The future of business: AI is here to stay; it’s up to us how we apply the tech
Be it Deep Learning, Machine Learning or AI, Indian firms and leaders are looking at technology as an opportunity to enable humans and remove mundane jobs.
Ray Kurzweil, American computer scientist, author and well-known futurist, has made an incredible number of predictions on Artificial Intelligence (AI) since 1990. What were they? That a computer would defeat a world chess champion by 1998, people would be able to talk to their computers through commands, and more. More recently, he predicted that AI would reach human levels in a decade and would have multiplied intelligence a billion-fold in 30 years.
Every company, from Google to Facebook, from Oracle to Microsoft and SAP, is now working on projects and platforms where software can learn and make decisions without hordes of people managing the business.
Indian companies and corporate leaders are also looking at AI as an opportunity to enable society rather than creating doomsday theories about the destruction of society because of software that can think for itself.
But first, let’s know more in 30 years about AI. For that, we need to begin with Deep Learning. The panel on the first day of the Bengaluru Tech Summit, titled: 'The AI Disruption wave -- How AI and Deep Learning will disrupt the world,' had Prithvijit Roy, CEO, Bridgei2i Analytics Solutions, Sudara Ramalingam Nagalingam, Head Deep Learning practice, NVIDIA Graphics, Manav Garg, CEO and Founder, Eka Software Solutions, and Sudhanshu Singh S, Senior Vice President, Analysis and Research, Genpact discussed the AI wave.
Deep Learning vs the brain
Sundaram Ramalingam, Head, Deep Learning Practice for Nvidia Graphics, said, “To understand Deep Learning, look at the brain. Software is trained to mimic the way the human brain works.”
For example, how does a child know that an apple is an apple. It is because the child is told by the parent that it is an apple and the brain processes the learning. Three days later, if you give a child a green apple, the child can correlate to the learning and say it's also an apple, provided the parent told the child that green apples also exist. The machine needs to be trained like this with multiple variants; you then write algorithms to extract features of the object and the machines learn about the object.
Apart from Deep Learning , there is Machine Learning where you extract features of the product (apple) and teach the machine about the details.
“It is a great field for all of us,” Sundaram said.
AI, on the other hand, works in predictive maintenance.
Sudhanshu Singh, SVP, Genpact, said, “We are using AI in aircraft. Every time an aircraft flies it is nothing but a bunch of computers and algorithms. We have been using historical data from several aircraft to model reliability of parts and forecast the failure rate. This is where real-time data helps.” He added that the field engineer knows what the problem with the plane is before it takes off. The airline can source parts quickly and fix the problem in any zone they are flying in to.
AI: recommend, act and learn
So, is AI really about the ability to learn, sense and act?
Prithvijit Roy, founder Bridgei2i Analytics Solutions, said, “The autonomous car writes its own software to make decisions. Even in healthcare a doctor gets suggestions on treatments when there is AI. So it is about recommend, act and learn to enable the human or replace the human.”
AI is all about self-learning software. AI can track fraud in finance on a real-time basis or even underwrite credit. The machine uses text and video at volume and variety to make these decisions. There are a variety of applications for AI.
“It is about outcomes that people get attracted to and about improving customer experience,” Prithvijit said.
However, all AI works when business learns how to apply the technology.
Manav Garg, founder, EKA Software Solutions, said, “From a business perspective, all old economy businesses, such as companies growing crops, for example, can use data to understand, on a real-time basis, weather and soil conditions to take precautions to sow crops.”
He added that even manufacturing will go through mass scale automation.
AI is all set to replace mundane and repetitive jobs in the IT and every other industry. There are problems, of course, because the quality of the application depends on the availability of data. Today, the data in old economies is difficult to get unlike Google and FB, which practically know everything about their customers.
“The challenge is to create great data models to build AI,” Garg said.
Despite being ahead of the curve in tech, we continue to lag in adoption. It’s imperative to manage how people will use technology and also understand its application in business. The biggest result? Customers will pay for efficiencies. That’s the reason AI makes sense.
Today Adidas moved its plant back to Germany, from China, because robots can manufacture better than people. This is the future. The biggest takeaway at the Bangalore Tech Summit is that AI is here to stay and society must decide whether it will reskill people or enable them with AI.
