AI - Is it at our doors?

I attended a conference by O’Reilly about Data Science in New York last week and on the last day, a keynote by Cassie Kozyrkov jolted my mind to every extent. Is really AI knocking on our doors? Is it really the time to put a foot on the accelerator? There are too many questions and too many speculations, but the reality is by far the scariest one for novice data scientists. I believe that in next three to four years everyone will be an analyst because if you are not then you might not be useful with reality. Every second we are generating data at a speed of 1.75 MB per person per second, which means that we are being surrounded by giant containers of datasets and the only way to survive in this puddle is by learning how to handle those data and get some meaningful insight from it. Also, automation is already making its space in the market, where Deloitte’s new product will eliminate professional analysts and anyone without a proper coding background can access meaningful insights from data or big data which can be handled much easier than anything else. But the question which is alarming is what’s next?

The world has changed and already there is vibration about AI in the market. We all must have heard that in the next 3-4 years AI would be in market, but that old news is now a reality and it is high time to understand that AI is future. If you are just starting in the Data Science world, buckle up because things are changing at the pace of ticking watch. If I want to explain what AI is, I would define it as a process where a user defines a few objectives and machine makes sure that those objectives get fulfilled in a timely manner with utmost accuracy. With GPUs making an entry in the market, computation power has increased exponentially, and the development of new libraries is making the life of analysts easy. But is it enough? Absolutely not because the industry has developed at a pace where AI is now taking grounds. Data science with the inception of AI has made analytics powerful and if you are limiting yourself to just analytics, you might be a little behind in line. There is one more speculation that AI can eliminate human involvement but if I leave my Hollywood spirit and be logical about it, I will say, are you kidding me? If you have learned even a little bit of coding, then also you can be very sure that whatever a coder defines that only a system does. So, AI eliminating humans from the picture is a farfetched idea and is only available as specials on Netflix. But should we be feeling secure with our mediocre knowledge? Absolutely not, here is a catch because AI was hyped too much 5 years ago when Google started making a shift with a google voice assistant, but now AI is at our doors. We as a community of Data Scientists need to elevate our game and start exploring horizons where we are still lagging in terms of approaching problems analytically with lots of data and making system more powerful to take decisions. If I must wrap this article with an example, I will say think of labor-intensive industries such as smelting jobs where labors literally burn themselves to make a product. We have sensors installed all over places, can we build a system more efficient and robotic which can save human life and reduce his labor pain? I believe the answer is yes, it’s just we are not exploring enough and on the other hand, industrial firms are not ready to change their approach. You like it or not, AI is at our doors.