I agree with Kelly that the Internet is the modern outgrowth of the publishing and media industry, though I find his focus and conclusions beyond that point lacking.<p>For starters, the Internet itself is far more about <i>distribution</i> than <i>duplicating</i>. It's <i>effect</i> is to achieve <i>distribution</i> of information. Computers alone are sufficient to achieve <i>copying</i>, but it was the <i>distribution</i> component which computers alone, or even low-speed or fixed-location Internet failed to achieve. And though the Internet's distribution involves copying of bits, that's not the critical function.<p>Information theory itself concerns the <i>origination</i>, <i>distribution</i>, <i>receipt</i>, <i>comprehension</i> (or decoding), and possibly <i>redistribution</i> of messages. Those are instantiated through such systems as webservers (or application engines), the Internet, protocols, and ultimately some human or nonhuman interpretation engine.<p>Kelly is <i>also</i> correct in that the Internet removes costs and frictions. Or alternatively, it <i>shifts</i> the cost points of media, and generally <i>reduces</i> the costs of originating and duplicating information, without increasing individual capacity to comprehend or filter it. This gets to the core of what's been called the attention economy, forseen in the 1970s and earlier by Herbert Simon and Alvin Toffler.<p>I've been looking at quantifications of information absorption capacity of people and suspect that that's generally quite <i>low</i> -- a few tens, possibly hundreds, of messages per day, but not much above that. The consequence of filter overload is a real problem. And the <i>attention</i> we can give any one message is directly <i>inversely</i> proportional to the number of messages received, divided by our time to dedicate to such messages.<p>Reducing costs gives rise to the Jevons paradox: It's not that the high-barrier-to-entry messages of the high-cost era suddenly become more available, but that a new flood of low-barrier, low-value (or negative-value) messages appear. Spam. Viruses. Junk mail. Phone solicitations. Web advertising. Clickbait.<p>There's also Woozle's Law of Epistemic Systems: as the audience using any given communications medium increases in number (and especially: in significance), the value of <i>manipulating</i> that audience will increase.<p>In market economics, low marginal costs become a problem as <i>marginal cost is the mechanism by which markets set prices</i>. For goods with high <i>fixed</i> and low <i>marginal</i> costs, market prices are simply insufficient to reward producers (or creators). This is the subject of recent books by Paul Mason (<i>Postcapitalism</i>) and Jeremy Rifkin (<i>The Zero Marginal Cost Society</i>).<p>Gresham's Law is another challenge, multiple ways. The popular form is "Bad X drives out good", though the general mechanism is more nuanced: Greshams mechanisms are complexity constraints applied where a "better" (more complex) good is valued identically to a "worse" one, for whatever reasons, within a given domain. The general result is that the worse goods predominate. The capacity to distinguish better from worse itself affects this, which is why a larger market tends to incentivise worse goods. And, if there is an alternative market in which the better goods <i>are</i> more highly rewarded, you'll see a flight of those goods to those markets. This is often exhibited as brain-drain: flight from low-paid professions such as teaching, government work, politics, or academia, to business, trades, or finance, for example.<p>It can also be exhibited as international flows of capital, coin, or talent.<p>And this plays out in information markets, where the lack of sufficient reward will see authors and creators either depart for realms in which they are appropriately rewarded, or to other fields. Political and ethnic persecution can have similar effects -- the flight of European Jews to the US in the 1930s and 1940s, or the flight of American blacks to Europe from the 1930s through the 1970s and even beyond, as examples.<p>Sorting out the implications of <i>changing relative costs and rewards</i> is key, and that's what Kelly, and a great many others, fail to do.