“AI Will Be Core Infrastructure For Global Economies”

Given the extensive AI developments emerging from the United States and China, what is your perspective on India's role and prominence in this field, especially in view of the comparatively lower visibility of AI initiatives originating from India?  

AI is still an emerging field, with Nvidia’s processors leading the way, and AMD and Intel making notable strides. A significant portion of innovation is also happening within the global open-source community, encompassing frameworks like PyTorch and various language models. This collaborative effort extends beyond borders, including the (possible) involvement of researchers from India. It isn’t just about the US and China. India possesses immense potential across various sectors and it should actively embrace open-source innovations. I am confident that India is working on creating solutions tailored to its unique needs, such as Large Language Models (LLMs) for Indian languages and I would be surprised if there weren’t ongoing initiatives in this regard. The opportunity still remains in the AI space for India. AI infrastructure has been in the spotlight since Nvidia struck deals with Reliance and the Tata Group recently. How pivotal will the AI-related infrastructure be for India from an economic perspective?
AI will be considered as core infrastructure by economies in a few years. Let’s face it, India probably has the most advanced open infrastructure in the world that’s being copied or being leveraged by a lot of different countries. That’s infrastructure. Amazing things are happening in India. Similarly, India is capable of creating things in AI as well. I am hugely optimistic to see what comes out of India next.
 
Lots of companies are addressing the problem of privacy in AI space. VMware has entered the space with 'Private AI' now. How does your approach differ from others?

We are not necessarily looking at it from a competitive angle. This is a problem that has not been solved for enterprises yet and that’s why we chose to tackle this problem. In fact, when we were collaborating with Jensen (Huang, Nvidia founder and CEO), he said: “Hey, this is exactly what VMware should be doing.” And that shows ‘Private AI’ is pretty unique. The scale at which we are tackling the issue makes it unique. Secondly, it is built on platforms that customers already have. Thirdly, VMware is in a position to bring these innovations wherever the customer wants it. Most of the other alternatives take your data to the public cloud – we don’t say that. While you clearly can do that with our solution, you can also choose to have it in your data centre, in your sovereign cloud, or in your edge – wherever it is that you want to.
 
With AI now at the crux of all tech conversations, supplanting topics like the Metaverse and Blockchain from previous years, what leads you to believe that Generative AI will emerge as the dominant technology of this next decade? 

All these technologies have their own adoption curves. And these adoption curves are usually dependent on two things: accessibility and range of use cases. In the case of Generative AI, one of the key reasons it took off so fast is because you do not need anything special to leverage it. If you have a browser, you can start experiencing the technology and getting value from the technology as an individual user. That is not the case with Metaverse. Certainly, that’s not the case with Blockchain. Secondly, the set of use cases relevant for enterprises to solve today’s problem is huge with Generative AI. For example, the technology can make developers 30-40 per cent more productive. 

Cloud will clearly be facilitating the proliferation of most technologies including Generative AI. What will VMware’s strategy be post the Broadcom acquisition?

VMware’s definition of multi-cloud extends beyond public clouds. It includes private clouds, edge and more. Hock Tan (Broadcom president and CEO) is very excited with that definition and the possibilities. He sees a lot of untapped opportunity in helping customers with their existing data centres and that’s not something that we ever disagree with as a standalone company as well. There are still a lot of customers that want to do a lot more in this space. He says: “Let’s capture that low hanging fruit, even as we continue to build out our multi-cloud portfolio.” That’s the plan.