Update: I had to block an angry libertarian on Twitter who declared me an "Uber hater" and was launching all kinds of contempt and overgeneralization my way. I also saw people retweeting this post and describing it as a general takedown of the sharing economy. So I just want to clarify: I think AirBnB is a terrific business, but Uber might be a terrible one. This is a post for rational adults who can handle nuance. If that's not you, there are plenty of other posts to read out there.
Sometimes an industry which is very tightly regulated, or which has a lot of middlemen, has these obstructions for a reason.
Working for any taxi service holds enormous opportunities for kidnappers or rapists, and makes you an easy target for armed robbery. In some cities, being a cab driver is very dangerous. In other cities, it's dangerous to take a cab anywhere - don't try it in Argentina - and in many cities it's dangerous for both the driver and the passenger. A taxi service with built-in surveillance could be even worse in the hands of a total maniac. And just to state the obvious, both Uber drivers and regular taxi drivers have logical incentives for driving unsafely.
In this context, I have no problem with governments wanting to regulate Uber literally to death, at least outside of California. California is a special case, in my opinion. Cabs are completely unacceptable garbage throughout California, and the startup scene definitely skews Californian, so the typical startup hacker probably thinks all cabs suck, but cabs are gold in London, Chicago, and New York, and probably quite a few other places too.
In Los Angeles, cabs are licensed by city. A cab driver can only operate in one "city," and can face disastrous legal consequences if they pick up a passenger outside of their "city." But the term's misleading, because the "city" of Los Angeles is really a massive archipelago of small, loosely affiliated towns. It's extremely common that a taxi cab will be unable to pick up any passengers once it drops somebody off. I lived in San Francisco before I lived in Los Angeles, and I had always assumed Bay Area taxis were the worst in the industrialized world, but Los Angeles cabs make San Francisco cabs look competent.
(In the same way that San Francisco's MUNI system makes LA's Metro look incredible, even though both suck balls compared to the systems in New York, London, and Chicago. I'm told the system in Hong Kong puts even London's to shame.)
By contrast, in London, you have to pass an incredibly demanding driving and navigation test before you can become a cab driver. Researchers have demonstrated that passing this test causes a substantial transformation in the size of the cab driver's brain, in the regions responsible for maps and navigation.
Disrupting something which works great (i.e., cabs in London) is not as impressive as disrupting something which sucks ass (i.e., cabs in California). I think the tech industry overrates Uber, because the industry's headquartered in a state which probably has the worst taxicabs of any first-world country.
Uber is obviously a disruptive startup, but disruption doesn't always fix things. The Internet crippled the music industry, but introduced stacks of new middlemen in the process. Ask Zoe Keating how that "democratization" turned out.
In music, the corporate middlemen are this pernicious infestation, which just reappears after you thought you'd wiped it out. Artists have to sacrifice a lot to develop their art to a serious level, and a lot of music performance takes place in situations where people are celebrating (i.e., drunk or high, late at night). So somebody has to provide security, as well as business sense. There's a lot of opportunities for middlemen to get their middle on, and disrupting a market, under those conditions, just means shuffling in a new deck of middlemen. It doesn't change the game.
In the same way that the music industry is a magnet for middlemen, the taxi industry is a magnet for crime. A lot of people champion Uber because it bypasses complex regulations, but my guess is that disrupting a market like that means you just reshuffle the deck of regulations. If Uber kills the entire taxi industry, then governments will have to scrap their taxi regulations and re-build a similarly gigantic regulatory infrastructure around Uber instead. You still have to guard against kidnapping, rape, armed robbery, unfair labor practices, and unsafe driving. None of those risks have actually changed at all. Plus, this new regulatory infrastructure will have to deal with the new surveillance risks that come along with Uber's databases, mobile apps, and geotracking.
Uber's best-case scenario is that the society at large will pay for the consequences of their meager "innovation." But if you look at that cost realistically, Uber is not introducing a tremendous amount of new efficiency to the taxi market at all. They're just restructuring an industry so that it offloads, to taxpayers, the costs and complexities of a massive, industry-wide technology update.
Uber got rid of inefficient dispatch mechanisms, and routed around licensing laws, but odds are pretty good (in my opinion) that those licensing laws will re-assert themselves. Those laws don't exist to preserve an inefficient dispatch mechanism, which is the only real technological change here. They exist mostly to guard against risks of criminal behavior. Those risks are still present. Like it or not, licensing laws are a very common response to risks of that nature in societies like ours.
It's possible that those licensing laws might be gone forever, but I wouldn't bet on it. I think it's a lot more likely that Uber has a huge IPO and its stock price withers to nothing a few years later, once the slow-moving systems which generate regulation in democratic societies have had time to react. If all you're doing is building to flip, sure, you can take the money and run, but if you want an actual business that lasts, Uber might not deliver.
I could totally be wrong. Only time will tell for sure. But I think Uber is a really good lesson for entrepreneurs in how a market can look awesome but actually suck. From a long-term perspective, I don't see how they can hope for any better end-game than bribing corrupt politicians. While that is certainly a time-honored means of getting ahead, and it very frequently succeeds, it's not exactly a technological innovation.