Home
Ai
Ai Features
AI That Solves for India Can Solve for the World: Google DeepMind's Manish Gupta

AI That Solves for India Can Solve for the World: Google DeepMind's Manish Gupta

Dr. Manish Gupta, Senior Director, Research, Google DeepMind, spoke with Gadgets 360 on AI, focus on India, and DeepMind's biggest milestones.

Written by Akash Dutta, Edited by Ketan Pratap | Updated: 8 April 2026 12:10 IST

Photo Credit: Reuters

Google DeepMind’s AI model that detects diabetic retinopathy was first tested in Madurai

Click Here to Add Gadgets360 As A Trusted Source

Highlights

DeepMind’s India team built the Matryoshka Representation Learning method
Dr. Gupta said AlphaFold is one of the biggest achievements of DeepMind
He says, “AI that solves for India can solve for the world”

The consensus is that generative artificial intelligence (AI) exploded into the mainstream with OpenAI's launch of ChatGPT in November 2022. But those who have followed this space closely know that it all began with a paper titled “Attention Is All You Need,” written by a team of eight researchers at Google Brain (now part of DeepMind) in 2017. The paper introduced the transformer architecture, a foundational neural network used in the Generative Pre-trained Transformer (GPT) by OpenAI and the backbone of all the company's frontier AI models ever since.

For Google, AI is not really a new frontier. The tech giant, which has played a pivotal role in shaping the modern Internet, has been working on the technology for decades. Before generative AI, there were predictive algorithms and sentiment analysis, machine learning and neural networks, and systems that challenged the Turing test. In particular, the company's DeepMind division has been developing such technologies since Google acquired it in 2014.

Google DeepMind's contribution to the AI space is undeniable. From groundbreaking models such as AlphaGo, which defeated world champion Lee Sedol in Go (a board game) in 2016, to AlphaFold, which led to Sir Demis Hassabis winning the Nobel Prize in Chemistry in 2024, the division has continuously innovated and developed technologies that solve real-world problems.

Google’s New AI Model Can Create Music Based on Your Image, Text Prompts

The division has also led several major projects in India and continues to work on new initiatives in the country's critical sectors. To understand its inner workings and vision in the AI space, Gadgets 360 recently had an exclusive interaction with Dr. Manish Gupta, Senior Director, Research, Google DeepMind. The conversation delved into DeepMind's journey after its acquisition by Google, its biggest milestones, and the impact of the India team on the global division.

The Inception and Journey of Google DeepMind

DeepMind was founded in 2010 by Hassabis, Shane Legg, and Mustafa Suleyman to combine machine learning and neural networks and develop artificial general intelligence (AGI). It was acquired by Google in 2014, and since then, it has operated as the research arm of the Mountain View-based tech giant. Later, in April 2023, the company merged Google Brain with DeepMind.

“Google DeepMind brought together two of the world's leading AI labs — Google Brain and DeepMind — into a single, focused, interdisciplinary team that is today driving both the next wave of AI-powered research breakthroughs across science and industry, and transformative products at Google,” said Dr. Gupta.

Today, it caters to both frontiers — academia and product. For the former, Google DeepMind has published research on challenges such as global weather forecasting, and for the latter, it has built and improved the foundational Gemini models that power a large number of Google products and platforms.

“Google DeepMind is the research engine, and in many ways, provides the foundational layer to Google's full-stack approach to bringing the benefits and advances of AI to everyone,” he added.

Battle of the Nerds: Godfather of AI, Google DeepMind Chief Argue Over AGI

DeepMind's Biggest Breakthroughs

When you think of the defining moments of a company, usually those closely affiliated with it can point to a single pivotal moment that shaped its course. But given the deep impact DeepMind has had on the AI space, doing so is tricky. When we probed Dr. Gupta about the biggest breakthroughs of the division, he told us about two distinct moments.

The first was in 2016, when a single move on the board game Go defined the future of AI. That year, the division's AI system AlphaGo became the first programme to defeat a world champion at the game, and with a move that initially looked like a mistake.

“Move 37 was highly unconventional, defying thousands of years of human Go theory, but demonstrated original AI creativity. It revealed that AI had evolved from mimicking human expertise to discovering entirely new problem-solving strategies on its own. It was also the ultimate 'proof of concept' for the deep neural networks and advanced search and reinforcement learning techniques that define the current AI era,” the Senior Director said.

While some may think that a computer beating a world champion might not be that pivotal, Dr. Gupta disagrees. “This shift is profound. We are no longer building models that just predict, but systems that can reason. And people are already interacting with these advantages. Our Gemini 3 family of models can not only retrieve information and better understand nuances of every interaction, but also use that information to understand context, synthesise conflicting ideas, and arrive at logical conclusions.”

The second moment is more recent. In 2024, Hassabis and John Jumper received the Nobel Prize in Chemistry for an AI model called AlphaFold. This was the first time in history that a large language model's development led to the highest recognition in academia.

Google’s Latest Weather Forecasting AI Model Is Available to Users

“AlphaFold solved the decades-long protein folding problem in just a few years. It has essentially helped decode the very foundations of organic life on the planet, predicting over 200 million protein structures, nearly all catalogued proteins known to science,” Dr. Gupta explained, adding, “The AlphaFold database is publicly available, and is being used by over 3 million researchers globally to accelerate research on humanity's long-standing issues – from autoimmune diseases, to malaria vaccines, and even cancer. This includes over 180,000 researchers in India.”

Highlighting the impact of the model, he shared that one researcher has been using AlphaFold to make one of the world's most critical crops, soybean, resistant to disease. The research can be a huge step towards ensuring food security.

Google DeepMind's India Impact

While Google DeepMind has proven its merits on the global stage, the question arises of its impact in India. With massive diversity and distinct socioeconomic circumstances, the country comes with its share of unique challenges and opportunities. The recently concluded India AI Impact Summit highlighted the huge potential of the market and the need to solve local, real-world problems.

“We've seen the Indian ecosystem excitedly embrace the promise of AI to solve for the country's pressing challenges: from using our AI models to screen for diseases like tuberculosis, lung cancer and breast cancer, to leveraging Project Vaani's open-source speech datasets to build automatic speech recognition models that are advancing Inclusive AI,” he said.

But that's not all. The Senior Director believes that by training AI models on the diverse languages and cultures in India, DeepMind can create a playbook for the rest of the world. To that end, the division has been building local AI infrastructure and forging ecosystem partnerships to enable it to scale on-ground impact. Citing one example, he says that the MedGemma models today are supporting AIIMS in building India's Health Foundation Models, with the ultimate goal of strengthening healthcare delivery and patient outcomes across the country.

“We've also seen that AI that solves for India can solve for the world,” Said Dr. Gupta.

He added, “Our model for detecting diabetic retinopathy, first tested in Madurai in India, has supported more than 600,000 screenings worldwide, and continues to support millions of screenings in India and Thailand. Our foundational agri models, first built to strengthen India's agricultural resilience, have recently been expanded to trusted testers in Malaysia, Vietnam, Indonesia, and Japan.”

On Balancing Innovation With Safety

Being a nascent space, innovation in AI is aplenty. However, innovation can sometimes also lead to security and safety risks. From prompt injections affecting AI agents to the problems associated with hallucination, there is a need to look at the downsides of the technology, and the Senior Director agrees.

“At Google DeepMind, we believe this progress must be earned responsibly. Safety, ethics, and scientific rigour are what make innovation durable and worthy of trust. That is why our work is grounded in a deep commitment to responsible development, with careful evaluation, robust safety frameworks, and ongoing collaboration across industry, academia, and government,” he said.

“When innovation is balanced with ethical considerations and safety, AI can become one of the most powerful forces for positive change, enabling humans to focus on creativity, connection, and problem-solving at scale,” Dr. Gupta added.

Comments

Get your daily dose of tech news, reviews, and insights, in under 80 characters on Gadgets 360 Turbo. Connect with fellow tech lovers on our Forum. Follow us on X, Facebook, WhatsApp, Threads and Google News for instant updates. Catch all the action on our YouTube channel.

Further reading: Google DeepMind, Manish Gupta, India, AI, Artificial Intelligence

Akash Dutta Email Akash Dutta

Akash Dutta is a Chief Sub Editor at Gadgets 360. He is particularly interested in the social impact of technological developments and loves reading about emerging fields such as AI, metaverse, and fediverse. In his free time, he can be seen supporting his favourite football club - Chelsea, watching movies and anime, and sharing passionate opinions on food. More