IBM Chairman and CEO Arvind Krishna on Monday urged investments by the government to develop sovereign capability around artificial intelligence. “Every country ought to have sovereign capability in artificial intelligence, including large language models,” he said, adding, “You might want to use it for purposes the rest of the world does not want to invest in… or for purposes that you may not want to expose to the rest of the world.”
He said that the country needed capability around computing, storage and datasets.
Asked what kinds of investments India should make, he said it could be a ‘National AI Computing Centre that helps create some government datasets, maybe in agriculture, health… where the government applies those use cases. So that creates a market.” Such a centre could be formed, he added, “under one of the existing bodies, or under a new body – the same way it did for supercomputing under CDAC.”
Addressing select media persons on his visit to New Delhi for the B20 Summit, Krishna said AI would create more jobs even as it enhances productivity. Asked if that would mean more AI jobs in India, he cited McKinsey’s estimate that artificial intelligence would drive $4.4 trillion of annual productivity by 2030. “For this to happen, you need 4.4 million AI workers, even assuming each drives a million dollars worth of productivity which itself is a high number. India could (potentially) contribute a couple of million people who are skilled in AI.”
“The entire developed world is witnessing a decrease in working-age population. To lift GDP growth, and raise GDP per capita, the only way is to have digital labour augment human labour. The only tool that can help us do that is artificial intelligence. India can be a great place as a source for deploying it for other countries.”
“India rode two waves very successfully – the BPO wave and the separate, IT services wave. This (AI wave) is the third one it can ride very successfully.”
India business to grow
On IBM’s India business, he said; “I fully expect that our business in India is going to grow.” He also said that he expected IBM’s total investment in India would keep growing, too; however, he did not elaborate on the quantum of investment.
Mr. Krishna said he was optimistic about technology spending globally. “Technology spending is typically between 2-4% faster than global GDP growth. (Given the IMF’s estimates for global economic growth), technology spending is going to sit somewhere around 5-6%.”
On the work in AI that IBM does in India, he said, “IBM India reflects our global footprint. Our researchers here do some of the deepest work that we have going on in AI anywhere in the world. Our consulting and BPO teams are helping deploy AI. We have teams building products that use AI. As to people using and deploying AI, (that would run) into the 10s of 1000s (of staff).”
Dwelling on the fact that long before the advent of AI as we know it today, IBM pioneered AI research. At its peak, the company‘s Deep Blue machine defeated chess champion Garry Kasparov in 1999 in a first. “The difference now is that, with the advent of large language models, the cost of creating these models has come way down. Earlier, the deep learning phase of AI was incredibly powerful, but very hard to accomplish. To create a model for something, we had to find lots of data, get lots of people to label the data, set it up and after hundreds of people worth of investment, and six months of training, we had a model that could do one thing!”
“Now, you can get a new task out of AI with one person over the weekend. That is why I feel the current AI will get these monetisation numbers, unlike the previous generations.”