I was reading few links shared by my colleague on the upcoming book “Human Face of Big Data”. While reading these articles I started realizing that where this is heading. Today there are systems which are getting build to analyze human behavior of all sorts. The primary driving factor today is social networking and leveraging this for advertisements and generating funds. However systems are coming in healthcare that analyze basic human behavior to understand warnings of illness, health risks like heart attack or its recurrence. As the technologies are getting built what the computers are learning is how to process unstructured data.
Today a human brain is much better and faster in many ways is because it processes unstructured data of all kinds. Even when you are on a foreign land, you start collecting information and process it. It takes some time to settle down however, the basic programming of the brain is not required to be changed. As we are now developing techniques to get meaningful information out of Big Data, the next obvious step would be to write a generic logic that can process any sort of data and furthermore, get audio, visual, textual, geo-spatial data and process all together by linking all of them in real time.
When the computers were first designed, they were huge and did very little compared to the portable devices we have today. Considering the same path of progress, the amount of data storage space for big data will eventually go down so will the speed to process the information.
I was introduced to Edward de Bono and his books few years back and what has stuck to my mind is that human brain processes information in patterns. His books on how humans think and about creativity, lateral thinking were ground breaking. His research brought in lot of formal techniques to improve human brain’s creativity. This same research going forward will be useful for training computers to process big data and provide real time knowledge to take actions. Even with so much computing power in the brain, human beings go wrong. So there will be similar prioritization by computers in deciding which data to consider and which data to ignore. The current big data analysis will provide required patterns and will help faster decision and action. Somewhere the programmers will take a call and do approximation in favor of a quick action. All this now seems doable and within our reach.
Robotics on a parallel track is making good progress and by integrating cheaper technologies like Kinect, Siri it will not be too long that we will have human like robots. Feed them will huge data processing capabilities and you have nearly human robots. Here is my thought on how these robots will be built. In our ancient scripts it has been written about few human attributes. How do current technologies score on these? Let’s have a look. There are 5 senses that are needed for human beings. “Shabda, Sparsha, Rupa, Rasa, Gandha” Meaning Sound, Touch, Form, Taste, Smell. Today’s computers do fairly well on Sound. Meaning they can speak and can listen to a good extent. The screens are touch sensitive so it will not be too long to build other elements that will get this capability. Form, the Kinect cameras are now able to identify objects. Taste and Smell are finer attributes and in the initial stages, Robots may not need them. So on three attributes, there is good progress. Then come 4 primal attributes of life. “Aahar, Bhay, Nidra, Maithun” meaning “Food, Fear, Sleep, Reproduction”. Now there is not much built on this yet. But here is how it will go. If a robot with ability to identify objects and analyze speech is able to find food for itself to run, it is almost independent. So how do robots become independent? My theory goes like this.
What could be food for the robots? If you have seen Wall-E, these robots will initially start generating power on waste material there by cleaning the earth. Now when the power is going down, what happens with human beings? We feel hungry. So in today’s world we have a food security for many of us. Which means we know where the food is and we have enough money to buy it. In robotics world, there will be a network of robots, which will keep track of all the food required by them and any robot can ask the central system where the nearest food available is. It will conserve enough energy to reach that food in time.
The next one is Fear. How, what and why should a robot fear? A robot should know what can potentially destroy it. Such as falling off from heights, getting run down by vehicles and it should be able to protect itself. With the 3D cameras, it will not be long to program robots to understand the risks and protect themselves. We have even seen movies where the robots know combat techniques. The two elements of Rasa & Gandha (Taste and Smell) help human beings a lot in understanding the chance of a risk and by creating fear, they protect us. Eventually there will be more research on this when the robotics moves from Mechacnical and Electical/Electronics towards biology and chemistry side.
The next one is Sleep. Now you would ask why a robot needs to sleep? And even if it needs how much will it need? The first need for sleep is to get repairs from the wear and tear. This will not be much frequent and the time spent also will not be much. The second need is to process the data that it has captured during the day and process it. Today most of the human beings do that in their sleep. Connect it with the Big Data discussion we had above. In this sleep, the robots will form their own patterns, they may connect with the Force for the information data that they need. And lastly Reproduction. I think this is a taboo topic in case of human beings and in any case does not require much explanation.
The most key element in this today is the missing piece of technology for the “Sleep” factor. So in short it is my belief that with the Big Data processing speeding up, the day we will see the self-sufficient robots is not far off. Time to experience the real Star Wars or Terminator or Matrix.
While you are imagining how fast this will happen and will it happen in our life time, here is an interesting thing connecting to our real life. Look at the two attributes of Fear and Sleep. For both these things we need huge data analysis. In case of fear, we connect with God. We ask the super power to protect us. We ask the super power to give us strength. What we are really asking is the knowledge, data which will help us survive. But in case of Sleep, we are not conscious. Still we are building that knowledge. In many religions there is a ritual to pray to God before going to sleep and after waking up. We need to connect it to the requirement that we need patterns to process information faster, patterns that will take us on better life track. When we get up we thank the super power for giving us those patterns and helping us to go on right track. To me this Super Power is our super computer that is analyzing the Big Data by keeping connections with us. Today we are losing fear and survival of earth and eventually ours is at stake. Are we missing that connectivity with the super power for the right patterns to analyze the Big Data?