AI Is Disrupting the Entire Economy, Not Just the Tech Stack – Brett Rochkind, SoftBank Investment Advisers

Artificial intelligence has been a theme at technology conferences for years. At the Jefferies Private Growth Conference in Santa Monica this April however, it was the reality shaping nearly every conversation. Few are better positioned to assess that shift than Brett Rochkind, Managing Partner at SoftBank Investment Advisers, whose firm has placed one of the largest single bets in venture history on the AI era.

Rochkind, attending his seventh consecutive conference, was effusive about the event. “This conference is amazing. This is probably the seventh time I’ve been and it gets better every year,” he said. “You talked about AI three or four years ago, and now it’s actually happening! So it’s quite relevant to what we do at SoftBank.”

Rochkind sat down to share his views on where durable AI businesses are being built, what founders need to prioritize on the path to the public markets, and why he believes the scale of the current technology wave is being systematically underestimated.

Separating Durable Businesses From the Masses

SoftBank’s conviction in the AI infrastructure layer is well established. By the end of 2026, the firm plans to have invested more than $64 billion into OpenAI, making it the company’s third largest shareholder and reflecting its confidence in both OpenAI’s foundational model and ChatGPT’s consumer outreach. But Rochkind is equally focused on the application layer: the vertical software companies taking those models and deploying them in ways that create measurable business outcomes.

He pointed to AlphaSense, applying AI to financial research, and Runway, building AI video tools for filmmaking and marketing, as examples of the category-specific disruption he is backing. In cybersecurity, he sees another wave forming. “Each category of software in its own way will be disrupted,” he said. “And so we’re investing in those AI disruptors.”

At the growth stage, where SoftBank primarily operates, the signal Rochkind looks for is real market leadership: customers who are expanding their usage beyond renewing contracts. He highlighted a portfolio company called Sierra, which applies AI to customer service, as an example of what genuine value creation looks like. Sierra, which replaces human agents with automation, is driving significantly higher resolution rates. “What Brett Taylor and the team at Sierra are doing is really driving a more intimate relationship between their customers and their customers,” he said.

One structural shift he noted is the move beyond seat-based SaaS licensing toward consumption-based models tied directly to business outcomes. In software development, the efficiency gains from AI coding tools are substantial enough that companies are willing to pay in proportion to the value delivered; a dynamic that changes both the economics and the competitive moat of leading AI software businesses.

Advice to Founders

With several of SoftBank’s portfolio companies among the most anticipated IPO candidates in the market, Rochkind is in regular conversation with founders about the path to the public markets. His counsel is deliberately simple. “An IPO is just a capital raising event,” he said. ” Build the best business you can, build the best products you can, address the needs of the customer in the most obsessive way you possibly can… and then the capital will be there.”

On the most common mistakes he sees at the growth stage, Rochkind was direct. Companies that raise significant capital often respond by pursuing multiple directions simultaneously, believing optionality is a strength. He disagrees. “You only have finite capital resources, time and management resources. Directing around where you see the greatest ability to add value to your customers should be your highest priority.”

The IPO Market: A Hallmark Year Ahead

On the broader IPO landscape, Rochkind distinguished clearly between two tiers of company. For the highest-quality businesses (OpenAI, Anthropic, and others of that calibre), the market is open whenever they choose to access it.

For other companies, the picture is more nuanced. Rochkind expects a robust M&A market to run alongside IPO activity, as legacy software companies and private equity-owned platforms look to acquire AI-native businesses to modernise their technology stacks. “They’re going to need to make those acquisitions,” he said. The result is a dual-track environment where the strongest companies have genuine choices (i.e. public markets or strategic buyers), while weaker companies may find M&A the more realistic path.

Beyond the Portfolio

Rochkind closed with a perspective that reflects SoftBank’s macro thesis under Masayoshi Son. The debate about an AI bubble, in his view, misframes the question. Prior waves of technology disruption were largely confined to advertising and media. What is unfolding now is categorically different.

“It’s not just the IT stack and disrupting the IT stack,” Rochkind said. “It’s really global GDP.” With AI, software, robotics, and data centre infrastructure now penetrating healthcare, manufacturing, logistics, financial services, and beyond, the addressable opportunity is measured in tens of trillions of dollars. “Technology is permeating everything in the economy.”