Matt makes a point that more people in the free software space should learn to appreciate:
Another issue raised in comments is the idea that a “fair use” by definition can’t be commercial. I was glad to see someone raise this point if only because I do wish we could re-inject more life into the commerce/non-commercial distinction for broad copyright purposes. But my goal would be to use the distinction to raise the scope of tolerated non-commercial copying, not to narrow the scope of allowable commerce. Commerce is a legitimate and important human undertaking, and the goal of copyright law should be to facilitate useful commerce. That includes preventing large-scale commercialized digital copying, but I think also means allowing commercialized sampling, quoting, and repurposing of existing material.
[...] I think it’s flatly wrong to consider private actors’ interest in public data to be uniformly problematic. We should be clear: we won’t tolerate those interests’ occasional attempts to lock public data into exclusive monopolies. I think our community has done a pretty good job lately of identifying such situations and stopping them, and of course people like Carl Malamud have been doing important work on this question since well before most of us ever heard of “open data.” But if commercial activity is enabled by data, that’s all to the good—the great thing about digital information is that scarcity doesn’t have to be a concern. Google Maps’ uses of Census TIGER data, for instance, is proprietary, motivated by profit, and unquestionably a huge boon to human welfare. And the source data remains free for anyone else to use! Cutting off those kinds of uses with noncommercial licensing would be nothing more than a destructive act of pique.
This really came into focus for me when I was in Berlin for a conference run by the good folks at the Open Knowledge Foundation. I admit that before I stopped to think about it, I never found noncommercial licenses that problematic, and would casually throw them on material I produced on the web. That way our vaguely-defined communal web society (so pure and untainted by the profit motive!) could use it, but “they” wouldn’t benefit from my hard(?) work tagging photos on Flickr. Honestly, this was dumb. I was never going to put in the time to try to make money off of those photos. If someone else could do the work to make them useful to others, why begrudge them that opportunity?*
I wouldn’t go as far as Matt about the goal of copyright law being to facilitate commerce (this formulation seems to ignore the kind of deadweight losses that Matt’s writing about IP is usually about). But he’s right about commerce being a “legitimate and important human undertaking.” Certainly the private sector is capable of excesses, but it’s also an incredible tool for identifying and satisfying human needs. We shouldn’t resent it out of some sort of ideological tribalism — particularly when we’re discussing digital goods, where things are rarely zero-sum and where (with apologies to Julian and Kash) the negative externalities (Mark Zuckerberg can infer your sexual preferences) are less severe than those found in the physical world (the chemical plant next door means your baby was born with fins).
* I should acknowledge that this is just a for-instance. As there’s very occasionally a market demand for unflattering photos of some of my friends by ideological press outlets, I’ve elected to keep somewhat restrictive licensing on my meager photographic output. But in less problematic cases — the code I put on GitHub, for instance — I’ve moved to open, nonviral licensing.