Nikhil Kamath’s newly launched podcast, WTF Online is quickly becoming viral on social media, more so because it promises to bring onto the table some of the world’s top voices working in the fields of business, technology, and culture. In this recent episode, Kamath sat down with Aravind Srinivas, CEO of Perplexity AI, to unpack the future of artificial intelligence (AI) and its disruptive impact on industries. Amid debates on India’s AI ambitions, job disruptions, and the rise of digital assistants, one insight stood out—Srinivas’ unexpected bet in the AI race.

When Kamath asked which company he’d back, Srinivas’ answer was surprising: Meta and not Google, or even OpenAI.

The answer behind this lies in the fact that for Srinivas, the defining shift in an AI-powered world isn’t just about better models or more efficient search engines. It’s about human connection. As AI becomes more capable, Srinivas argues, people will seek authenticity, trust, and organic interactions—factors that play directly into Meta’s strengths, given its larger-than-life presence in the world of social media.

“Among the listed players within the ecosystem, I would invest in Meta,” Srinivas said during the podcast. “In a world where AI works increasingly well, the human-to-human connection becomes even more essential. And there's literally no way no one is disrupting that in Instagram or WhatsApp.”

His argument is simple yet profound: Google faces an existential dilemma as AI-generated responses challenge the traditional search engine model. If AI assistants deliver answers directly, the need for search-based advertising—the foundation of Google’s business—diminishes. Meanwhile, Meta’s core strength lies in its ad-driven ecosystem of social connections, which remains intact even as AI evolves. In fact, Srinivas believes AI could supercharge Meta’s business, making its advertising more precise, personal, and indispensable.

This perspective stands in contrast to the growing anxiety around AI’s impact on jobs, a topic that dominated much of Srinivas and Kamath’s conversation.

The Perplexity AI CEO, in fact, painted a stark picture of the short-term labour disruption AI will bring, particularly in the tech sector. “Not as many people are needed to get work done anymore,” he noted, predicting that large-scale layoffs and hiring slowdowns will be an unavoidable reality.

But there was a ray of hope as well.

In this changing landscape, new opportunities will emerge. Srinivas pointed out that India, with its burgeoning startup ecosystem, has a golden opportunity to build homegrown AI infrastructure. “India should definitely train its own models. There’s so much more work to do to make models reason, think, and become truly agentic,” he said, arguing that the country should aim to create its own DeepMind-like research hub.

The discussion also touched on India’s potential to establish itself as a global AI infrastructure leader. With data sovereignty becoming a pressing concern, Srinivas stressed the need for India to develop its own data centres and AI training facilities. “If we can provide training GPUs faster and at cheaper prices, there’s a huge opportunity,” he said, pointing to the growing demand for compute power in the AI age.

Kamath, always keen on practical insights, pressed Srinivas on what Indian entrepreneurs should focus on. The answer? Distribution and scale. Building the next Google or Instagram isn’t just about having better tech—it’s about cracking the cold-start problem and finding ways to reach users at scale.

As the conversation wrapped up, Srinivas returned to the key theme of human connection. In a world where AI influences everything from search to commerce, the platforms that facilitate real, organic interactions will thrive. And in his view, that makes Meta a bet worth making.