
Recently, the tech world has been rocked by a seismic event. The tech-heavy Nasdaq index dropped 3.1%, while Nvidia, a driver of the AI revolution, saw its stock plummet 17%, wiping out a staggering $593 billion of its market value. The catalyst for this upheaval? A relatively small but groundbreaking large language model (LLM) from China called DeepSeek, and its latest iteration, DeepSeek-R1. With a self-proclaimed development cost of under $60 million, DeepSeek has outperformed established giants like ChatGPT, Gemini, Perplexity, and Claude in numerous benchmarks. The proof, as they say, is in the pudding. Users who have experimented with DeepSeek at https://chat.deepseek.com have been astounded by its capabilities. Reddit is abuzz with developers downloading this open-source model onto their smartphones and running it locally.
The dramatic fall in U.S. tech stocks can be attributed to a simple yet profound realisation. If such a powerful AI model can be built so cheaply, the sky-high valuations of tech companies may be grossly inflated. As the rest of the world unravels how DeepSeek achieved this feat, a new AI world order is emerging—one that promises to be more equitable and inclusive. Despite India's geopolitical tensions with China, the rise of DeepSeek is a development worth celebrating. It represents not just a technological breakthrough but a potential catalyst for India's growth, innovation and ambition.
The DeepSeek Phenomenon: A Game-Changer in AI
DeepSeek, founded in May 2023 by Liang Wenfeng, is a Chinese artificial intelligence company based in Hangzhou, Zhejiang. Initially backed by Liang's hedge fund, High-Flyer, DeepSeek has focused on developing open-source large language models (LLMs) that excel in reasoning tasks. The company began making waves globally since January 22nd, 2024 when its researchers released a paper titled "DeepSeek-R1: Incentivizing Reasoning Capability in Large Language Models via Reinforcement Learning." The paper revealed that DeepSeek-R1 was trained for a mere $5.5 million, a fraction of the cost of its competitors. For context, Gemini is estimated to have cost between $30 million and $191 million, while ChatGPT-4's technical creation cost ranges from $41 million to $78 million. These figures don't even account for staff salaries and other operational expenses.
But the real game-changer lies in DeepSeek's operational efficiency. The model charges $0.55 per million input tokens and $2.19 per million output tokens, compared to OpenAI's ChatGPT, which charges $15 and $60, respectively. This dramatic cost reduction is made possible by DeepSeek's innovative architecture. Despite a massive 671 billion parameters, DeepSeek-R1 activates only 37 billion per forward pass, making it significantly more resource-efficient. This means it doesn't have to rely on the expensive, high-powered AI chips currently monopolised by a handful of companies in the West. Moreover, by operating under the permissive MIT license, DeepSeek allows developers to inspect, modify, and use its models freely, even for commercial purposes. This open-source approach effectively breaks down the entry barriers plaguing the AI industry.
The Implications for the Global South
The rise of DeepSeek is not just a technological milestone but a harbinger of a new world order. To understand why, we must recognise a fundamental truth: human civilisation is built on applying intelligence and logic. From the labourer carrying sacks to the CEO steering a multinational corporation, every job relies on some form of intelligence. The advent of artificial intelligence has externalised this intelligence, turning it into a commodity. Those who control AI, therefore, wield immense power—they can dictate the cost of intelligence and, by extension, the cost of nearly everything else in the world.
This dynamic has been particularly detrimental to the Global South, where access to cutting-edge technology is often limited by cost. The high price tags of Western AI models like ChatGPT and Gemini have placed them out of reach for most individuals and businesses in developing nations. This has created a widening gap between the Global North and South, accentuating existing inequalities. DeepSeek, with its dirt-cheap pricing and open-source model, disrupts this status quo. By democratising access to artificial intelligence, it makes it affordable for the very regions that need it most.
Consider this: how many of us in India can afford the $20 monthly subscription fee for ChatGPT or Gemini? DeepSeek changes this equation entirely. By offering an affordable and efficient model, it removes AI from the exclusive domain of the wealthy. This isn't just a victory for individual users but entire nations, as affordable AI can drive innovation, boost productivity, and create new opportunities in every sector of every nation.
A Blueprint for India's AI Ambitions
The rise of DeepSeek also holds valuable lessons for India. If a small Chinese company can build a world-class AI model in a relatively short time and at a fraction of the cost, there is no reason why India cannot do the same. The Indian government should take note of the policies and strategies that have enabled China's AI boom and seek to replicate them. This includes investing in research and development, fostering a culture of innovation, and creating an ecosystem that supports startups and open-source projects.
India's strengths in software development and a nearly endless pool of engineering talent position us well to become a significant AI player. However, this can come to fruition only when the government prioritises AI as a strategic sector to actualise this potential. This can only be done by allocating resources, incentivising private investment, and building the infrastructure to support AI development. It also means learning from China's approach to AI governance, which has balanced innovation with regulation to ensure that AI serves the public good.
The Broader Implications for the World
The rise of DeepSeek isn't just a challenge to Western dominance of AI tech but to the very structure of global power. Historically, technological superiority has been a key determinant of geopolitical influence. By making AI affordable and accessible, DeepSeek and similar models can level the playing field. This is particularly significant for the Global South, which has long been marginalised in the global economy. Affordable AI can empower developing nations to address pressing challenges, from climate change to poverty, and to participate more fully in the global economy.
Moreover, the open-source nature of DeepSeek fosters collaboration and innovation. Building on its model creates a virtuous cycle of improvement and adaptation, a stark contrast to the proprietary models of many Western companies. In this sense, DeepSeek represents not just a technological breakthrough but a philosophical shift—a move towards a more open, inclusive, and equitable AI ecosystem.
A Call for India to Embrace the AI Revolution
Much as it has woken the world, DeepSeek's rise is also a wake-up call for India as it proves that the future of AI need not be dominated by a handful of Western players. This signals India to leverage its strengths to build its own vibrant, inclusive and open AI ecosystem, much like China did. By adding to the democratisation of AI, India can join China in helping transform economies and empowering individuals to eventually reshape the global order.
Hence, we in India and the GoI should not view the rise of Chinese AI as a threat but as an inspiration. By learning from our neighbour's successes and building upon our unique strengths, we can play the AI revolution on the front foot.
The time to act is now. The future of AI is not just about technology; it is about power, equity, and opportunity. In this future, India must play a vital role. Let us cheer the rise of DeepSeek, not just for what it represents, but for the possibilities it unlocks—for India, for the Global South, and for the world.