AI’s integration into the enterprise sector in India has matured significantly, becoming a cornerstone for businesses seeking to enhance product efficiency, streamline supply chains, and reduce time-to-market.
Recent research underscores the growing adoption of AI within the Indian enterprise landscape. A study conducted by Morning Consult, commissioned by IBM, found that approximately 59% of enterprise-scale organizations in India actively incorporate AI in their operations. Moreover, the IBM Global AI Adoption Index 2023 reveals that early adopters have not only embraced AI but are also doubling down on their investments, especially in areas crucial for future growth, such as research and development and workforce reskilling.
The AI discussed here encompasses what’s known as classical, traditional, or discriminative AI, in contrast to the emerging wave of generative AI (GenAI) introduced by OpenAI, notably through its development of ChatGPT.
In just two months of its release in December 2022, ChatGPT had garnered more than 100 million users. Today, OpenAI is valued at over $100 billion and has around 180.5 million users.
For perspective, the internet took 10 years to get to 500 million users, while mobile phones took six years. ChatGPT may take only three years to get to 500 million users, which would indicate its arrival as a mainstream technology.
A major reason for this is that while traditional machine-learning, an AI technique, was largely limited to observing and classifying patterns in content with the help of predictive models, GenAI models rely on self-supervised learning (deep-learning, which again is akin to machine learning) to pre-train on humungous amounts of data, and not only analyze data but also create new designs and propose ways to improve upon existing ones.
The result: we have many so-called ‘Big Daddy’ tech AI models today that include large language models (LLMs), large multimodal models (LMMs), and small language models (SLMs)—OpenAI’s GPT-4 and Sora, Google’s Gemini, Meta’s LLaMa, Anthropic’s Claude-3, and Elon Musk’s Grok, to mention a few.
India’s venture into the GenAI domain is both ambitious and fraught with challenges, notably due to its rich linguistic diversity. With over 400 languages, the need for India-specific large language models (LLMs) is paramount. This necessity has become a focal point for local innovators, from tech giants to startups, all aiming to transform conversational AI by shifting from rule-based chatbots to LLM-based models.
However, the journey is complex, marred by high computing costs and a scarcity of comprehensive Indian datasets. Despite these obstacles, there’s a concerted effort within the Indian tech ecosystem to surmount these challenges, harnessing GenAI for both commercial success and public good.
India is home to 113 unicorns across sectors with a combined valuation of more than $350 billion. But it has just two AI unicorns—Fractal Anaytics, and Ola Electric founder Bhavish Aggarwal’s Krutrim, which is developing an India-based LLM and so is technically a GenAI unicorn.
Companies like Krutrim and Tech Mahindra say they are building local LLMs from scratch. Sarvam AI has said it will work with Indian enterprises to co-build domain-specific AI models on their data. It also hopes to use GenAI atop India Stack (Aadhaar, UPI, Account Aggregator, etc.) “specifically for public-good applications”.
Bangalore-based AI and Robotics Technology Park (ARTPARK) and the Indian Institute of Science are partnering with Google India to launch an LLM called Project Vaani. CoRover has BharatGPT, while the Bharat GPT programme (involving seven IITs, two IIITs, and healthcare firm Vizzhy) is building the Hanooman series.
India’s AI ambitions are set against a backdrop of global competition and cooperation. While the US and China lead the AI race in terms of investment and innovation, India is carving out its niche, focusing on areas where it can make a significant impact. The Indian government’s allocation of approximately $1.1 billion towards its AI policy signifies a strong commitment to nurturing an AI ecosystem that can propel the nation forward.
These ambitions are well-timed, given that an estimated 40% of enterprise applications will have incorporated GenAI by the end of 2024, according to Gartner. The main drivers of this adoption in India include the accessibility of AI tools, the need to reduce costs and automate, and the integration of AI into readily available business applications.
AI’s trajectory in India mirrors a global trend towards deeper, more integrated technological innovation. Yet, it also highlights a path uniquely tailored to India’s strengths and challenges.