Q4B: to continue from my last post:
https://lnkd.in/g8vqnFVi
* on what India is doing to beat USA and China in the race for ai supremacy”
But philosophy and conceptual strategy aside let’s look at the science and technology involved.
Granted India lost the first lap of the ai race. But that is because of our omissions not now but ten years ago. You cannot play catch up by trying to mount a rocket which has taken off. You have to mount it before it blasts. Ideally even before it is assembled for takeoff. SpaceX isn’t a trillion dollar stock because it did something last week. It is a two decade journey. It is probably about as old as Bitcoin. And most forget, Bitcoin is an Indian invention. And even today, Bitcoin has a market cap more than SpaceX. Though not as much as Nvidia.
Where India will be in 2030 will be a function of what India did after 2019. If we look at the history of independent India, it has a few stages with well defined separators.
The first milestone was 1947 when India got the right to self determination.
The second milestone was 1997 (fifty years later) when the combined forces of liberalisation, IT services leadership and teledensity increase accelerated the rate of socio economic progress.
The third milestone was 2022 (twenty five years later) when the combined forces of geopolitics, rise of frontier technologies (cyber and crypto) and post Covid challenges / opportunities changed most of the foundational hypothesis that stock market valuations were built on. Now take the three years before and after 2022. What we did in this period will define where we get to in 2032 +/- 3 yrs.
2019 is when the FCRAs of various non profits like the Gates Foundation and their partner entities like PHFI were cancelled. It is also the year which marked the emergence of large domestic non profits like the IIT Alumni Council. For the first time, the act of doing good ceased to be a government monopoly.
So let us track the fundamental changes from. 2019 to 2025. What we do now will probably yield results in 2035 by when the ai sweepstakes would be settled. The core Indian non-profit sector strategy hinges on three core paradigms:
– first is a distributed data centre and not a centralised one. Possibly distributed enough to go into the house or into a desktop or phone. There were zero believers in this and now even Nvidia speaks of home data servers for distributed compute.
– second is a new network architecture for ubiquitous networks. The current data networks emerged from legacy networks. Hierarchical and circuit switched. This architecture doesn’t scale. The new one based on hdvsl scales massively. Starting at 155 gbps type rates.
– third is getting inference to work on remote data and doing training and inference from the same system with a self learning framework that scales to artificial general intelligence.
These approaches are unique to India. Niether China nor USA are following this.Activate to view larger image,