Kamal Ghorui
There’ll be no global cooperation in AI. But Indian entrepreneurs can pick their niches
Around this time six years ago, world was waking up to dangers of a brand-new virus. Was Covid a manmade pandemic? We still don’t know. What we do know is that AI, which is in its infancy, already has, or is close to acquiring, knowledge to “create and release” biological weapons. In an essay titled ‘The Adolescence of Technology’, Anthropic CEO Dario Amodei warned last month that AI could guide “someone of average knowledge and ability” to do so. That’s why, as India AI Impact Summit winds down, Modi’s call for AI to be guided by clear human values is significant.
Every global AI summit, starting with Bletchley Park in 2023, has closed with a declaration of guiding values. And safety has been a running theme. India’s latest mantra for AI development is MANAV –Moral, Accountable, National, Accessible and Valid. It’s a laudable goal, but the reality of AI market makes it hard to practise. That reality was on display on the Summit floor Thursday, when Amodei and rival Sam Altman refused to hold hands for the camera. Let’s be clear, AI is evolving amid cut-throat competition. None of the big names has revenues to justify its stratospheric valuation. It’s a winner-takes-all game, so everyone is intent on outrunning everyone else. Cooperation? That’s a three-legged race. Ethics? They require slowing down sometimes, and even changing course.
Forget AI, global cooperation on anything at all – peacekeeping, trade, climate, nuclear arms – seems passé these days. Let’s hope it happens, but assume it won’t as India goes about growing its AI capabilities. That opens two paths for us. One, as the backroom – not in a bad way – of the big players. Our scientists research, develop, test and implement models for foreign firms, gaining know-how. But we don’t compete with the likes of ChatGPT and Gemini, because we don’t have hundreds of billions of dollars to do so. The other path leads to more indigenous foundational models, like Sarvam’s 105B and 30B, tailored for India’s needs, and AI apps based on them as well as foreign LLMs. Like we said yesterday, AI is not one race. Indian firms will need to pick their niche and excel there. It could be AI for the garment units of Tirupur, or the coffee growers of Coorg, or even an early-warning system for pandemics. Govt can enable this AI revolution by improving ease of doing business, and providing funding for critical projects. Indian enterprise can take it from there.
END OF ARTICLE
