Should we hurry up with policies around AI?
There has to be a significant engagement by the policymakers. I don't know if there is enough understanding of the pace at which the AI space is developing. Eventually, we do need policies but it has to be done – given the nature of how fast it is evolving. Moreover, it has to be done in a very thoughtful way by people who have a fairly deep understanding of the issues.
A recent Goldman Sachs report mentioned that as many as 300 million jobs could be lost due to AI systems. Do you think such job losses could happen in the next 10 years?
I don't think so. Generative AI technologies require a human in the loop for a lot of activities because it can make spectacular errors. You need human judgment. But even as this develops, it could completely replace human functions in certain areas but there will be other newer things that will be created – that's always been the history. I mean, they predicted that accountants will go out of business when Excel was first introduced. Now, there are more accountants in the world, per capita, than there were at that time.
So, the nature of jobs will change?
Yes, there will be more job categories or jobs that we do not know anything about right now. For instance, nobody knew that social media influencers/ content creators would be a job category 20 years ago. But the dystopian scenario of everybody sitting around with nothing to do and the state has to feed them, I think that is a little bit farfetched.
Do you think countries like India should have AI-first policies to foster innovation at a national level? Because monopolies could easily form on the global stage, especially, with the US and China being far ahead in AI innovation?
I certainly think this will become a contentious topic in the future. Because national competitiveness is affected by the degree of utilisation of machine learning and AI models. So, certainly governments like India should be investing in AI – not to build giant models which will compete with the ‘latest and greatest’ from OpenAI or Google – but they need to foster the creation of these models and the usage of these models as part of the national infrastructure. This could be considered critical national infrastructure.
You mean, something like the India Stack?
Yes. India Stack was done with the support of the government. You need similar efforts here because this thing is going to be one of the most significant determinants of technological superiority among nations. And it's going to affect all forms of competitiveness, including economic, military and more. Hence, it's super-critical for each country, certainly a country like India, as it aspires to be the top three global powerhouses in the world.
Should we be wary about anything with AI development?
The companies that are building these new AI models and developing the technology, they understand the power of these things better than most people like you and me – who are on the outside. Hence, the onus is on them to both educate and put the safeguards as needed. That is where the higher sense of responsibility needs to come from.