Machine learning is a type of artificial intelligence that provides computers with the ability to learn without being explicitly “programmed”. It focuses on the development of computer programs that can teach themselves to grow and change when exposed to new data. They can also look at events that happened in the past and take a new decisions. This for me is a true “intelligent” system.

Working on one of my Machine Learning projects, I was slowed down by the time taken by my laptop to train a machine. It used to take me some 45 minutes. We decided to assemble a new PC with a Graphics Card (GPU) and train machine on that. The system had 16GB memory, 3.2 GHZ QuadCore processor with 12GB Graphics Card RAM and massive number, 3,000 GPU Cores. Using this hardware, the same machine got trained in less than 1 minute which is approximately 50 times faster. The $1,000 spent for this super fast learning machine, was money well spent!

Pondering on the days of past, recalled the era when I used to play Prince of Persia 2D on my HP Pavilion. Exchanging music files in 1.2 MB 8" Floppy Drive. Waiting for dial-up to connect and open YahooMail. Life was so different in those days or should I call it slow! This was an era where most computer programs were not so intelligent, they used to work as they were programmed. They were dedicated to do some specific tasks and are not capable of having knowledge.

Going further back in 1969, Apollo Guidance Computer took Man on moon had approximately 64Kbyte of memory and operated at 0.043MHz. Today my smart phone runs on 3GB memory and 2.1 GHz QuadCore processor. It is roughly 200K times faster that Apollo11. Magical !! Isn’t it??

All this has come true due to advancements in the field of hardware and software and one man was true about his predictions. In 1965 Intel co-founder Gordon Moore wrote his vision of how the performance of computer hardware would double every 18 months using the same size. This has hold true so far and the here is a clear example from my life:

Moore’s observation transformed computing from a rare and expensive venture into a pervasive and affordable necessity. All of the modern computing technology we know and enjoy sprang from the foundation laid by him. From the Internet itself, to social media and modern data analytics, all these innovations stem directly from his findings.

The field of tech has changed tremendously. Desktop computing is moving towards distributed cloud computing, desktop apps are moving to WebApps and Mobile Apps, hardware has gone cheaper, internet has gone cheaper and computation cost has gone cheaper too.

The conditions were never so perfect for computationally extensive operations like ML. Apparently most of ML concepts that we use today were presented in 1980s but they were ahead of their time. Researchers couldn’t use these concepts because the hardware needed was not affordable.

In past, it used to take around 8 days to train a machine just to identify signatures. Today, my PC can learn to categorise 1000 different objects from images in some minutes of learning with a very high accuracy. All of this has come true because Moore’s Law was correct. Because to this small firms are innovating and using ML/AI to disrupt the businesses which were predominantly occupied by Googles and IBMs of the world.

Today machine learning is used in all fields like Financial Markets, Medical, Technology, Weather Forecasting, NLP, Pollution Control and many more.

Today’s state of art machine learning programs can :

1. Drive your car better than you
2. Play games better than humans
3. Summarise content for you in fewer words
4. Read human expressions better than humans
5. IBM Watson, a machine, that can play jeopardy better than any human because of the content it can learn and store
6. Recommend products that you will like and purchase
7. Predict game results
8. Process Natural Language in multiple languages even in noisy conditions

And Many more…

Things are just going to get better, hardware is going to get cheaper, smaller and faster. This will encourage Machine Learning experts to be more courageous to create smarter & faster machines. More and more things will be done by machines, this doesn’t mean that humans will just sit and eat like in the movie Wall-e. Instead we will work on things which are more difficult and needs more creativity and intelligence than a machine can ever have. I am up for this challenge, are you?