The growth of Artificial Intelligence (AI) technologies and their deployment has raised questions about privacy, monopolisation and job losses. In a discussion moderated by Prashanth Perumal J., Ajay Shah and Apar Gupta discuss concerns about the economic and privacy implications of AI as countries try to design regulations to prevent the possible misuse of AI by individuals and governments. Edited excerpts:
Should we fear AI? Is AI any different from other disruptive technologies?
Ajay Shah: Technological change improves aggregate productivity, and the output of society goes up as a result. People today are vastly better off than they were because of technology, whether it is of 200 years ago or 5,000 years ago. There is nothing special or different this time around with AI. This is just another round of machines being used to increase productivity.
Apar Gupta: I broadly echo Ajay’s views. And alongside that, I would say that in our popular culture, quite often we have people who think about AI as a killer robot — that is, in terms of AI becoming autonomous. However, I think the primary risks which are emerging from AI happen to be the same risks which we have seen with other digital technologies, such as how political systems integrate those technologies. We must not forget that some AI-based systems are already operational and have been used for some time. For instance, AI is used today in facial recognition in airports in India and also by law-enforcement agencies. There needs to be a greater level of critical thought, study and understanding of the social and economic impact of any new technology.
Ajay Shah: If I may broaden this discussion slightly, there’s a useful phrase called AGI, which stands for artificial general intelligence, which people are using to emphasise the uniqueness and capability of the human mind. The human mind has general intelligence. You could show me a problem that I have never seen before, and I would be able to think about it from scratch and be able to try to solve it, which is not something these machines know how to do. So, I feel there’s a lot of loose talk around AI. ChatGPT is just one big glorified database of everything that has been written on the Internet. And it should not be mistaken for the genuine human capability to think, to invent, to have a consciousness, and to wake up with the urge to do something. I think the word ‘AI’ is a bit of a marketing hype.
Do you think the current regulatory system is equipped enough to deal with the privacy and competition threats arising from AI?
Ajay Shah: One important question in the field of technology policy in India is about checks and balances. What kind of data should the government have about us? What kind of surveillance powers should the government have over us? What are the new kinds of harm that come about when governments use technologies in a certain way? There is also one big concern about the use of modern computer technology and the legibility of our lives — the way our lives are laid bare to the government.
Apar Gupta: Beyond the policy conversation, I think we also need laws for the deployment of AI-based systems to comply with Supreme Court requirements under the right to privacy judgment for specific use-cases such as facial recognition. A lot of police departments and a lot of State governments are using this technology and it comes with error rates that have very different manifestations. This may result in exclusion, harassment, etc., so there needs to be a level of restraint. We should start paying greater attention to the conversations happening in Europe around AI and the risk assessment approach (adopted by regulators in Europe and other foreign countries) as it may serve as an influential model for us.
Ajay Shah: Coming to competition, I am not that worried about the presence or absence of competition in this field. Because on a global scale, it appears that there are many players. Already we can see OpenAI and Microsoft collaborating on one line of attack; we can also see Facebook, which is now called Meta, building in this space; and of course, we have the giant and potentially the best in the game, Google. And there are at least five or 10 others. This is a nice reminder of the extent to which technical dynamism generates checks and balances of its own. For example, we have seen how ChatGPT has raised a new level of competitive dynamics around Google Search. One year ago, we would have said that the world has a problem because Google is the dominant vendor among search engines. And that was true for some time. Today, suddenly, it seems that this game is wide open all over again; it suddenly looks like the global market for search is more competitive than it used to be. And when it comes to the competition between Microsoft and Google on search, we in India are spectators. I don’t see a whole lot of value that can be added in India, so I don’t get excited about appropriating extraterritorial jurisdiction. When it comes to issues such as what the Indian police do with face recognition, nobody else is going to solve it for us. We should always remember India is a poor country where regulatory and state capacity is very limited. So, the work that is done here will generally be of low quality.
Apar Gupta: The tech landscape is dominated by Big Tech, and it’s because they have a computing power advantage, a data advantage, and a geopolitical advantage. It is possible that at this time when AI is going to unleash the next level of technology innovation, the pre-existing firms, which may be Microsoft, Google, Meta, etc., may deepen their domination.
How do you see India handling AI vis-à-vis China’s authoritarian use of AI?
Ajay Shah: In China, they have built a Chinese firewall and cut off users in China from the Internet. This is not that unlike what has started happening in India where many websites are being increasingly cut off from Indian users. The people connected with the ruling party in China get monopoly powers to build products that look like global products. They steal ideas and then design and make local versions in China, and somebody makes money out of that. That’s broadly the Chinese approach and it makes many billion dollars of market cap. But it also comes at the price of mediocrity and stagnation, because when you are just copying things, you are not at the frontier and you will not develop genuine scientific and technical knowledge. So far in India, there is decent political support for globalisation, integration into the world economy and full participation by foreign companies in India. Economic nationalism, where somehow the government is supposed to cut off foreign companies from operating in India, is not yet a dominant impulse here. So, I think that there is fundamental superiority in the Indian way, but I recognise that there is a certain percentage of India that would like the China model.
Apar Gupta: I would just like to caution people who are taken in by the attractiveness of the China model — it relies on a form of political control, which itself is completely incompatible in India.
How do you see Zoho Corporation CEO Sridhar Vembu’s comments that AI would completely replace all existing jobs and that demand for goods would drop as people lose their jobs?
Ajay Shah: As a card-carrying economist, I would just say that we should always focus on the word ‘productivity’. It’s good for society when human beings produce more output per unit hour as that makes us more prosperous. People who lose jobs will see job opportunities multiplying in other areas. My favourite story is from a newspaper column written by Ila Patnaik. There used to be over one million STD-ISD booths in India, each of which employed one or two people. So there were 1-2 million jobs of operating an STD-ISD booth in India. And then mobile phones came and there was great hand-wringing that millions of people would lose their jobs. In the end, the productivity of the country went up. So I don’t worry so much about the reallocation of jobs. The labour market does this every day — prices move in the labour market, and then people start choosing what kind of jobs they want to do.
Ajay Shah is Research Professor of Business at O.P. Jindal Global University, Sonipat; Apar Gupta is executive director of the Internet Freedom Foundation
