It used to be that you'd have to google the number of the local taxi companies (if you're lucky enough that they have a web presence (pre internet you'd ask the hotel concierge, or check the yellow pages, and many of them never updated past that model)) then call the dispatcher, who would grunt something at you and hang up, and if you were super perceptive you might be able to tell whether they said a cab is coming or not. If you were leaving the airport or a really busy hotel then you'd just walk into the next cab to pull up with no dispatch, though.

There were pre-Uber attempts at app dispatches for cab companies, but different cab companies would use different ones, they were all shit, and most of them didn't do it.

So the cab companies did fuck this up and Uber did a better job, and Uber replaced them, and everything "improved" exactly how it's supposed to under capitalism. But cab companies and Uber aren't people; the actual people (drivers) all got fucked over.

Oh yeah apps make things more convenient for sure. But functionally the app is just pre-populating ride request information in the system for dispatch instead of taking it down over the phone like in ye olden days. Nothing of the economics of fleet management changed, in fact if anything Uber’s decentralized system is less efficient, the costs from the inefficiency just get pushed to the gig worker drivers.

A lot of taxis now use similar apps for requests and dispatches; NYC has been using Arro for years for yellow cabs, and you can request accessible taxis to boot. There wasn’t anything particularly proprietary in the app, which meant the established industry would pick it up sooner or later.

I think this is why Uber has been pushing Uber Eats as of late, as there they’re just competing with Seamless/Grubhub, who also employ gig worker drivers for restaurants that don’t have delivery drivers.

Others have gotten the main points, but I'll add one more.

Before Uber, there was no guarantee that your taxi would ever come.

You had to tell the dispatcher both your location and the destination, it would be relayed, and then the drivers would decide if it was worth the effort.

Often short suburban trips outside weekends would not be as profitable to drivers and so they wouldn't pick you up and there was no way to tell except repeatedly calling up.

So, imagine 11 year old me, standing outside alone in 40 degree summer heat for 90 min waiting for a driver that never came. This happened several times.

Uber really did make things better for awhile.

There wasn't much in the way of innovation in Uber/Lyft. Private Taxi companies and the regulated cab commissions were able to get competing apps going pretty quick. They just couldn't compete because they had to charge some kind of fee to the Taxi operators, while Uber was giving users a huge discount.

The venture capital firms knew from the beginning that Uber wasn't about technology or innovation, it was about being the next tech monopolist.

Get in between businesses and customers in an industry that's going through a technological change. Subsidize to the benefit of the businesses and the customers to prevent competition. Grow until you have monopoly power. Increase the costs to the users, then increase the costs to the businesses. Next step is usually offering to sell the customers to the businesses(usually by selling ads), but I could see it being some kind of Uber driver gold subscription or something that gives you priority in the app.