With the flag on the south pole of the moon, India continues to progress and show its technological prowess to the world. While that just happened, a new debate seems to be popping up. Just that, it is not about going to Mars or elsewhere - it is about developing a local version of ChatGPT.
And why shouldn't it be? Most of the technological workforce and leadership around the world is still Indian or of Indian origin. Whereas most GenAI action is still limited to the US and China, their ideologies are quite different. The other larger group, the European Union, is still sitting on the fence.
So the question is - should India develop a local version of ChatGPT?
In my view, this is not a question of capability.
Can India develop? Of course, it can.
Should it? Well, that needs to be the question of priority.
A couple of points in favour of development seem more relevant. The first is cultural and linguistic relevance. For a diverse nation with over 1,600 spoken languages across its vast expanse, limiting GenAI to English only is limiting. Catering for various Indian languages, dialects, and cultural nuances can be a game changer. A local version of ChatGPT can address it by tailoring its responses to local languages and cultures. That will make it more relevant and effective for Indian users.
And more importantly, a matter of data sovereignty and privacy is more critical. These are paramount issues in the digital age. A local version will enable greater control over user data. Data generated and processed by the AI model can be kept within the country, reducing concerns about foreign entities being able to access sensitive information. This approach also aligns with India's evolving data protection regulations.
But, as raised earlier, the question of priority is still paramount. Generative AI and other similar technologies are resource-intensive by nature. It requires a substantial investment in talent, infrastructure, and computing power. While load-shedding is still a reality in many parts of India, going for this may seem like a misplaced priorities. Resources like energy may find a better place in addressing other critical national projects or development initiatives.
What about global versatility? Existing AI models like ChatGPT are in some way highly versatile and can cater to users worldwide, including many (but not all) in India. Furthermore, a globally available model can help tap into international markets more easily.
The point is
The question of developing a local version of ChatGPT or any such large-language model in India is very much nuanced. It involves weighing the advantages of cultural relevance, data sovereignty, and the challenges of limited resources.
India doesn't have to participate in a race that doesn't seem relevant to its aspirations. Just because we can doesn't mean we should! Development choices are guided by specific priorities, available resources and long-term vision.
A balanced approach may involve a combination of global and local solutions, leveraging the strengths of existing AI models while addressing the unique needs of the Indian population.
Ultimately, the goal should be to harness AI technologies benefitting Indian society while maintaining a global perspective in the ever-evolving world of artificial intelligence.
Just because we can doesn't mean we should!