Hacker News new | past | comments | ask | show | jobs | submit login

What exactly is precluding someone from building this today, regardless of the tumor?



It exists. There have been a number of decentralized cab hailing apps. I was never able to get a single cab on any of them.

The problem is decentralized apps have shit advertising budgets.


Which is because there's no money in it.

If you build a centralized platform, you get to act as a middle man and skim off profit off every transaction, which you then reinvest in advertising and feature development.

On the other hand, if you build a decentralized platform, you've essentially commoditized yourself: you can't skim off profits, because if you do, a cheaper node will just pop up, leading to a race to the bottom.


Ironically, the actual system that exists and existed pre-Uber in NYC (TLC) is about as close to a real decentralized system as you're going to get. Like most good decentralized systems, it relies on a small centralized core (the government) to enforce a few basic invariants (taxi drivers must be trained, licensed, pass background checks, vehicles require insurance and must pick up passengers in certain zones and not in others, etc) and offer a few basic primitive operations (get driver license, get FHV car license, get base license, etc) to get involved with the market.

Beyond that it's all decentralized -- anyone can, after jumping through the right hoops, buy a taxicab or medalliion, affiliate with a base, become a driver, etc. A passenger can easily find a car by walking about half a block to the nearest avenue, putting their arm up in the air (in much of Manhattan) or by using the Curb app (in less busy areas).


The total number of cabs and the prices they charge being set by a central authority is decentralized?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: