The Lily Drone made the headlines in 2015 promising to change the way people would take photos. It was portable, easily fitting in a backpack, it could be launched from the palm of your hand and would take care in complete autonomy to photo your walk, canoeing, snowboarding with no need for guidance. It would follow you using a beacon strapped to your hand and you can push a button at take off to tell it how you wanted to be filmed, from the back, from the side, from the front.
The winning proposition was its being autonomous. And that could be achieved through sophisticated AI.
In 2016 it was awarded the “product of the year”, in July 2017 the company went bankrupt before having been able to sell a single unit, in spite of the thousands of pre-orders that brought 30+ million $ in the pockets of Lily Drone. You can read the full story on Wired.
There is no doubt that this was a failure and it also show how quick the press may be to give hype to something that is not a product by naming it “product of the year”.
Reason I am posting this news is that in spite of this failure I do believe that it is just going to be a matter of time. By early the next decade I am betting we will have these types of autonomous drones, may be coming from an established company like DJI, and more than that, we will consider them as perfectly natural. Why would you need to “pilot” a drone? You just tell it what you want and it will take of all technicalities.
Think about your point and shoot camera. You tell it you are taking a portrait photo and it will take care of all the technicalities, like focus, exposure, white balance, background bokek … and much more. These technicalities are so obvious that you don’t actually know them, even less how to accomplish them!
Autonomous vehicle are already here, just not in the price point that can make them affordable in many areas. But that will (actually is) change.
We often think about autonomous systems as “complex systems”. In reality they are winning the market because in an ever more complex world they are the easy way to go. And once you have reached real autonomy you have also reached the possibility to increase system complexity since all the various parts (systems) will take care of managing the complexity by themselves making life easy for us.
There is, will be, of course a big issue. As our world will become at the same time way more complex and way more easy, what could happen if something breaks down and no longer takes care of managing the complexity on your behalf? What if once we will be used to self driving car the car (self driving capability) breaks down and you will have to take the wheel? Will we be ready for that if we have established a symbiotic relation with the car?