For along time, a standard component of my war on the current intellectual property system has been to point out how incredibly absurd our current laws will appear when computers can, with very little direction, churn out patentable and copyrightable material. That day is now. Popular Science has an article on a man(John Koza) who is using what is, at its heart, an incredibly simple machine(if composed of about 1000 cpu's) to brute-force new ideas until an 'innovative' one is found. One such idea has actually been successfully patented, one of the first ever patents granted to a machine designer(in reality, of course, the patent was granted to the human who uses the machine).
What if(as is likely within a short period of time) your average Joe has a computer of this power factor in his home and is able to run a similar program on it? Who gets the intellectual 'property' that is 'produced' by it? Does credit go to the program's designer, or to the man who typed in "find a good engine design, please, mr. computer."???? The natural progression of technology in this area will quickly blow our existing intellectual property system to shreds. Don't believe me? Take note of this: before designing the optical lense that was ultimately patented, Koza's beowulf cluster duplicated scores of already patented designs by starting from simple parameters with a simple goal(at no time was any portion of these designs in any way inputted). It seems patent offices everywhere are going to be scrambling to do the one thing I've been pointing out they are unable to do: define what exactly 'obvious' actually is.
Patent laws require patentable ideas to be 'non-obvious.' What that actually means nobody has the foggiest clue. Is a design obvious if a simple computer program can generate it automatically given a basic understanding of the target context? How about a complex program? How about a complex program run on a complex machine? How about a complex program run on a complex machine that is in everybody's offices and homes? The reality is that without the ability to pick and choose individually who will receive the incredible economic incentive that our intellectual 'property' laws actually are(and in most cases what we need here is a good weeding) our current patent and copyright laws are fundamentally meaningless. Unless we can redefine the transfer of intellectual 'property' as the service it actually is, our entire system of intellectual 'property' law is likely to end up in the biggest mess we've ever seen.
Monday, April 24, 2006
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment