|
| The cult of the inventor |
|
navigational aids: News ticker:
|
02 November 05.
Many think that inventors have an ineffable je ne sais quois which the rest of us don't. They should be rewarded for having It. The handy OED explains that "genius" shares the same root as "genie" and "genesis": it is a spirit breathed into certain individuals at birth, which allows them to do great things without effort or prior training. In either world, the question of how people get rewarded for their ideas remains. Now, it is hard to find examples of people who are not in any way enriched from innovations that they were the first to write down. Employers and entrepreneurs spend a lot of time looking for people who are better than everybody else, and a world-moving product is pretty good evidence of exceptional ability. This guy who hires programmers explains that programming ability is not linear, and you should pay a great programmer a whole lot more than a good programmer. This is especially the case in the incrementalist world, because Joe wrote down the idea first because he worked longer hours or thought a little faster than the next guy, and Joe is likely to continue doing so into the future. In the revolutionary world, Joe was visited by angels, so the invention on his resumé provides no evidence that he deserves the big bucks from new employers. So in the revolutionary world, we need to make sure that Joe gets paid for the revolutionary idea, which is outside of Joe, not for his efforts or labor, which he will continue to carry with him. In the incrementalist world, he'll get tenure somewhere. The next question is the rent-splitting question. Jane invents new device, ManuCo manufactures it for a dollar, MarketCo advertises and distributes it for a dollar, and consumers consume the item for $100. That gives us $98 to split between Jane, ManuCo, and MarketCo. How? This question is entirely outside of economics. On the ethical side, we all say that Jane should get a 'fair share', but as with any ethical position, what is fair is hard to define. The guys at MarketCo are expending a great deal of effort in perpetuity, whereas Jane may have made her entire contribution over the course of a few seconds. In the incrementalist world, the moral weight of her contribution wasn't that huge, since the guys at ManuCo could have put the puzzle pieces together themselves if they'd put enough effort into it. But in the revolutionary world, Jane's je ne sais quois was a unique and essential ingredient to production, and a drop of it is included in every unit sold. A few weeks ago, I had to do a little lit review of the books on patents at the local Borders and Noble (I was asked by a publisher to review a book proposal). There were some that were very direct and honest, mostly consisting of pictures of forms and occasional discussion of legal strategy, but many were imbued with the cult of the inventor. They explained that now that you, dear reader, have rubbed two brain cells together to form an idea, you deserve riches, and they could be yours provided you file the right forms to ensure that the bastards don't steal your idea. Notice how much this sounds like every other get-rich-quick scheme: you're the best, so there's no need to work—it's just a question of getting the system to work for you! Maybe I'm just not cool enough, but I don't know anybody who has built a career off of pure ineffable genius. I do know a lot of people who had a good idea, which was not necessarily a unique idea but was a step or two ahead of its time, and who put a great deal of work into writing products, pitching products, and generally working hard to turn the fugitive fermentation of an individual brain into a career. But wouldn't it be great if you didn't need all that? If you could just have the idea and let somebody else do all the work? The attractiveness of the concept means that the cult of the inventor won't ever go away, and the media will always provide new tools to fuel fantasies of money for nothing. Back to softwareIt is impossible to determine what is revolutionary and what is incrementalist, so the Patent and Trademark Office doesn't try. It is entirely rooted in the philosophy of the revolutionary inventor, and if your idea looks new, it is revolutionary. In an ideal world, we'd be able to ferret one type of invention from another; in the world we have, the best we can do is search for certain proxies, such as, oh, subject matter restrictions. I could not imagine a more incremental field than software. Everybody builds on everybody else's code base, and there hasn't been a true-and-honest paradigm shift in decades. What, object-oriented programming, where you can put your functions inside data structures? (hint: I put pointers to functions inside data structures using K&R C.) Storing data in XML instead of ad hoc data structures? (hint: XML is a special case of SGML, which is a generalization of HTML.) The desktop metaphor? (hint 1: it's a metaphor; hint 2: it was all done at Xerox in the 70s-80s.) For many decades now, people have been talking about writing down something truly different from the function-oriented paradigm embodied in C, and nothing has materialized but nifty trends and syntactic sugar. The nifty trends are indeed nifty and do make our lives better, but they've all been incrementalist in nature. Why is software exceptionally incrementalist? First, I would argue that to a great extent, it isn't exceptional: all research builds upon research. If your chemistry paper doesn't have a lit review, it won't get published. Second, the structure of code is incrementalist: if you want to write a program of nontrivial length, you need to use preexisting libraries to do the lower-level work. That is, everybody who wants to write a program is staring at the same set of tools, trying to assemble them in a new and better manner, and this is a recipe for incrementalism. That's pretty far from materials science, where your new material may be a combination of old, but is as likely to bear only distant relation. Some computer geeks are working on the truly low-level data manipulation routines, but they're mathematicians, whom the ethicists feel are not inventing but discovering, so they can't lay claim on their work. So the conclusion from this perspective is the same as every other perspective: patents shouldn't apply to software. If you're a smart and hard-working programmer, you'll get paid, without the right to exclude others from the code base you wrote down first. Because your code base did come first, and you're the sharp one (and don't forget the lock-in and network effects that are prevalent in software), you have a good chance of making a whole lot more than the guys who show up second. Whether you deserve to make more than the guys who wrote down the idea second is an ethical question with no answer, but in an incremental world like software, it's hard to make the argument that you not only deserve to make more, but deserve the right to tax the second-comers for using what is now declared to be your invention.
[link] [No comments] |
