Is the AI boom real or a bubble? Nikhil Kamath asks Silicon Valley founders

The artificial intelligence boom has attracted billions of dollars in investments and pushed startup valuations to unprecedented levels. But are these investments justified, or is the industry heading towards a bubble? Nikhil Kamath raised this question during a conversation with technology founders in San Francisco on WTF podcast. The discussion included Akshay Kothari, co-founder of Notion; Azalia Mirhoseini, co-founder of Ricursive Intelligence, and Anna Goldie, its co-founder and CEO; and Vijay Krishnan, co-founder of Turing. They examined AI's economics, its impact on businesses and the opportunities it could create for India.

AI spending soars, but where are the returns?

One of the biggest concerns was the gap between investment and revenue. A participant estimated that committed AI spending could reach around $700 billion this year, compared with approximately $100 billion in enterprise AI revenue.


Krishnan questioned whether major companies outside the AI industry had demonstrated meaningful revenue growth or cost savings from the technology. He acknowledged, however, that software teams can now run at half to 60% of their former size. Goldie added that AI coding tools have allowed Ricursive to hire only a small team of deep experts.

Kothari argued that AI's economic benefits are taking time to spread beyond Silicon Valley. Many people still use AI primarily as a search tool, rather than exploring how it could transform entire business operations. He identified two potential benefits: completing existing work with fewer employees and enabling teams to pursue projects that were previously too expensive or difficult. Today, he said, “the biggest barrier is just imagination.”

Why AI chips and computing power matter

The discussion also highlighted the growing demand for computing infrastructure. Mirhoseini said AI inference demand was increasing rapidly and suggested that the world could need 100,000 times more computing capacity in five years under a conservative estimate. Scaling data centres alone will not meet that demand, she said.

What should India do to benefit from AI?

Founders said that, for India, the panel identified computing infrastructure as a priority. The country also needs to develop AI businesses that can compete globally as the technology reshapes its traditional IT services industry.

The discussion highlighted opportunities in expert data services, AI models for Indian languages and technologies that deliver complex solutions at lower costs. However, startup funding and administrative hurdles remain challenges.

When Kamath asked where an entrepreneur should invest $100 million, Mirhoseini emphasised computing efficiency, while Goldie stressed the importance of building something competitors cannot easily replicate.

Is the AI boom a bubble?

The conversation offered no definitive verdict. AI is already changing how companies develop software and organise work, but questions remain about whether these gains can justify current investments. If adoption accelerates, today's spending could help build a much larger market. If revenue and productivity fail to meet expectations, valuations could come under pressure. For now, the central question is whether AI's economic benefits will grow fast enough to match the enormous sums being invested in it.