"Paper has been traced to China in about AD 105. It reached Central Asia by 751 and Baghdad by 793, and by the 14th century there were paper mills in several parts of Europe. The invention of the printing press in about 1450 greatly increased the demand for paper."*

Paper has endured for almost 2000 years, printing for more than 500 - and they're both still large, vital industries. Think about other technologies that have endured that long: then think that computers and networks actually drove greater paper and printing consumption - printing out emails chews up a lot of trees and is a big reason that Epson and HP and a dozen other huge businesses exist.

Dave recently wrote about the incredible drop in publishing costs in the last decade: from printing press to laserwriter to Web server to blog. It really is amazing: not only has it become cheaper, by several orders of magnitude, but the lag time approaches zero, unlike printing, where even the best distributors can't manage much better than a day's delay.

Web servers are a classic industrial innovation: so much cheaper, so much better than previous technology that it will take people a generation to catch up with them. Thus it was with steam, internal combustion, electricity et al.

Web servers won't kill printing: new media almost never kill old media - they just change how they're used. Web and email have already moved printing from something that mostly happened in a factory, to something that happens in offices and homes.

And printing still has a killer advantage: you don't need any technology to access printed pages. No compatibility problems, no down time due to ISP troubles, no browser bugs. Printed pages just work.

But, industrial processes that offer orders-of-magnitude reductions in cost inevitably set the world on its ear. Printing did. I don't think that the Renaissance and Industrial Revolution just happened to occur shortly after printing (and postal service) arose.

Blogs are a link in that chain: they represent, as Dave suggests, the reduction of costs of publishing to near-zero. The fallout is likely to be enormous. Anybody know any bloggers who use free computers - in schools or libraries - and use a free blogging service?
Comments 10:31:16 PM    

In part 1, I offered the opinion that the important part of intellectual property is the order of the 'atoms' - the basic units, be they musical notes or bytes of code.

So how is it that some people come to find the uncommon, and unusually valuable patterns, and many don't? We are all, after all, walking around with the most evolved pattern-recognition system on the planet, and maybe the universe, sitting on our shoulders.

Part of the answer is that the usual noise operates here: some are lucky, some are more gifted, some people discover great things at the wrong time, some discover mediocre things at fortuitous times, some have better marketing etc. etc. So understanding that order is the magic ingredient doesn't just solve the problem of creating value out of thin air, which is the essence of IP.

What it does suggest is that there may be non-intuitive strategies for uncovering the most useful orders. It might, for example, be informative to try large sets of different orders in a Darwinian competition to see which ones seem to fill the bill best - whether that means 'appealing to teenagers' or 'crunching numbers faster and more accurately'.

Indeed, there is a music service now that compares new songs to a database of hits and tries to divine if the patterns in the new tune are more or less like songs that have been successful before. And it's not likely IMHO that invention will be able to proceed based strictly on automated processes trying large numbers of things.

The state space of almost any field - music, literature, code - is incredibly huge, and not susceptible to a 'brute force attack' in any timeframe that is likely to be useful - like a human lifetime. But it is possible that our superb pattern-recognition machinery could do better with some tools to help cull the field down to a more manageable size. It would be interesting to see what a meta-pattern software tool would look like... one that could compare any construct against any other...is there a common thread between, say Linux and Mozart?
Comments 10:11:56 AM    

Chris Gulker's view from Silicon Valley - in words and pictures