It's a good argument and all, whether it's made by Matt or the economist. But, why does everyone focus on money? Is wealth relevant or not? Surely giving away free stuff creates more wealth, if it is sustainable. The real answer to this dilemma is to somehow redirect the outflow of wealth partially back to the producer without the use of money.
For example, say you created free, online computer games where people spent all day reconfiguring proteins. The game is completely free, thus providing wealth to people, and people pay the service back by using it.
The only reason money is ever mentioned is because it is the simplest, catch all solution. But, like the no free lunch theorem for search, money is your Monte Carlo baseline. For a limited problem domain there is quite likely a better technique than random sampling, and therefore a better wealth redirection technique than money. People just need to be more creative. (And, perhaps the comp sci analogy is more than just an analogy)
I've just finished reading Accelerando by Charles Stross, and one of the characters there is what he calls an "agalmic entrepreneur". He doesn't actually start businesses himself, just has half a dozen brilliant ideas before breakfast, finds the right person to implement those ideas and gives that person the idea (and the crucially missing connections to make it happen, etc). In that novel, he has no money - lives off of favours that he can call in at any time from the numerous people he's made rich.
Now, that's obviously not very realistic, but perhaps a mix between those two ideas might bear fruits.
There was a good article around called something like "the well that Google built"... (Ironically I can't find it on Google).
But the basic idea is like building a village well. The rich people of the village may band together and build a well - naturally, they don't make money off it, as the general populace would be too poor to pay.
What happens is that the prosperity of the whole village increases. Disease goes down, productivity goes up - so the demand for what the richer inhabitants have to offer (e.g. goods & services) goes up.
The Internet is all about this - Google isn't the only example, but is a prime one. Whilst many of their services do not directly make any money, they make the Internet a more useful, better place... And the more that people turn to the Internet, the better off Google is.
Edit: pg touches on this too, with his comment about a startup tackling an issue like malaria: http://www.paulgraham.com/good.html
Curing malaria isn't going to immediately/directly make anyone any money - but it's possible the downstream effects could generate a lot of wealth.
Hosting companies, grocery stores, and landlords only accept money as payment. That is the main reason.
Google already does what you are hinting at: they give away free services and then harvest the input from that usage to make their services better. However, at the end of the day, Google has significant expensive that need to be paid in "real" money.
For example, say you created free, online computer games where people spent all day reconfiguring proteins. The game is completely free, thus providing wealth to people, and people pay the service back by using it.
The only reason money is ever mentioned is because it is the simplest, catch all solution. But, like the no free lunch theorem for search, money is your Monte Carlo baseline. For a limited problem domain there is quite likely a better technique than random sampling, and therefore a better wealth redirection technique than money. People just need to be more creative. (And, perhaps the comp sci analogy is more than just an analogy)