Many have long believed that robots would reach a point of sophistication where they would be welcomed in homes around the world and make a huge difference as helpers and companions. Science fiction has instilled in public consciousness a vision of how robots should look and behave After a long wait, there some signs that such expectations could become a reality within the next few years.
Japan’s telecoms giant SoftBank has introduced what its hoping to be its robotic game changer, teaming up with Aldebaran Robotics (the team behind the robot NAO) to make sure it gets it right. This is Pepper. He’s adorable and hilarious.
Another interesting recent creation is Cynthia Breazeal’s little home robot Jibo that has just secured a massive $25M first funding round on the back of a hugely successful crowd funding campaign. [link] (this is almost double OculusVR’s first round of VC investment).
Do these signs really mean that it is finally time for the home robot revolution like it was for the home PC some decades ago?
Lets’ focus on Pepper for a moment. Pepper in particular looks like a very well designed product. If it is to be a household robot, it needs to be physically stable. If they gave it legs instead of wheels, it would eventually fall over and get damaged. Science fiction has created a culture of fear and mistrust of robots. With Pepper- it is impossible to be scared of this fellow. Also, they didn’t fall into the trap of trying to make something look "too human", because very human-like robots often fall into the "uncanny valley" problem, where they end up looking like an animated dead person, which can be very creepy.
The Importance of robot software evolution
For home robots to be successful, they will have to be constantly evolving products that do not get blandly predictable or boring. It is important to think of robots as a software ecosystem (with the enabling hardware), in a similar sense to smartphones. Constant updating of software will keep the human-robot interactions unpredictable and make the owner feel like they are interacting with the robot for the first time, over and over again.
If it can’t do this, it might become boring and "used up". Think of it this way: would you buy a $2000 computer (roughly Pepper’s price) that has only skeleton software and no programs to install or no means to access cloud services?
Star Trek (and to a large extent, much of SciFi) picture robots as technologically isolated, self contained machines. I remember when Data died in Star Trek: Nemesis – the crew was mourning at his funeral. Apparently they live in a universe where they can beam a steak sandwich out of thin air but can’t back up a computer. (That isn’t the only thing in Star Trek that doesn’t make much sense-I say this as a fan).
The actual robot revolution when it happens will probably look like this: It will mainly be a constantly evolving software cloud that comes with some cool additional hardware (the robot). The survival of the robot will not matter that much, you can get your bot a new and improved body when the next model comes out- the brain is essentially its online account in the cloud, which is constantly being upgraded through wireless connectivity.
To compare with the iPhone the iPad, it is important to note the applications as a critical feature in their success. When Apple first launched the iPhone it had no App Store. Steve Jobs wasn’t sure if he wanted to have one because of his desire to totally control the experience. Eventually he relented. But even if Jobs had been on board with an App Store from day one, there’s no way it could have been ready right away. It’s a major community undertaking.
Not all at once- perhaps light-entertainment first
Sooner or later, someone will give home robots a chatbot program for entertainment purposes. Chatbots are not great at convincing you they are human, but it can be really funny watching them try to process and react to what you said.
This could be a good focus for the robot’s initial selling point -its entertainment value – rather than trying to be able to do everything amazingly well straight away. It should talk to you, and tell you jokes or interesting facts. If you promise the product to be amazing at everything you run the risk of delivering on nothing.
My view would be to focus on making the product fun right from the start. Then let the developer community greatly extend the functionality in both predictable and unforeseeable ways.