We cannot predict right away if AI will turn all terminator-evil in the future or be the good guy like Jarvis but one thing is sure that visualization will play an important part in its effective utilization. Data Visualization plays a dual role when it comes to AI. One, it will visualize the AI by plotting the relevant outputs on a visual canvas. And two, it will paint the picture for the complex processes that take place behind that smooth functioning AI you see. A visual representation of the process will give better insights about its working than reading a manual or text files. AI may be as difficult as rocket science but its worth the time and money invested. At this very moment the major players of the tech industry are working on some sort of AI inspired element to go with their products. And if you happen to own a smartphone then chances are that your phone will be having an AI component. Google Assistant and Siri are spin offs on the very same AI concept we are talking about. So its not that uncommon and gradually its becoming a part of your life and helping you out in ways you may not be noticing.

But one core question that lies at the bottom of all this effort and research going into this field is “WHAT are we trying to achieve here?”. With the minimal understanding of AI that i have, i will say that in essence we are trying to replicate the human brain. The very prospect of giving the “ability to think” to machines is very fascinating. Till now man has created things that have their defined roles and do exactly what they are meant for and not a single thing more. However, if we succeed completely in achieving AI in its fullest form then we would have created something that will have the ability to change and adapt according to situations to obtain most optimum results.

This brings me to the next segment of the blog- “How do we achieve this?”. Since we are trying to mimic the human brain then its only right to take inspiration from the same. Our brain has neurons that are linked to pass information to all parts of the brain. Similarly, we build a network of digital neurons that communicate with each other to achieve an objective. They send each other messages and store and analyse the responses. The connections and interactions that lead towards the objectives are strengthened while the non-fruitful interactions are eliminated. These are stored in memory so that the next time a similar situation presents itself then the same path could be followed.  The process ensures that the system keeps learning form the data that is being fed to it. And hence rises the need for data. The more valid the data, the better are the results. You must seen a pop up in you iPhone that asks your permission for the diagnostic feedback. Such are the types of data that they can very well utilize in their AI research programs.

Daniel Smilkov’s visual interprets the beautiful relation that data visualization and AI share. Smilkov partnered with Shan Carter from Google to develop an interactive Data Visualization system that you can tweek as per your wish and play around a bit to have an idea about how AI works. Every change done will have consequences. Just like every interaction we have alters the outcome in minutest of the ways, similarly any change at the smallest of the level will have changes in the output. Have a look at it and see for your self.

You may be talking soon to an android in the future. Its not a fantasy anymore. Fiction is about to enter reality.