My fascination with Algorithms started when I was quite young. In my teens perhaps – when algorithms and the emerging world of computers seemed to be enticing and promising in equal measure. My brush with them began in a high school computer class, when we were introduced to these archaic boxes of off-white bulky computers with a grey or black hard keyboards.
It was the early days, computers were a relatively new invention and being able to see one live in front of us was quite exciting. The first language we learnt was BASIC, and then graduated to more cognitively expensive ones like C and C++. You could make the computer do enchanting things, by giving it the most complex earth shattering instructions and then watch with pleasure as it bends over backwards to do your bidding. Indeed, how dramatic!
And to add to that, these computers were primitive and heated up rather quickly so they needed enclosed and air-tight rooms with air conditioners in what was known as the “Computer department”. If you grew up in a small town with the harsh unforgiving Indian summer – spending time in there was quite a treat. Computer classes were the favorite even among students who didn’t fancy programming.
Until, one day, almost ten years later – I discovered the magical world of Machine Learning.
Machine learning, as a concept was a new paradigm where computers do not need to be programmed with explicit instructions about what needs to be done, but can be taught to learn purely by observation. And at the core of it is the concept of Learning. So first you train your algorithms to learn from the past, and with this knowledge of the past learnt primarily by observation, your algorithms can predict the future and take actions. All this may sound very mysterious but there is plain logic and and a lot of math behind all this.
My discovery of Machine Learning was not accidental. I started with reading books and spinning experiments of my own. And slowly, applying these experiments in many work projects exposed me to the inner workings of these digital beasts.
And the more I knew – the more it astounded me. Imagine a machine that can observe how a system has performed in the past and develop a complete knowledge of the system from Day One. The power of this capability is mind boggling , and frightening in equal measure.
Today, this is the core of what I do. And yet, the more I discover it, the more enamored I am by how much of this world fits into certain patterns, and how much of it can be discovered through pure math. It is also surprising to me how these algorithms reveal our hidden beliefs and desires, some of them which we might not be aware of ourselves.
There are very many fears on what this means for our future, and where this technology will take us. And it also raises provocative questions.
- What can we learn from these super powerful algorithms?
- What are the benefits and limitations of using these technologies?
- How do we leverage these technologies without succumbing to the inherent biases they come with ?
- What are the key challenges we would face — as strategists, programmers, individuals, society and humankind in general?
There are many versions of answers to these. I am hoping to discover my own answers through these pages.