That said, India does not have a semiconductor fab to build chips required to power a large language model (LLM) or LLM-powered chatbot. LLMs typically require below 10 nanometre (nm) graphic processing units (GPUs), while India plans to build a 40 nm semiconductor fab, implying that it will have to buy graphics processing units (GPUs) and use cloud services to bear the costs of building an LLM.
Cut to 16 November, a day before Altman was asked to leave OpenAI, Satya Nadella, chairman and chief executive of Microsoft, told a select group of journalists at a roundtable in his Redmond office, that he had thought “very, very deeply about” the instances of countries aspiring to build their own LLMs. Acknowledging that India would have “legitimate reasons why they want to do things that are important”, he said governments would need to strike a balance between national and economic interests, which “is definitely going to be key”.
In fact, to drive home his point, Nadella cited a study by an economics professor of Dartmouth College, Diego Comin, who concluded that countries seeking to take advantage of “real paradigm shifts” such as the Industrial Revolution, need “to get the new input…which is cutting-edge technology, into your country fast”.
“Sometimes, when you think about what’s in your national interest and what’s in your economic interest, it turns out that…you have access to the best technology (which) is probably the most important thing…In other words, don’t get stuck reinventing the wheel. (Instead) use it intensely to create new value-add on top (of it),” Nadella said.
“Ricardo (19th century British economist, David Ricardo, known for theory of ‘comparative advantage’) was right; there is comparative advantage in India…(that) should be expressed on top of it versus (the urge) to start from scratch, because if you start from scratch, then you risk falling behind in a massive way,” he added.
When Nadella was asked what would be his reaction if Microsoft’s relationship ever soured with OpenAI, he said, “Without the Intel-Microsoft partnership, would Wintel ever have happened? There is no way Microsoft would be Microsoft without that defining partnership with Intel. And all these years later, Intel remains a massive partner…Partnering is actually an art that helps grow enterprise value.”
As for OpenAI, “it’s not like, they wouldn’t have been able to build without our systems. Our systems are powering them. We are building on top of them. They are building on top of us. So, there’s a lot of co-dependency…We are not going to compete against partners”.
Following the sacking of Altman by OpenAI, Nadella posted on X, “We have a long-term agreement with OpenAI with full access to everything we need to deliver on our innovation agenda and an exciting product roadmap; and remain committed to our partnership, and to Mira and the team.”
Nadella said while the benefits of the Industrial Revolution “spread only to a few countries…this time around, there is no reason why this technology (AI and Generative AI) can’t spread to every part of the world”.
In March, for instance, Air India became the world’s first airline to successfully deploy a Generative AI virtual agent, Maharaja, powered by Microsoft’s Azure OpenAI service.
Air India claims Maharaja has dealt with over half a million customer queries and handles over 6,000 queries a day in Hindi, English, French, and German.
Microsoft is also working with MakeMyTrip to make travel planning more inclusive and accessible through voice-assisted bookings in Indian languages, besides partnering Tech Mahindra to enable generative-AI-powered enterprise search, and to develop intelligent applications with HCLSoftware.
Nadella also underscored the role that India has played in developing “digital public goods. And one of the digital public goods is language translation–speech to text, text to speech”. He cited the example of Jugalbandi–an AI chatbot jointly developed by Microsoft and AI4Bharat–that runs on WhatsApp, aiding users in multiple languages, translating and streamlining responses, and helping them benefit from government programmes.
Nadella recalled a demo of how a farmer, who wanted to inquire about “some government subsidies”, queried a GPT-4 bot and got the task done. “That is really benefiting a rural Indian farmer–getting real access to government subsidies”, and stimulates the economy. Nadella believes this “is one of the most enlightened things to have happen, which is you’re not waiting for some Indian GPT-4 to emerge before the rural Indian farmer can benefit…Let the rural Indian farmer, who really didn’t have a stake even in the industrial Revolution, have a real stake. That, to me, is magical”.
Thoughts on AGI And what does Nadella think about the divisive views on artificial general intelligence (AGI) or artificial superintelligence (ASI)? On the one hand, chiefs of global companies including Elon Musk and Masayoshi Son, and AI experts including Geoffery Hinton and Yoshua Bengio believe the phenomenal growth of generative AI models indicates that machines will soon think and act like humans, a trend we refer to as AGI or ASI. They fear that if humans are unable to fully understand the workings of unsupervised networks, the sophisticated AI models could automatically evolve into Syknet-like machines that achieve AI Singularity or AGI.
On the other hand, an equally accomplished group of experts including Yann LeCun, Fei-Fei Li, and Andrew Ng, believes AI is nowhere close to becoming sentient. They underscore that AI’s benefits, such as powering smartphones, driverless vehicles, low-cost satellites, chatbots, and providing flood forecasts and warnings, far outweigh its perceived risks.
Nadella, on his part, believes it’s better to focus on how we can use AI to drive economic growth that benefits everybody instead of speculating on AGI or ASI. “But at the same time, we are not completely hand-waving AI safety, but really doing the engineering work and participating in all the forums to be able to say, let’s make sure that the benefits of this technology–not just in the future but even today–are really widespread and the risks, mitigated. As a company, I want us to do the hard work, versus hand-wringing about some distant future or ignoring the present danger.”