they use it themselves internally for some stuff, and they describe it in an academic paper. These algorithms tend to be pretty simple in themselves; 3 months later, an estimable 3rd-party academic creates and releases an Apache licensed implementation of his own, done from the starting point of the published paper.
It took one developer 3 months to implement the algorithm from the paper, but it took 1000 developers 3 years to choose it from among others. (In hindsight, it need not have taken that long, but they had the money at the time.) How does one estimate the amount of effort Alphabet has saved me when I build a product using a tweaked version of algorithm 1984, which I basically took off the shelf?
Can 10 million lawyers approximately solve this problem in the next 1000 years?
I copyrighted the idea of eliminating business method and software patents in 2016.
I patented the patent.
I found some prior art in a 10,000 year old painting on a cave wall in France, it’s public domain.
Am I doing this right?
I am surprised you like this article. It is utopian and while claiming to address real-world problems is completely detached from reality. It is very much in line with the academic mindset that there are infinite resources provided by loftier human instincts (or, in the exception that proves the rule — unfettered greed overcoming any elastic effects) that is freely permeating eg Warren’s campaign.
Look at this choice piece: “A more fundamental explanation lies in the dominance of nonpecuniary motivations in artistic creation. The overwhelming majority of creative works don’t sell much, but the intrinsic pleasures of artistic self-expression are so powerful that people will engage in creative pursuits regardless of the economics. ”
Truly? This line of reasoning is along the lines of speculating that people get so much enjoyment out of buying a lottery ticket there is no need to actually pay anything out.
And second, ‘engage in creative pursuits’ just sweep a ton of problems under the mat: professionals and amateurs are miles apart — if in doubt, go listen to Vengerov play and then go to your local high-school where kids who are exceptional and have taken ten years of violin play, and you will still be astounded by the difference. And that extra investment he made did not come free nor part-time.
” As society gets richer and more people have the leisure to engage in creative activity, and as new technologies continue to drive down the costs of self-expression, we can expect an ever richer bounty of cultural works no matter what the economic payoffs for a lucky few might be. Strong copyright protection may be important for bolstering the profits of giant media companies, but it is far from clear that it is needed to secure a vibrant cultural marketplace. ”
Sure, except ‘society becoming richer’ is ridiculous — Bill Gates making more money makes society richer but riches do not just distribute themselves to everyone. This picture realistically will play out like this: a Wall street banker makes his money in a lucrative specialty and gives it up at age forty-five to make … cruddy movies. Meantime no one spends years honing their creative craft in their twenties or thirties, because, as much as creatives tend to do a poor job of doing a proper accounting of probability of financial success, even the most mathematically obtuse understand the difference between zero and something larger.
I do not think the quality of culture will be pushed by more out-of-focus cat videos, nor mathematical research by poorly written tutorials by someone who just yesterday discovered there is a dimension beyond pointland.