Why ‘now’ is the time to be a ML developer (in India)?

Few application developers that I speak to tell me that ‘machine learning is all hype. they don’t want to invest time in learning it.’ Even if this phenomenon happens, they claim that huge companies (Google, Microsoft) are building APIs and they are happy to build applications by leveraging them, no real need to learn it. Here is my sense on why we need to get involved deeply and ‘right now’ is the time to do it.

More than a millennia ago, there were many dynasties that ruled India for many centuries. In South India specially, we had Pallavas, Cholas, Pandyas, Hoysalas to name a few. They built stone temples of great grandeur and at the same time artistically aesthetic. Hoysalas at Belur & Halabeedu, Pallavas at Mahabalipuram, Cholas at Thanjavur, Dharasuram are some of the well-known among the hundreds of temples they built. I wonder, if I were a sulptor, would I rather have lived a millennia ago working hand in hand with some of the greats that built these temples or live in obscurity today?

Indian railways started predicting the ‘probability of a wait listed ticket to be confirmed’ a few months ago. Now, they are building ‘smart coaches’ with sensors that will aid in predictive maintenance. Liv.ai has built a speech recognition system from the ground-up for Indian languages.

Lot of similar great stories in Agriculture (our core strength), Education, Manufacturing are being built today and will continue in future too!

The fundamental building block for these applications is Annotated data. (At least for now, this is the only AI that seems to work well). We have it in abundance now if reaped correctly.

Niti Aayog published AI strategy for India that talks at length on India’s strenghts and challenges. It calls out, India are just getting started in AI whereas most of the countries are ahead of the curve. However, it envisions that, we have huge opportunities to build an inclusive India with AI. Novel ideas are the need of the moment in India, it explains and we have to play this game to our strengthens on the backbone of our younger population.

Top it all, people in ML boast of an open culture sharing knowledge as much as possible with others. Any new starter is always welcome and has lot of opportunities to learn. This is the best of times to be a software developer in 2.0 paradigm (as Andrej karpathy explains in his article)

Above all, one of my friends is fascinated by the back-propagation (heart of Neural network learning) and he says ‘I’ll learn ML on any day just for the beauty of backprop and the maths involved’

Like the sculptors of the last millennia that left ever lasting temples and monuments as their legacy, developers like us have an opportunity today to build stuff that can transform industries, bring poor out of poverty, healthcare and quality education accessible in villages of India and many more with technology driven by AI.

These are good enough reasons for me to take the plunge. What about you?

Venugopal Rajaraman

Written by

Software developer, Machine learning enthusiast