"Best Way To Predict Future Is To Invent It": What Entrepreneur Vishal Sikka Told NDTV

Vishal Sikka, founder and CEO of Vianai Systems, discussed the promises and perils of artificial intelligence. In an interview with NDTV's Senior Managing Editor Vishnu Som, the entrepreneur spoke of the need for careful oversight and proactive responsibility in its deployment.

Sikka began by highlighting the fundamental difference between early AI systems and today's generative AI. “Current-generation AI, which is generative AI, is inherently hallucinatory,” he said. Unlike rule-based AI of the 1980s and 1990s, which provided predictable outcomes for applications like credit approvals or medical diagnoses, today's AI relies on deep neural networks that synthesise vast amounts of data.

#NDTVExclusive | Is future AI, crisis or catalyst? Vishal Sikka, founder and CEO, Vianai Systems answers in this exclusive conversation with @vishnundtv pic.twitter.com/0vrPYjhBJJ

— NDTV (@ndtv) February 21, 2026

“These large language models don't actually understand the questions you ask,” Sikka explained. “When you ask, ‘Who was Barack Obama's father?' it doesn't know who Barack Obama is or what a father is. It's generating the next most likely word. In that sense, every response from an LLM is a hallucination.”

Sikka also addressed concerns over AI developing agency. “These models will act as if they are protecting themselves when tested, but they have no idea what testing or self-protection actually means. They are merely imitating patterns observed in training data.” He warned that AI's seemingly empathetic responses could mislead people, making accountability for AI behaviour critical.

On the impact of AI on jobs, Sikka acknowledged the potential for significant disruption. “The best way to predict the future is to invent it. By training people, teaching them how AI can augment their work, and instilling safeguards, we can prevent massive job loss.”

Sikka also touched on ethical concerns around AI in military applications. Autonomous drones and weapon systems, he noted, raise critical questions about accountability and the absence of standardised ethical frameworks. “Giving control of computer or messaging systems to AI agents at hyper speed and scale can lead to catastrophic consequences,” he said.

“We are on the cusp of major medical breakthroughs, from cancer research to genetic insights. AI can perform millions of calculations to identify optimal treatments,” he said.

On India's AI strategy, Sikka stressed the importance of self-reliance. “India is big enough and important enough that we need to develop all layers of the AI stack ourselves, while learning from what others might offer, as long as it aligns with our framework.”

Sikka also dismissed the notion that AI is an exclusive domain for a select few. “It is not that God divined AI so that only a chosen few can dictate how it works. That is complete nonsense,” he said. “It is simply a matter of learning about it, making up our minds, and putting the right guardrails in place. We have done this before.”

He said that humanity is not a passive observer. “It is within us to fix this. We can demand it. The Prime Minister this week has called for AI governance that is powerful, grounded, and aligned with democratic values. AI must act in the interest of people, not pose dangerous risks.”

“One of the things the Prime Minister has articulated is a goal to be 10x on nuclear power by 2047. I think that is a really powerful and right goal,” he said. Sikka emphasized the importance of SMR (Small Modular Reactor) technology for nuclear power and the need for long-term research into fusion energy, positioning India as a leading player in both.

He linked energy strategy to AI advancement, noting that “the current generation of AI is very energy-intensive. We need massive energy access to power these technologies. Beyond self-sufficiency, India has the potential to export energy globally.”