Coverage by Bhat Dittakavi of Variance.AI on “AI Capabilities” by BVR Mohan Reddy ji at Zettamine on 12th May 2018.
Almost 36 years ago, on a Friday evening
in Pittsburgh, I was talking to Raj Reddy, only Turing award winner from India. Raj was very optimistic about AI at that time. He said driverless car would happen. When I asked him the time line, he took me around in the campus and amazed me with a driverless car. It was twice the size of cars of current generation. It had a truck attached to it. It only drove certain paths. We discussed few limitations.
Today, tremendous amount of programs have been made. The truck was carrying processing computers then. Over the decades, with the Moore’s law, power of computing has gone up while the hardware has become smaller and cheaper. This is the secret behind today’s AI.
Connectivity is the next big thing. Internet and HTTP. People started to connect any machine and people any time in the world. Ww waited for 8 years to get telephone connection to home in my days. Today an young man on the streets got two phones and is connected to the world. Today we got 20 billion things and people are connected in the world.
Third one is a sensors that detect many things. Temperature sensor used to be a big.boz but now we have a 3mmx3mm sensor. My wife is a diabetic and her doctor gave her a wearable. I couldn’t see. See showed it to me.just below her blouse. It is a 3mmx3mm supplant with one rupee size equipment. Doctor told her to write the food habits for 15 days and data from the chip and her food data are superimposed to analyze.
Miniaturization of devices. 15 days and 15 minutes of daily data and it costed just Rs1800. Buy the battery is a concern. During the night you out a calf over the wearable to charge overnight as the patient is sleeping.
Sensors are producing enormous amount of data. Majority of our business is into aircrafts. On a flight, we got 4000 sensors that send data continuously. Information such as engine vibrations and so on. They out 12-18 Tera flops of data by the time fought comes down.
Computing power, connectivity and sensors are complete if we have algorithms that make meaning out of them. These are four accelerators. Now we add amplifiers.
One technology rides on top of another: Ola and Paytm. One depends on the other. Paytm depends on digital payments. Amplification happens because of globalization. Countries are separated by lines in sand. Small wind changes the contours. Technology will be all pervasive.
The truck had disappeared and large size communication equipment is gone. Now enormous amount of AI is part of our mainstream of life.
Future or AI
US is on the leading edge with largest investment going in AI. Growth is 48% CAGR. 2025 will have $50B in funding. AI can be applied anywhere. Gartner is predicting 2.3M people are needed to work on AI applications by 2022. 75% of all IT applications go on AI. If AI I’d there our GDP would get additional 1.3%. AI research institutions we have. NASSCOM is setting AI centre of excellence in Hyderabad.
Will AI talk off jobs?
Yes. But AI will create more number of jobs. Net net there will be job creation. That has been the trend over the eras. Stone to steel to industrial revolution 1 to industrial revolution 2 to industrial revolution 3 and beyond.
I remember a 1904 picture from a vantage point in New York that has got just one car and all horse buggies. There is another if recent times that has all cars except one car buggy. We didn’t kill the horses or Horsemen. They got re-skilled. Machines handle repetition very elegant and people get more productive. AI makes doing different things differently. Look at companies like Mynthra, Lenskraft, PAYTM and many that use AI in India. Even Telangana government does it too.
Healthcare and agriculture excite me. I invested in 16 and 14 of startups as early stage invesfor. All of them are tech startup last. We are running a project with drones in India. Because of the way we have structured drones, it turned out that these drones have become multi-purpose. Drone is now a service provider. Extent of land is small but we can use the drones for inspection fo rail tracks. Hazards can be taken care of AI. Inspecting huge chimneys require scaffolding that get
built in 15 days
. Kids from IIT Chennai came on a Monday morning to Coramondel plant in Kakinada a and gave a report using drones all within 24 hours.
After we flew the drones in agricultural land, we could get soil, water and time to harvest information where our intent originally was different.
An ICU bed costs $100,000 and an IoT costs $100. Accuracy may not be there but direction of accuracy is never wrong. ECGs have now 10 leads and all clumsy. Now the sensors are in a glove. There is a washable t-shirts that track your data. Data is not anymore sinusoidal. It is just red, green and yellow lights. Lifestyle helps us predict when the person may get heart attack.
Data. People. Technology.
People make technology. Key driver for all of this is technology and talent. We need extremely talented pool of people.
Global AI pool:
One credible report showed 10,000 to 20,000 is the total strength of AI people in
India got a thousand of them. Tencent analytics India thinks India for 30000 out of 300000 of world. We got 4 million IT professionals in India. Only a small percentage are into AI. Hence there is a need for skilling in India. STEM is key. We produce 7.5 million graduates each year. How many are placable?
India suffers in university education and multi-faceted stakeholders.
NASSCOM Skills of the future: AI is one among them.
Beauty of technology disruption is that it makes all of us extremely obsolete very quickly. There is a positive spin too. In 1976 I was at IIT. I went to a library and the librarian told me the book was given out and I was third in queue. I would take six weeks to start learning. Today we do anytime and anywhere.
Collaboration is the way forward. I am delighted NIT and CSI and Zettamine coming together.
Mohan Lakkamraju is doing emerging technologies things too. IIIT is doing. NASSCOM too. Many of us shall do it.