Open Source Uber

uber

Parts of the Uber software can be replicated and provided as a open-source product. Uber consists of

  • Location mapping
  • Payment processing
  • Pricing based on complex demand algorithms

What if an app had only the first two components? And the third was replaced with manual bidding by drivers?

Here’s how it would work:

  1. User tells the app they want a ride, just like in Uber
  2. The 15 closest available driversĀ are alerted and respond with a fare they will work for
  3. The quickest 5 offers appear in the app, and the user gets to see where they are, estimated times, and the rating of each driver
  4. The rest works just like Uber

Such an app would give all power to the drivers, who would operate the app as a collective. The overheads would be substantially less than the commission charged by Uber, meaning cheaper fares for riders and higher income for drivers.

Because Uber drivers famously aren’t employees, they would be able to use the co-op app at the same time as Uber, meaning it is low risk to try.

Note Libre Taxi seem to be partway there…

Natural v Sheeple

sheeple

The whole automated home thing bugs me. It seems to be the dividing line, it separates the sheeple from those who can think for themselves, and feel for themselves.

No free-thinking, spiritual/artistic being wants a fridge that orders the thing you just ran out of.

No soulful individual wants a world that is 19.0 degrees Celsius permanently.

Nobody with a heart desires an algorithm that spews out music that will always please you pleasantly.

And I’ve yet to see a dating system that comes vaguely close to matching me with someone suitable.

Shun easy. Revert to old ways. Fight the demise of the beating pulse of humanity.

Are you a natural? Do you dare walk into a warm room, turn on the aircon, and then after 10 seconds of extremely mild discomfort feel the change of temperature and smile?

 

Mystery Bot

Dear Elon Musk,

I have 50 ideas each day, just like you. Mostly mine never get done. So I’m gifting this one to you.

A little robot, like a robotic vacuum cleaner, that wanders around your house. And that’s it. People who buy it don’t know its purpose. That purpose will be revealed one day. Until then, it is a mystery.

People will go crazy not knowing why it exists. People will refuse to own one because of what it might see or hear. Which means everyone else will own one because they have nothing to hide.

It will be the most viral, perverse thing ever.

Oh, and to activate it, you need to give it access to your Facebook šŸ˜‰

 

Robot Theft Will Be a Big Problem

Robots and autonomous vehicles won’t be allowed to arm themselves. Which makes them easy pickings for thieves and and anyone who just hates robots (perhaps they lost their job to one).

An autonomous car is easy to stop – you just step out in front of it. What about a truck? Lots of valuable cargo might be in it.

To thwart the bad guys we will need:

  • Vehicles and robots that are very difficult to break into
  • Cameras that are monitored in real time
  • More surveillance cameras in general
  • Networking of robots and vehicles

There’s no point havingĀ video of someone stealing a robot, if they are wearing a mask. You need surveillance cameras that pick up where they arrived from and when they put the mask on. And if all the cameras are networked, AI can spot bad guys before they actually commit the crime.

And so the debate will begin – is increased surveillance a fair price to pay to protect out robots?

How Banks Could Distribute Cryptocurrencies

For any digital currency to take off it needs to come from a business people trust, and it needs to be widely distributed, and perhaps it needs to be forced onto people.

While governments will ultimately be the only winners in this space – because they can legislate against anyone else and they can’t lose control of currencies, banks could make a play. Banks obviously have the technology and infrastructure, and people already use their products to make payments.

Distribution can be easy and fair – by attaching it to either interest paid or received by customers. One BankCoin for every $100 in interest.

All the bank needs to do is make the BankCoin redeemable in enough places for people to benefit from it. Perhaps public transport, convenience stores and major online shopping destinations.

It could be seen as something similar to a loyalty card, that in effect gives a 1% discount – for example FlyBuys in Australia. The average mortgage repayment in Australia is around $2000, so that would be $20 per month in BankCoin. You’d want it to be around $100 per month to be widely adopted, so that is a 5% discount. Perhaps too much?

But what benefit do the banks receive for giving away 5% of their profits?

Micropayments. They are a product that will definitely happen one day, will work best if provided by a bank, and are only really achievable with a digital currency.

Once there is enough BankCoin out there, banks can stop issuing it, and just reap the benefits.

Oh, and all the banks in one country could work together on this…

Security Watch

 

We can do a lot on our phones these days, that saves us carrying other things around (like an iPod or a camera). The latest is the ability to pay with your phone, no more carrying a wallet…

And the Apple watch is now shifting those functions to your wrist.

So, in the near future, when I leave my apartment, go for a run, get a drink, pop into the gym… I still need to carry things.

  • Loyalty cards
  • Keys for my apartment
  • Key fob to get into the apartment building
  • Key fob to get into the gym

I rent, so changing the lock of my apartment isn’t easily achieved, and I still need a fob to get into the building.

My car has keys, but if it had a fob, well that’s another fob.

I think there is an opening for this last mile to be completed. Forget home-automation, and move into the territory of your phone/watch unlocking things. Actually, phones get lost/stolen.

Introducing the Security Watch

Either as a standalone or part of a smart watch. But when part of a smart watch, it has a separate battery for the security features, so a flat watch battery from too much chatting won’t keep you from getting inside your home.

Has to be standardised. Ideally a thrid party designs this, because Apple/Samsung/Amazon/Google will want to use their own proprietary systems.

Works universally. Allows access and entry to your car, home, workplace, gym. Used for loyalty cards. Used for VIP access to a club. Anywhere that one way ID is needed (so not purchases).

Kids can wear them.

Waterproof and shockproof, obviously.

As a standalone device it needs to last one month on a charge. As part of a smartphone, it recharges when you charge the phone.

Here’s a key concept. There is always a physical version of the key/card/fob you are replacing. And that key/card/fob is what you use to get permissions for your Security Watch.

1. Push the connect buttons on your Security Watch and key/card/fob at the same time to connect them
2. The key/card/fob sends a message to the manufacturer to check if more users are available (so a house might get a 10 user limit, a car 5, a gym pass one), and registers them.
3. The user verifies their ID on the Security Watch by PIN and biometrics.
4. If 2 & 3 are successful, all is set up.

So your entry card for work gets activated on your Security Watch, but you keep that card somewhere safe at home, in case you ever need it.

(And keep, a spare house key with a friend…)

Reporta – Free App Idea – Could Be Huge!

obikes

A very easy to build app could be massive if you get the right stakeholders on board:

  • Police
  • Council
  • TV/Radio/Newspapers
  • VicRoads (local road authority)

The concept is basic – take a photo (or video if need be), the GPS location is tagged, you type short message (or nothing if the photo tells the story), and you choose which authority to send it to. Not anonymous, your name and phone number will be included.

See any of these and report them, quickly:

  • Crime
  • oBike in a stupid place (or damaged)
  • Graffiti
  • Potholes
  • Damaged public property, like road signs
  • Storm damage

For something like a pothole, if enough people report it, the council will be prompted to make fixing it a priority.

There is no revenue model, except maybe if everyone ends up installing it, media might pay for photos or videos

AI Squared: AugMe

Scenario: it is 2am, you are drunk, you are asking Alexa to order the latest microscopic drone.

Without Assistance:

You: Alexa, order me that XYZ Zippy Drone please
Alexa: Sure, no problem. Do you agree to pay $800 plus a $65 expedited delivery fee?
You: Yep

With your own AI Assistance Bot:

You: Alexa, order me that XYZ Zippy Drone please
Alexa: Sure, no problem. Do you agree to pay $800 plus a $65 expedited delivery fee?
Assistance Bot: Rob, you only have $200 in savings, do you really want to go into debt for this? It only has a 3-star rating and has negligible utility value
Rob: I want it, whack it on the card
Assistance Bot: Delivery is free if it takes 3 days, would you like that instead?
Rob: Hell yeah!
Assistance Bot: Alexa, please proceed with the order but with the free delivery option please

This will be the next big thing, aroundĀ  5 years from now – Augmented Me.

Imagine being at the supermarket, and your AugMe negotiates with the automated checkout:

Assistance Bot: He doesn’t want to redeem points, he is paying with this card, here is his PIN, he doesn’t need a receipt.

 

 

Self-Driving Car as Video Studio

Self-driving cars are coming…
Some people already record themselves using dash cams while driving…

In the future, some people will have self-driving cars that are essentially a video studio where they can film themselves. That way the hour of commuting is use efficiently by those who want/need to video themselves everyday.

  • One bench seat at the rear of the car, or a pair of swivel seats
  • Green screen behind the seat
  • Microphone above the seat
  • Multiple cameras
  • Blacked out windows

You could even have a teleprompter…

AI Checking Startup

AI-Democracy-6

Artificial Intelligence will have a major influence on everyone’s lives (yes, even not digital folk, who apply for a mortgage) very soon. The problem with AI is that it makes decisions that can’t be deconstructed. We just need faith…

While the decisions can’t be deconstructed, they can be tested. I propose a service that tests and rates AI-based services.

The testing will be done by an automated system, guided by real people in real time.

For example, a restaurant booking system. A script will try booking restaurants at various times and locations, for different cuisines and numbers of guests. A human observes the process and rates the service on accuracy. The human also looks for bias and favouritism.

Or for a system like Amazon’s Echo, the test can be for finding obscure music using a strong accent.

Users can leave feedback that give the business new aspects to evaluate.

Monetise? Free for everyday folk, and for a fee businesses can have their products tested prior to launch