We spent a weekend contemplating fun uses for the Uber API. Here are a couple ideas that we abandoned - feel free to use as you wish:

* "Bar roulette" where you would hit a button, an uber would show up, then take you to a random bar sourced from Yelp.

* Last-minute ticket sales to events (e.g. in the hour before they start) that include transportation

* Having bars or clubs offer to drive people there for free and bid based on demographics of the user like an ad - e.g. "We need more people in this age range, of this gender, etc." They would set a bid, and the app would ping people that matched the criteria and offer to drive them there for free if the estimated fare was less than the bid.

That third one reads like something out of a Gibson novel. Might work best with a pre-defined group of people (e.g., models in an agency) or something similar.

"like something out of a Gibson novel.", "models in an agency".

While not politically correct, it would be interesting to run an outcall escort service with this. Throw in an AirBNB API...

Disclaimer: The above is just a thought. I do not condone this use case.

This is what chatroulette should have turned into.

Or Tinder even.

An app for blind dates. Combines Tinder, Yelp, Uber, and Airbnb, with a pretty workflow over the top. Call it 'Dial-a-date'.

Airbnb might be a bit presumptuous.

You're giving Uber some ideas.

There could be all kinds of uses. Imagine a product launch event - "We need to pack a room so the press is impressed. Want anybody and everybody - tell them we have free food."

That'll get them to the location, but how long will they stick around.

As long as the food is free and they are not yet satisfied. Drip feed them.

Offer return rides at the appointed time :)

I worry that they're just going to follow Twitter's lead: build an API, get developers to work on interesting new ideas, pull in the good ones and the slowly squeeze out devs by locking up the API.

This is completely different. Usage of the API will translate directly to more rides on the platform and more $$ for Uber. The company will have every incentive to keep developers happy if apps are increasing ridership.

It's crowd sourced innovation.

The Uber API is pretty nice to work with. I built a mobile-web surge avoider: http://nomoresurge.divshot.io/client/

If you load that up on your mobile phone you can search for the location nearest to you with low or no surge, and then deep link into the Uber app to actually call the car.

Now I can implement requests directly, which is very cool.

I'd definitely use this. I've noticed that I can often change location by a couple of blocks to avoid surge charging

Could a "traffic jam on demand" be built using this? A la Chris Christie?

Sounds like something that would be against the acceptable use policy... and wouldn't be testable without quickly getting you banned.

Interesting... DDoS for hire comes to real life.

Once the Uber Affiliate Program pays out as a % of revenue (or even some sort of flat-fee CPA), then I'll be very very interested

Good news!

https://developer.uber.com/earn/

This only pays for signing up new users

What's stopping someone building an interface that points to signup, validates that sign-up and then hooks into the endpoint for [insert creative use here]?

I admit this was my one complaint about their API when it was released, since it meant I couldn't do thinks like write Uber for Glass. Their branding document is very detailed and handy to make their buttons look good on mobiles too, so I liked the quality of their work otherwise. Since they fixed it I'll have to take a look at doing the hackathon, I guess. Although maybe I'll have to put it on a watch instead of a Glass since Glass is dead.

Would it be possible to create something that effectively books an Uber for a later time/date?

An example would be I know I'll need an Uber tomorrow morning at 7AM so rather than get out my phone at 6:50AM I do it the night before and it just delays the request until the morning? Perhaps even add in some push notifications or something to check that you still want to book it and all you have to do is hit yes/no.

This would be great! This is where people still hire cabs in my city--If you have an early flight, do you trust that there will be an Uber driver available? Or do you pay a little extra and call a cab service with a scheduled pickup?

The difficulty is that there's still no guarantee that an Uber would be available. It'd just take care of the request part. Would need to be some thinking around what if there's none available in the area etc.

You mean a cron/at job with curl and json?

(cron repeats, at is for one time jobs)

Effectively, yeah. Thinking from the RoR world something like sidekiq or delayed_jobs could do the same thing.

Should definitely work. I use a cron job in the winter to kick off the remote start on my wife's car through Viper/Directed's JSON API.

SOME UBER APP IDEAS:

* An app that allows people to search up the products of small local businesses near by, pay for them via mobile, and then have it delivered to them via a trusted Uber driver. Good for small businesses since they wouldn't be able to afford a fleet of cars. Uber delivery will increase business since they will no longer be limited by their location.

* Tinder integration would be perfect. Users can agree upon a location via a google map in the app, and then schedule Uber drivers to pick them both up (option to share ride too). No need for either person to reveal their address to the other. Will make participants feel more safe.

* Vacation package integration. Not sure if this is already being done, but vacation packages can now include a dedicated Uber driver, along with the standard hotel + plane package. Will make vacations even smoother.

* Office lunch app uber integration. An app that lets co-workers pool money together to order lunch from local restaurants, and then have it delivered via Uber. Business for restaurant is increased since they are no longer limited by seating capacity (esp during high volume times - business plaza at 12pm for example).

* This idea is a combination of some of the others. Pizza Hut, Domionos, Chinese restaurants all do delivery... why not all of the other restaurants? Imagine an app that aggregates a bunch of restaurants and allows users to place orders. As the food is being prepared the app calculates all the nearby Uber drivers, the delivery locations, etc. It then finds the most efficient 'load' for a single Uber driver to carry. For example, pickup food from restaurants A B R since the drop off locations are all near to each other. This allows an Uber driver to take food from more than 1 restaurant to more than 1 customer, maximizing the amount of $$ they can make and decreasing cost for the end customer.

Nice, I'll have a look at integrating it into my Android Widget (https://play.google.com/store/apps/details?id=com.factorialp...) so you can order an Uber with one hit from your home screen!

Pity the affiliate program only gives you Uber credits, and not actual cash.

What I thought Uber Request Endpoint was before reading the article:

A box containing a captive iPhone and a credit card reader. When you swipe your card, they call an Uber to that location. So you can use Uber without a smartphone or an acocunt.

I still think that would be neat.

Not sure if Uber is already working on this, but I would love an Uber app for Android Wear/Apple watch. Would be pretty convenient to just tap twice and get a ride.

The Uber Apple Watch app was demoed at the Apple event. Exactly as you say, you can tap to get a ride at your current location.

Uber Proxy: use uber in cities where it's banned but pay the driver by drawing an equi-distance route in a city that it's not banned in.

What about a babysitting app of 27k users? Parents are willing to pay a uber to bring back babysitters.

I'm sure this will cost me some of that sweet HN karma, but:

Sweet; Now my automated dating life will really take off!

http://robrhinehart.com/?p=1005

https://news.ycombinator.com/item?id=6773841