Vishwagurus of AI
Finding needles in a haystack
The IITians inventing the future
As one of many jury members to choose the top hundred ai experts of Indian origin, I get to speak to the who’s who of Ai. Whilst ai may seem to be a new field to many, actually it is not. Five of the eleven jury members are PhDs in Ai. Two of them – Jyoti Joglekar and Swami Suryanil 🙏🏻 are engineering graduates of the eighties. Jyoti has been teaching Ai for a couple of decades and Swamiji was doing post doc work in Ai at the max planck institute in Germany decades ago after completing his PhD in Ai way back in 1992. He is now a monk of the Ramkrishna Mission order – so in a way entering van prastha.
My college batchmates – Dr. Harrick Vin , Raminder Bajwa , CP Mammen and Ambarish Malpani are all major fundamen in ai. They are my goto guides on ai. Harrick is the global CTO of TCS. Bajwa was part of Google and Mammen is Nvidia. Thanks to them, I at least now know the right questions, the answers hopefully will follow.
Just when I thought we had the final shortlist of 500 done to get to the top 100 – one or two more pop up daily. My mailbox today showed up Alok Aggarwal , Chief Scientist of Scry Ai in the valley. I first met Alok when he had set up Evalueserve. It was the first of the KPOs and it was quite a popular employer. Of memory serves me right, it was the first to offer those Rs 50 lac type packages in IIT. The others like Google then realised that it was a great way to build a brand, so a race to offer obscene salaries started. Those salaries were rarely paid out but they did up the suicide rate in IITs because every IITian was expected to start his career at usd 100k or more in India.
Alok plans to write a book and thinks that the limits of ai are determined by the economics of ai, the accuracy of ai and the safety of ai. 90% accuracy is ok for a google search equivalent but not for anything serious.
To take an analogy from Covid and genetics. At 95% accuracy, u won’t know if it is man or a monkey. At 92% you won’t know whether it is a shark or a monkey. And at 99.9% accuracy u won’t know if it is a Caucasian or an Asian. To know the likely color of his hair, we need to look at 99.999%. And we may be wrong.
All of the above which ai struggles with at times, are obvious to a kid, probably without even much schooling – leave aside college.
Alok raises some interesting points in the presentation that he shared with me. He is clearly someone who will make it to the shortlist, if he applies that is. I am looking forward for someone to nominate him at applications@iitalumnicouncil.org
A nomination for Aravind Srinivas from the Chennai campus of IIT hit the nominee list last week and I am hoping he will make it. He is the founder of perplexity and the person nominating him seems to believe he has a good chance of making it. This is one interview I am looking forward to.
Americans do hype things well. But substance seems to be nearer to India.
