On Tuesday (May 19), as Sundar Pichai presented a slew of AI updates at the annual I/O developer conference in Mountain View, Calif., the company’s president and chief investment officer, Ruth Porat, took the stage in a sizzling New York City to explain the big financial numbers behind AI’s AI push.
Google’s capital expenditures (CapEx) nearly doubled from last year, rising to $180 billion to $190 billion by 2026. The company has committed to spending about 40 percent of that impressive budget on data center construction, with the remaining 60 percent allocated to other AI infrastructure such as chips. Speaking at Fast Company’s Most Innovative Companies Summit during an on-stage interview with editor-in-chief Brendan Vaughan, Porat explained that this high spending is a response to a “platform shift” in the entire industry that Google can’t afford to miss.
“We’ve never seen anything this deep in our lives. You don’t want to be behind the curve,” he said. “It’s an incredible privilege to be alive today, especially if you focus on what you can do with technology to advance science, drive economic growth, improve the delivery of critical social services and make advances in health care, education, cybersecurity. Those superpowers run deep. And to deliver it, we obviously need computing power.”
To protect that capacity, Google has been stockpiling AI chips. While it continues to stock up on Nvidia GPUs, it’s also producing its own in-house Tensor Processing Units (TPUs) to power its Gemini models.
While the current AI boom began publicly with OpenAI’s launch of ChatGPT in late 2022, Google has long been a leader in AI research. Ten years ago, the CEO of Photos announced that Google was “we’re moving from mobile-first to AI-first,” which means “we’re going to invest heavily to lead in AI, and we’re going to have a full-stack approach,” Porat explained Tuesday, noting that vision continues to drive senior leadership decisions. “It’s models, it’s chips, it’s research, and it’s application in all our platforms,” he added.
AI tools widely available today are far from perfect. Examples of large languages are still prone to “missing sight.” Eliminating these errors requires more computing power and intensive training capabilities.
“One area we are very concerned about is bringing everyone to a high standard, especially when the models look funny,” said Porat. “If you wake up in the middle of the night and your child has a fever, and you want to give Tylenol, you better be ready. Google stands for quality, and that was very important to us.”
Porat’s leadership lessons from Wall Street
Porat took over as Google’s president and chief investment officer in September 2023, after serving as CFO for eight years—the longest tenure in the company’s history. Prior to Google, Porat was a chief financial officer at Morgan Stanley.
During the financial crisis in 2008, Porat worked closely with then-Treasury Secretary Hank Paulson, from whom he says he learned valuable leadership lessons.
At the time, he had a business that included banks, insurance companies and asset managers as the global economy collapsed. When Paulson needed an elite advisory team to analyze the impending collapse, he tapped Porat to lead a group of about 40 financial experts.
“When I came to Google later, I was asked, ‘What are the lessons from the financial crisis?’ It struck me as “It’s a strange question coming from a place where things have been going up,” he recalled.
His takeaways from working with Paulson boiled down to three key leadership principles, which he now applies to the fast-moving AI space.
First, “identify your biggest source of vulnerability and protect against it early, because you can’t protect against it when you need it most,” he said. During the financial crisis, that risk was not having money in the bank. In the age of AI, the risk is stagnation. “If a competitor uses [a revenue opportunity] and you’re not, the steep curve makes it harder to find later.”
Second, Porat emphasizes that leaders must have both the will and the financial means to take action. “Most of the time, when you put your team together and you have a desire, you can’t make money to be successful,” he said.
Finally, he highlights the importance of building a team that provides “horizontal vision”—the ability to look at multiple organizational units and patterns and trends. “Technology can help, but in a rapidly changing world, a broader perspective is important,” he explained. “As I often say, give me a horizontal perspective. Then I can connect the dots to the future…These lessons remain relevant to leadership today.”
