>has always been and continues to be the special, important, and valuable product that publishers provide. I'd have to disagree with that. While I definitely

>has always been and continues to be the special, important, and valuable product that publishers provide. I'd have to disagree with that. While I definitely think aggregation is a valuable service, it has never been able to hold a candle to distribution. Without distribution, a work could only be sold or acquired in extreme geographic proximity to where it was created. Distributing it far and wide is what creates the entire market, and makes it possible for publishers to exist. Once anyone at all is able to distribute things far and wide, however, that's no longer valuable and the publisher will hopefully simply go away. As for aggregation, there are 2 issues I see with seeing this as very valuable - first, publishers are terrible at it. See the book "A Drunkard's Walk" for some statistics on the matter, but the bottom line is that no one in publishing is capable of telling the difference between what will be a success or a failure. 9 publishers passed on the Harry Potter series before one 'took a chance' on it. It's not really their 'fault' exactly, it's just that the public responds to entertainent products randomly and anyone who claims to have noticed a pattern in the noise is a (usually self-deluded) charlatan. So, from that aspect, publishers cannot offer anything superior to what anyone else can offer. Secondly, machine-learning-based recommendation systems such as Amazon, Netflix, etc have completely automate the aggregation process and reduce the value of aggregation nearly as much as the Internet has reduced the value of distribution. >They also give them a soapbox that has a big stamp on it that reads, 'trust us, she's a great writer. All this stuff is true.' If there were evidence that publishers had an accuracy rate greater than random chance at this (there isn't), this would certainly be something they could profit from.