The battle is between the proprietary rights represented by American tech firms, which keep their technology guarded and protected, and the Chinese who follow open-source AI
Artificial Intelligence (AI) is now a hot subject as it highlights the growing race between the United States (US) and China in the frontier area of technology that stretches the cerebral capacity of humanity. This also reeks of and is reminiscent of the bygone bitter arms race between the erstwhile Soviet Union and the US!
Even as AI is touted to surpass the electricity, steam and computing revolutions of the industrial era, its unconscionable use of data and electricity risks suggest that the pathway ahead is replete with portents and possibilities, more the former than the latter, with big tech companies of Silicon Valley neck-deep in it! How AI is going to be developed and deployed is at the core of the clash between the US and China, with other fringe players keeping their fingers crossed that the resulting benefits, if any, would spill over to them. This is so as the battle is between the proprietary rights represented by American tech companies, which keep their technology guarded and protected with guardrails against poaching or purloining to uphold their competitive finesse, while the Chinese have increasingly followed open-source AI.
Compared to earlier general-purpose technologies such as steam engine, electrification and information technology, which heightened workers’ productivity by extending their capabilities with better tools, AI could traverse the extra mile to enhance the production of knowledge itself.
Given the grim reality that AI has given nefarious people a chance to foster credible deepfakes and voter-attraction ways to facilitate resourceful parties in power to play havoc with the outcome of even free and fair elections in democracies, the burgeoning use of open-source AI models only aggravates the restiveness of people disenchanted with optical illusions and delusions determining the outcome of polls being bitterly fought among several political parties.
Be that as it may, experts say that business process outsourcing (BPO), in which India gained supremacy through its software skills in capturing overseas markets as service providers, was made possible by technological progress triggered by cheaper computing, with the price of data transmission touching rock-bottom on the back of cheap bandwidth following the dot-com bubble that left fibre-optic cables in large places lying idle for use in assorted tasks across the world cheaply. Verily, India benefited significantly from the concomitantly inexpensive internet to build the global back office.
As big AI firms spend humongous amounts on chips, data centres and power purchases to keep the fastest mass-computing machinery spick and span, their borrowing binges, co-opting supply chain owners, might suck the entire system into a vortex if the chips are down, figuratively speaking. This is the warning bugle sounded by the Bank for International Settlements (BIS) in its latest Annual Report. The portent of such an eventuality happening in the absence of any breakthrough in ensuring limitless productivity improvement but human obsolescence — what technologists refer to as AI reaching “singularity” — a point at which machines could improve and invent themselves, rendering humans redundant and eliminating labour reinstatement through the creation of new tasks, is made feasible!
One needs to reckon that AI just simulates human intelligence through learning and reasoning by repetitive practice, processes massive mounds of data to pinpoint patterns and gain insights, and persistently improves performance using experience and feedback. It has myriad models, including machine learning, generative AI, Natural Language Processing (NLP), Large Language Models (LLMs) and Expert Systems.
Interestingly, in response to a starred query in the Lok Sabha on July 22 in the ongoing Monsoon session, the Union Minister of Electronics and Information Technology, Ashwini Vaishnaw, proclaimed that the India AI Mission builds on its vibrant information technology (IT) sector, with annual revenue of $300 billion and a robust workforce of six million. He noted with verve that the India AI Mission has seven pillars, including the development of AI models, applications, datasets and talent, and affordable computing through empanelled cloud service providers.
It is also particularly pertinent that the Government of India is supporting Indian start-ups and institutions to develop large language models, under which 20 indigenous sovereign models have been identified for support, comprising a dozen LLMs and eight small language models to address healthcare and language technologies, with agentic AI systems at various stages of development and deployment.
It is not clear whether the government would help successful IT companies by handholding them to branch out and become AI techies before long, for, after all, they were the original trailblazers to make a momentous foray into Silicon Valley by sheer software strength. This is crucial because today the cost of running an AI model eventually boils down to its physical ingredients, ranging from semiconductors and cooling to electricity, with the whopping outlays invested in data centres. In a country where the voltage system is volatile, making adequate and steady access to reliable power supply the decisive factor is crucial if the intention is to emerge as an AI power in the near future.
(G Srinivasan is a senior economic journalist based in New Delhi)
