Not much has changed over the past three decades, with the exodus growing in scale each year, according to data from the Bureau of Immigration.

This brain drain phenomena is a critical challenge as India strives to become a global tech powerhouse. With so much happening on the technological front - AI and then some more AI - it got us wondering: Are we also losing the crème de la crème of our AI talent to the west?

The Globetrotting AI Professionals Of Tomorrow

As it turns out, yes. A significant proportion of India’s AI talent is going abroad to study, finding jobs and eventually settling in the US or Euopean nations. In 2019, nearly all Indian AI researchers opted to pursue opportunities abroad, primarily in the United States. Though others also opted for Australia, the United Kingdom and other European countries. 

Studying ‘abroad,’ as we call it, has neither been easy nor cheap. And it still isn’t. So what has motivated students? The Secretariat spoke to a small proportion of this ‘crème-de-la-crème’ AI talent, some of whom are enrolled in the world’s top universities.

Specialising in a subfield called "interpretable" machine learning, Suraj Srinivas completed his PhD in Machine Learning from the EPFL research institute in Switzerland and is now a postdoctorate research fellow at Harvard University.

Like many other students that we spoke to, Srinivas was motivated by the ability to do more impactful research, access to better opportunities, and getting to interact with a diverse set of researchers and professors from all over the world.

“Broadly, there are a much larger number of opportunities to be a top faculty or a researcher outside India,” said Srinivas.

There are, although, few places in India which are considered pioneers in AI research, such as the Indian Institute of Science (IISc), select IITs, Microsoft Research, Google Research, among others. 

Srinivas also holds an MS in machine learning from IISc. Even though he believes that IISc offers rival courses, comparable to EPFL and Harvard, the number of such places is very small compared to the number of such opportunities in Europe or US.

This is evident in rankings as well. None of India's institutes figure in the top 25 category for Top-Tier AI Research in the world. In the top AI institutions in the Asia-Pacific region, only Indian Institute of Technology (IIT) takes the 17th position.

The Secretariat spoke to an alumni of IIT (Madras), Abhishek Sharma, who like Srinivas is also enrolled as a PhD (machine learning) candidate at Harvard.

“I don't think retaining top talent is a priority for India. The government funding for research is low, and the industry is not as mature as in the US. However, the situation is improving from what I hear,” said Sharma.

Megha Kalia, an IIT Kharagpur graduate, who is now a post-doctoral research fellow at Harvard Medical School said that she was grateful to have studied at IIT where she got to know about so many career paths and opportunities abroad.

But both Kalia and Sharma echoed Srinivas in the feeling that the US has better infrastructure, funding and research opportunities. And it didn’t hurt, he added, that the jobs in the US come with a much higher pay. 

“The demand for Senior AI talent is very high abroad (both in industry and academia) and they are also able to pay a very high salary to hire these talents,” explained Chirag Jain, CEO and founder of Textify AI, who completed his education from Seoul National University, South Korea.

For example, OpenAI invests about $1 million annually on each AI researcher. While India has companies with market capitalizations comparable to global giants, the disparity lies in their R&D spending. Indian companies generally allocate significantly less towards research and development compared to their international counterparts.

Sharma’s story, however, as an alumna of IIT, speaks to a larger problem in India.

A  2023 study of the 1,000 highest scorers in the 2010 IIT joint entrance exams found that those who attended the top IITs (Madras, Delhi, Bombay, Kanpur, Roorkee, and Kharagpur) are more likely to migrate for graduate school.

The scale of the ‘problem’ in these prestigious centrally-funded universities is that of the top 100 scorers, 62 per cent leave the country whereas around 36 per cent of the top 1,000 scorers migrate abroad.

To get a sense of the difference in graduate experiences, The Secretariat spoke to Khurram Afroz, who is pursuing PhD in Machine Learning from IIT Delhi.

“Brain drain is a significant concern among top IIT graduates and their peers. Many feel that the opportunities, research facilities, and financial rewards abroad are more attractive,” he said.

However, Afroz also said that the quality of research at IITs is excellent, with many students publishing their work in top AI conferences and journals on a regular basis.

In the scheme of these developments, what would have been the natural path for India was better research infrastructure, an increase in funding mechanisms, better public-private partnerships and tapping the young scientists to pave the growth of a robust scientific ecosystem.

But that quite didn’t happen. India, even though it produces the third highest number of elite AI researchers* and highest number of PhD holders after the US and China, still doesn’t have the relevant infrastructure.

Note: *The “most elite AI researchers” are defined as authors of papers selected for Oral Presentations at NeurIPS, which represent the most prestigious class of papers. The Oral Presentations acceptance rate was 1.8 per cent in 2022. Source: Global AI Talent Tracker

Unequal access to AI infrastructure, that is the percentage of the Indian student population getting access to AI tools, is one part of the problem. India's investment in research and development has consistently been identified as an area needing improvement.

In 2020-21, India spent 0.64 per cent of its GDP on research and development, according to Ministry of Science & Technology data, falling behind major developed economies which are spending gigantic amounts of money into R&D, such as China (2.4 per cent), Germany (3.1 per cent), South Korea (4.8 per cent) and the United States (3.5 per cent).

So, can India catch up in retaining AI talent? The government seems to be doing its part in improving the infrastructure back home. In March, the Modi government announced it would be injecting over Rs 10,000 crore into its IndiaAI Mission. But there is still a long way to go.

“The entire ecosystem is in shambles in some sense. Look at China, they are now at a level-playing field with the US. But India is nowhere close. And to compete with them is not just with investment. Should we just invest 50,000 crore instead of 10,000 crore? Or should we do something that would give us a competitive edge?,” asked Ajith Sahasranamam, founder and CEO of Ongil.ai.  Sahasranamam holds a PhD in Computational Neuroscience from  the University of Freiburg, Germany.

In comparison, China has an AI talent pool much larger than India, evidenced by a greater number of top-tier AI publications and deep tech startups, despite similar population levels. Some of China’s initiatives make it easier for their AI talent to work and innovate in the country itself. 

“(Indian) conditions need to be more alluring for high skilled workers, similar to China’s framework, to retain talent,” said Sarwari Das, an Indian pre-doctoral fellow at Harvard Business School, who builds AI models to predict an entity’s financial behaviour.

The Chinese framework, as Das put it, includes policies to support AI development, including tax incentives, subsidies for tech startups, and streamlined visa processes for foreign AI experts.

Their global standing is underscored by their position as one of the largest producers of elite AI talent worldwide. According to the Global AI Talent Tracker, China's contribution to elite AI researchers in the Asia-Pacific region surged from 67.2 per cent in 2019 to 81.9 per cent by 2022.

India's top tier AI talent who published elite papers on the other hand has shrunk from 18.6 per cent in 2019 to 8.2 per cent in 2022. The shrinking does not reflect on the number of papers produced by Indians but on the huge explosion of AI talent within China.

On a global scale, China contributes significantly to the pool of top AI researchers, with its share rising from 29 per cent in 2019 to 47 per cent in 2022. This trend explains why more Chinese talent are opting to work within their domestic industry.

An AI Homecoming?

Afroz, like Sharma and Srinivas, had considered studying and working abroad but his decision to stay in India was motivated by personal factors. “Staying in India allows me to contribute to the local ecosystem and be close to family,” he added.

Afroz is not the only one. It seems that despite being unable to bridge the gap between funding and retaining top AI talent, some students are returning home. In 2019, nearly all Indian elite AI researchers opted to pursue opportunities abroad, but in 2022, 20 per cent of them stayed back.

The Secretariat spoke to Anand Louis, an Associate Professor at the Department of Computer Science and Automation at IISc. He was a postdoctoral researcher in the Department of Computer Science at Princeton University, and pursued his PhD in from Georgia Tech. He too, like Afroz, is a graduate of IIT Delhi.

Louis said his motivation comes from wanting to be in the country he grew up in and being close to his family. 

“Indian universities have some well documented shortcomings. Instead of sitting abroad and complaining about them, I wanted to do my part in helping overcome them,” added Louis.

Factors affecting the decisions of Indians abroad to return home also include distance from family and cultural roots, a higher living cost and visa challenges, as well as difficulty integrating into local societies in some places.

Louis’ colleague at IISc, Danish Pruthi had similar reasons. He did his Master’s in Machine Learning and a PhD from Carnegie Mellon University, which is one of the most prestigious universities for AI learning in the world. In the time he was in the US, he got the opportunity to work in the AI labs of Amazon, Google, Facebook and Microsoft. So, why come home?

“India is an important country. And we are one of the leading economies, we have one of the largest populations, but unfortunately, we don't have a lot of AI research labs, especially in leading Indian universities,” said Pruthi. 

Pruthi felt that there is a much larger gap to fill in India. His classes are consistently well-attended, reflecting a large pool of talent. He noticed that despite the outflow of highly trained individuals leaving, the interest and enthusiasm in learning about AI in India remain robust and many want to contribute locally.

“The job anyway requires me to operate pretty much independently and develop a group from scratch. And I could do it anywhere. So why not try this in India,” he added.

India stands at a pivotal intersection where it can either be a key voice in the AI industry or watch the action unfold from the cheap seats at the back. So, how can India ensure that its first-class AI talent stays back and contribute to an industry whose market size is expected to reach US$28.36 billion by 2030?

The obvious solutions are investing in research facilities and universities, so that India can harness research potential early on.

“The basic principle is: hire the smartest researchers, offer them the best resources to do open-ended research with minimal bureaucracy. Top talent tends to congregate around other top talent, so this model can lead to a positive self-affirming loop,” said Srinivas.

In the end, The Secretariat  did a dip stick survey and asked AI researchers settled abroad if they are planning on coming back to India anytime soon. Almost all said no.