There is too many assumptions to make a case either way. There is zero info about actual unit economics available publicly. We don't know which cities are witho

There is too many assumptions to make a case either way. There is zero info about actual unit economics available publicly. We don't know which cities are without subsidies (on both driver and rider) and which ones are, only Uber has it available internally. Heck, due to the growth in a city, the economics themselves change as the company scales in a city. Only Uber can actually project out to a certain reasonability (which are almost always different from reality even after that). Projections are just broad strokes to see if something will definitely not work rather than a proof that it definitely will. All these arguments could have been made for Google, Amazon, Facebook etc at a certain point in the company's life, and then BAM - in a couple of years there was some "magic point" where the whole company became a cash minting machine. Also, Amazon is not a "pure software" company - there is a huge operations backend which was scaled along with the company. People tend to look at a given snapshot in time and confidently proclaim that it won't work - ever. But the world is a complex place - nobody has a fking idea what's gonna happen next. Instead of finding reasons to proclaim "This is not gonna work - ever.", it would be better if we can ask - "Okay, it looks like it doesn't work in the current phase - but what are these people doing which makes it possible to work." In short, its moot to discuss why Uber will fail - atleast right now - there isn't enough data that we can confidently project to make that conclusion. Its better if this sort of analysis is done in hindsight because right now there are too many variables which are changing too fast for anyone to make any prediction about anything.