The AI Acceleration Question
Global progress in Artificial Intelligence is showing no signs of stopping anytime soon. At the same time, governments are scrambling to regulate AI, but with different approaches, threatening to potentially throttle its growth. What approach should India take?
Does regulation hinder innovation?
The most radical changes in any sector are a result of new players innovating to compete with the established players. In the transportation sector, for example, Uber innovated by offering ride services through mobile apps, competing with giant traditional taxi companies. Similarly, the retail sector saw a breakthrough when companies like Amazon and Alibaba tried to compete with brick-and-mortar stores.
More regulations are nearly always detrimental to the new players, because not only do they raise the cost of development, but also discourage experimentation.
The nature of political lobbying by big companies means that these regulations may also reflect the influence of established companies seeking to protect their market dominance from new entrants. The best example of this would be the USA, which arguably has the strongest political lobbying in the world.
As states move on AI, tech lobbyists are swarming in — POLITICO
Which three countries are doing the best in AI? Short answer- USA, China and UK.
The UK Leads Europe and Ranks Third Globally in Artificial Intelligence | Business Wire
Interestingly, these three countries have completely different approaches to AI regulation, providing India three templates to consider.
Is the American template worth looking at for India? Not really. America’s dominance in AI is built upon its traditional advantages including the availability of a diverse talent pool, a culture of innovation and the availability of funds- all of which India lacks. Established tech giants like Alphabet’s DeepMind and Microsoft-backed OpenAI have contributed (and lobbied) as much as startups have. However, the very-recent US President’s Executive Order of 30 October 2023 sets new standards for safety, privacy and civil rights, and is something India should consider looking at. In general, however, it could be argued that America’s global leadership in AI is not because of its AI laws but despite them.
China’s model, however, bears important lessons. The Chinese government has been supporting AI nationally since as early as 2016, and aims to do so till at least 2030. Moreover, by incentivizing research and development, China leads the world in AI patents. China has preferred to regulate only those sectors that it considers “of strategic national importance”.
The UK has a different approach, which it calls “pro-innovation” and “light touch”- basically creating an environment with minimum regulations to promote growth of AI instead of throttling it with restrictions. This involves performing the bare minimum checks on companies’ AI development to ensure that nothing is detrimental to the public. The result? In most AI metrics, the UK stands next only to the USA and China, despite lacking the strong technological base they (and other countries) have had.
India’s AI Acceleration Question — the right way to proceed
It’d be foolish to suggest that India follow any one of these approaches without looking at where each country stood at the time it started enacting these regulations. China, for example, has been brimming with cash, fueled by its massive export surplus and ownership of hard assets; so, financing for government investment into AI is easy. Similarly, the USA is a natural ground for AI companies to thrive, with leading global tech giants also investing heavily; so, a more regulatory approach focused on ensuring that the rapid growth of AI does not come at the cost of safety and civil rights makes sense.
Hence, the Indian answer to the AI Acceleration Question should be grounded in India’s own reality. Firstly, just like China, it could identify sectors of national importance and only place regulations on those, letting companies freely experiment in the others. Secondly, the nature of regulations in these sectors should be like that of the UK, involving a “regulatory sandbox”- a framework that allows companies to freely test new technologies, albeit in a controlled environment.
More regulations of the kind the US has should be reeled in gradually only once India gains significant advancement.
What does the future hold?
India holds immense AI potential. With a huge amount of data, a fast-growing economy and a plethora of tech startups springing up, the future of AI in India is bright if the right regulatory steps are taken.
The view that AI regulations hinder its growth is only partially true. In fact, there are examples where specific AI regulations have even fostered innovation. This has happened in China, which has rules that mandate how algorithms are operated, helping remove uncertainty for companies around what’s ethical and what’s not, providing a clear framework to operate within.
The right AI regulations also help build public trust in AI, by not only assuring people of its legitimate use by companies, but also by bringing AI closer to the people through its use in government schemes. Moreover, some large-scale applications of AI require government involvement, such as in healthcare, agriculture and education. This is being done in India on a large scale.
Finally, the AI Acceleration Question is not about whether AI should be regulated. Instead, it’s a question of how much and where. The answer lies in staying true to principle, asking two questions every time:
- Does the AI progress in consideration create enough value to justify the potential harms it could have?
- What way of mitigating risks due to AI creates the least hindrance for future innovation?
