 | « Back to article | Print this article |
'You don't need massive industry or huge energy resources.'
'You could start with just 3 or 4 people in a coffee shop and invent the world.'
Sebastian Thrun, the pioneer in artificial intelligence, robotics, and self-driving cars, sees an exciting and huge opportunity for India in AI.
Speaking at the Invest Karnataka 2025 summit, Thrun said India is uniquely suited for AI innovations, without the need for massive infrastructure or resources.
"This is your chance, India. You have the brightest people in the world because most of our CEOs in America are Indians. If you are doing something right, you export to us, like Satya Nadella," said Thrun during a discussion with The Economist's Michelle Hennessy.
Thrun is known for co-founding Google X, the innovation lab responsible for groundbreaking projects like Google Glass and Waymo, Google's self-driving car initiative.
He explained that India has huge potential in AI because it primarily deals with software.
"You don't need massive industry or huge energy resources, and you don't have to be Germany to build this.
"You could start with just three or four people in a dorm or a coffee shop and invent the world," he said.
Thrun cited the success of a group of Chinese novices who, from the confines of a dorm room, built Deepseek -- a low-cost competitor to OpenAI -- without any formal AI background.
"So this is a moment where you can truly democratise and be on par with the best of the best in the world, which you, in many ways, already are -- if you take that attitude to innovate," said Thrun.
"If you take the attitude of Europe to regulate, then good luck."
AI Regulation and Ethics
When asked whether a global framework for AI is necessary or if each country should handle it independently, Thrun recommended that India focus on investing in technology, making it accessible to people, and engaging in a broad dialogue about what AI truly means for society.
"When you see abuse and technology being misused against the interests of the people, legislators and regulators step in and ask how to prevent this. This works for many technologies, not just AI," he said.
"There are very few exceptions. Let's say nuclear weapons should be regulated differently. But AI is not a nuclear weapon. It is a technology that helps you create content and do research."
He noted that regulators exist to protect society and minimise the misuse of technology, such as cybercrime and deep fakes.
However, this does not mean the entire technology should be deemed illegal; instead, the focus should be on the opportunities it presents.
The AI Race
Regarding the global race in AI and whether other countries are catching up with the US and China, Thrun said, "We're just at the beginning, and the opportunity is open to everyone."
He pointed out that, in the past, there was a belief in the US that building an AI network would cost $100 million, but now it is clear that it costs only around $5 million.
Additionally, most tech companies are focusing on open-sourcing, making the latest models accessible and creating a level playing field.
"It doesn't really matter where the model was made - -whether in Abu Dhabi, China, or France," he said.
"The race is now about turning these models into productivity and something that we care about."
Impact on Jobs
When asked about how AI would impact the job market, Thrun said he firmly believes that technology complements people and that this is not a zero-sum game.
Historically, he noted, society has undergone major innovations, such as the steam engine, which transformed the world from an agricultural to an industrial economy. Despite these shifts, they did not lead to a shortage of jobs.
For instance, in Europe, agriculture once employed 95 per cent of the workforce, but now it accounts for only 2 per cent. However, new jobs have emerged, such as airline pilots, software engineers, massage therapists, and TV anchors.
"In this new world, the first thing you will see is more wealth, greater capabilities, and the democratisation of skills," he said.
Moonshot Ideas
When asked how to assess the potential of moonshot ideas, Thrun said that, for him, it comes down to three key elements:
A significant societal problem to tackle
"Every major company has these kinds of individuals, but they need to be discovered and empowered," he said.
At Google X, the focus was on identifying such people and giving them a clear milestone to aim for.
"The most famous one for us was the self-driving car, where we had a thousand miles carved out in California, and we had to demonstrate to Google founders Larry Page and Sergey Brin that the car could drive 100 per cent autonomously," he said.
AI in Warfare
When asked about the potential of AI destroying humanity, Thrun explained that these systems lack objectives and emotions necessary to carry out such actions.
"To destroy humanity using AI, people would need to be involved in directing it. But my biggest concern is deep fakes," he said.
"We live in a world where we take things for granted, and now AI can fool us."
He also predicted the growing use of swarm drone technology in defence, where multiple drones collaborate to complete tasks.
Reflecting on lessons from the Ukraine-Russia war, he noted that low-cost swarm drones would make it economically unfeasible and challenging for countries to invade others.
"I once sat in a room next to an operator using a $2,000 drone to drop a bomb on a $5 million tank, instantly destroying it," he said.
Advice to Entrepreneurs
His advice for young entrepreneurs pursuing moonshot projects is to "dream big and build something as bold as a self-driving car."
"One of the most important characteristics for success is grit, the willingness to learn, being self-critical, and reflecting on mistakes instead of blaming others -- and being happy about it," he said.
Feature Presentation: Ashish Narsale/Rediff.com
