Data doesn’t lie. According to a study by Stanford University’s Institute for Human-Centred Artificial Intelligence (AI), India leads the world in key areas of AI and machine learning (ML). It, however, lags in other AI/ML areas.
Dice the data. India has the world’s second highest cohort of AI/ML engineering talent. According to the Stanford study, the United States has 786,000 AI/ML engineers, followed by India with 446,000 and China close behind with 400,000.
Cynics dismiss that as a natural consequence of India’s population and an existing software ecosystem upgrading to AI/ML. The argument hits a wall when relative AI skill penetration rate is measured. Stanford data reveals that for the period 2013-23, India had the “best relative AI skill penetration and prevalence of AI skills across occupations.” India scored 2.8 on this metric, ahead of the US (2.2), Germany (1.9), Canada (1.7), Israel (1.6) and Britain (1.6).
India does less well, however, on private investment in AI. In the decade between 2013 and 2023, India ranked seventh worldwide. The US led with private investment in AI totalling $335.2 billion, nearly triple China’s $130.7 billion and far ahead of India’s $8.9 billion.
While the number of Indian AI startups has surged in recent years, the country has a lot of catching up to do with developed economies. For example, between 2013 and 2023, 5,509 AI-related startups were set up in the US compared to just 338 AI startups in India. Nonetheless, India compares well with Japan (333 startups), France (391) and Canada (397) – all of whom have larger AI ecosystems and longer legacies. India’s deep reservoir of AI/ML and software talent, as the Stanford data suggests, will lead to the universe of Indian AI startups expanding rapidly in the next few years.
What the data holistically shows is that India is rich in human talent but poor in resources. The government’s budgetary allocation for India’s startup universe is expected to receive a boost when the full 2024-25 Union Budget is tabled in July 2024.
*Transformative
Kunal Shah is founder of the startup Cred. He earlier founded Freecharge. As a financial services entrepreneur, Shah believes AI is “brewing a revolution, like the steam engine and the internet did.”
In an article in The Times of India (May 15, 2024) Shah wrote: “The steam engine reduced inefficiency in transportation and started the Industrial Revolution; fertilisers reduced inefficiencies in agriculture and we called it the Green Revolution. The internet reduced the efficiency in information transmission and we called it the Digital Revolution. The largest remaining inefficiency is availability of intelligence and judgement of the highest quality for every individual. AI solves this.”
How can AI transform work in our daily lives? Shah provides an example: “For a short time I ran a teleradiology project where well-qualified radiologists from India assisted US radiologists in reading MRIs and X-rays, making them 3x more productive. It was quite successful but hard to scale because of constraints on expertise. AI can potentially do this within seconds. India produces a few hundred radiologists a year for our vast population; AI-assisted or AI radiologists can benefit both doctors and patients, and accelerate the accuracy, speed, and cost-effectiveness of healthcare.
“Imagine walking into an X-ray room and getting a high-quality radiologist’s report in 10 seconds for 50 rupees! With AI, this is a real possibility and the principle can be applied beyond healthcare, wherever strategic analysis is accessible only to a few experts currently.”
*Third industrial revolution
As a colony, India missed the first industrial revolution in the 18th and 19th centuries. India was quick to grasp the second industrial revolution in the 20th century driven by information technology and the internet. India is in a good position to be at the forefront of the third industrial revolution, charged by AI and ML.
An early indication, quite apart from the Stanford data, is the fact that India is the second largest contributor (19 per cent) to AI projects on GitHub, a software development platform owned by Microsoft. US developers form the largest contributors on GitHub (27 per cent). China is third with nine per cent.
Analysts expect Indian software developers to overtake US developers in absolute numbers on GitHub by as early as 2027. As the world’s largest developer platform, GitHub has AI-focused projects where China, the US and India are the most strongly represented.
OpenAI’s ChatGPT-4o (the o stands for omni) was unveiled recently. It claims to add “reason” across audio, video and text in real-time. What this means is that you can type in a simple command – say, recreate the climax scene in the movie Titanic – and you’ll get a reasonably good quality short video on how the Titanic sank with the ship’s band continuing to play as they went down.
Creators are worried that such advanced AI models will make scriptwriters, designers, filmmakers and even authors redundant. They needn’t worry. New AI models may have been programmed to add “reason” (code for emotion) to large language models (LLM) but AI still lacks original creativity. It can recreate Titanic, even tweak it, but ask it to make a brand new movie and – as yet – it can’t.
It’s important that AI is seen for what it is: a tool. It speeds up tasks. That helps researchers in, for example, life sciences to sequence molecules within hours instead of months in the search for life-saving drugs.
AI’s malignant potential to make deep fakes and clone voices has already led to the creation of a whole new ecosystem to combat the danger. Most AI models from OpenAI (in which Microsoft has invested) and Google have built-in software to identify deep fakes and other malign behaviour.
Like every revolution, AI comes with both risks and rewards.
