Building large language models (LLMs) is no easy feat. Only a few enterprises worldwide possess the necessary resources. In fact, Microsoft had to construct a supercomputer for OpenAI to train its models.
Recently, OpenAI CEO Sam Altman’s visit sparked a hot debate on whether India should develop its own LLM. Tech Mahindra CEO CP Gurnani hinted at the possibility of having their own LLMs. Similarly, AI4Bharat, an initiative of IIT Madras, is reportedly working on a multilingual ChatGPT alternative, but the details are scarce.
When it comes to building LLMs, India can draw inspiration from the UAE. The Technology Innovation Institute (TII) in Abu Dhabi recently unveiled Falcon, a fully open-source LLM. Falcon boasts three variations (1B, 7B, and 40B) and ranks highly on the Hugging Face OpenLLM Leaderboard.
Lessons from UAE
Established in 2020 and funded by the Abu Dhabi government, TII has made remarkable progress. In just two years, they have created the open source LLM that surpasses Meta’s LLaMA. This achievement is truly remarkable and presents valuable lessons that India can learn from. Not just in Abu Dhabi, but across the Emirates, the governments want to accelerate scientific breakthroughs, develop local talent, attract top global researchers, and promote knowledge-based economic growth, a step taken to reduce their economy’s dependence on oil.
Interestingly, both governments have made a significant commitment towards the development and implementation of AI technologies as part of its national strategy. Both the UAE and the Indian governments have announced their desire to be a leader in AI and released their AI strategy in 2017 and 2018, respectively. However, India seems to be lagging behind when it comes to output.
In 2017, UAE made history by appointing Omar Sultan Al Olama as the world’s first Minister of State for AI. Contrastingly, in India, the government has miserably failed to appoint the right person in various top positions. For example, in January 2022, the government said it will hire someone with over 25 years of experience in the semiconductor industry and more than 10 years of experience at global level to serve as the CEO of the India Semiconductor Mission (ISM). However, currently, “The CEO is a joint secretary of the Ministry of Electronics and IT (MeitY) and has absolutely no semiconductor background. The CTO is a scientist at MeitY and he too had no real exposure to the semiconductor industry throughout his career,” Arun Mampazhy, semiconductor analyst, told AIM.
Moreover, in 2019, the UAE’s AI minister unveiled an ambitious roadmap aiming to position the UAE as a global leader in AI by 2031. This roadmap outlines a vision to leverage AI’s full potential for the country’s development. In contrast, India currently lacks a similar roadmap, although reports in April indicated that the government has established a task force to draft a roadmap for the AI ecosystem.
Not only that, in 2019, the government set up the Mohamed bin Zayed University of Artificial Intelligence (MBZUAI), a institution solely focussed on driving AI education in the country. The UAE’s focus on AI education and skill development can inspire India to enhance its education system to produce a future-ready workforce. Incorporating AI-related curriculum, promoting AI literacy, and offering skill development programmes can help India meet the demands of the AI-driven economy.
Although India has taken steps towards AI development, such as establishing three centers of excellence, the UAE government’s dedication and enthusiasm in driving AI growth is a notable example that India can emulate.
Moonshot approach
Most of the research in this country happens in the AI labs of our educational institutions. However, when compared to the West, most of these are still languishing in the theory stage. In a previous interaction with AIM, Prof Amrutur Bharadwaj, research head and director, ARTPARK, said, “I think it has to be driven by the government and they need to pump in money into the ecosystem for a certain period of time.”
Just as India has successfully executed ambitious missions in nuclear power and space exploration, it now requires a similar approach in AI, Bharadwaj said, “We can embark on ‘moonshot’ mission goals with adequate funding support to change the narrative. We have done so in the space and nuclear sectors – so it is doable.”
The government also needs to ensure active collaborations between academia, industry, and government entities to drive AI research and innovation. India can emulate this approach by promoting public-private partnerships, encouraging knowledge exchange, and facilitating collaboration to accelerate AI development.
The post What Can India Learn from the UAE in AI? appeared first on Analytics India Magazine.
