From Search to AI Supremacy: Sundar Pichai on Google's Next Decade
Sundar Pichai, CEO of Google, discusses AI-first strategy, regulatory pressures, and the decade ahead for the tech giant
“AI ends up being the biggest opportunity ahead and so I’m focused on what we can do to innovate as a company and I think if we can do that well, our success is in our hands.”
— Sundar Pichai, CEO of Alphabet and Google
Sundar Pichai recently sat down for a wide-ranging interview with Andrew Ross Sorkin at the New York Times’ DealBook Summit. The conversation spanned critical topics such as AI innovation, competition in the tech industry, regulatory challenges, and the company’s long-term vision. Andrew posed a series of thought-provoking questions, and it was both fascinating and reassuring to hear Sundar’s insights and ambitions for the future of the tech giant.
Sundar emphasized that Google has taken a “deep full-stack approach to AI”, building foundational models and tools that serve over 3 million software engineers worldwide. He stated:
“We do world-class research; we are the most cited institution globally.”
Google’s AI infrastructure, including the 6th generation of Tensor Processing Units (TPUs), was highlighted as a key competitive edge. Tools like Gemini have already been integrated into Search, benefiting over 1 billion users globally.
AI Predictions and Competition
Responding to Satya Nadella’s claim that Google should have been the “default winner” in AI, Sundar pushed back, pointing out that Microsoft’s AI tools rely heavily on OpenAI’s models rather than in-house innovations. “I’d love to do a side-by-side comparison of Microsoft’s models and ours any day” he quipped.
Sundar acknowledged strong competition from OpenAI, Anthropic, Meta, Amazon, Apple, and xAI, but emphasized Google’s data access, infrastructure, and research pedigree as distinct advantages.
Regarding AI’s future, he predicted 2025 will require deeper breakthroughs in algorithms and reasoning abilities, moving beyond scaling compute power:
“In the current generation of LLMs [Large Language Models] a few companies have converged at the top but I think they are all working on the next versions too. I think the progress is going to get harder when I look at ‘25. The low-hanging fruit is gone. The hill is steeper. […]
I'm very confident there'll be a lot of progress in ‘25. I think the models are definitely going to get better at reasoning, completing a sequence of actions more reliably, more agentic if you will. […]
When you start out, quickly scaling up you can throw more compute and you can make a lot of progress but you're definitely going to need deeper breakthroughs as we go to the next stage.”
AI, the “Blue Link Economy”, and Google’s Search Evolution
Andrew raised a provocative question about Google’s position in the AI race. He noted that while Google was the pioneer in AI research, it seemed to have been overtaken, an implicit reference to OpenAI’s breakout success in 2022 with ChatGPT public launch. He pressed Sundar on whether Google’s cautious approach to AI advancements stemmed from a desire to protect its core “blue link economy”, the search-based business model responsible for generating hundreds of billions of dollars in revenue.
Sundar firmly rejected the notion that Google had been slow to innovate, emphasizing that AI has been deeply embedded in Search for over a decade. He pointed out Google’s early use of Deep Neural Networks (DNN) in 2012 for image recognition and speech processing, as well as its development of sequence-to-sequence learning in 2014 to enhance text understanding. By 2015, Google introduced RankBrain, an AI-driven ranking system designed to improve search relevance, a milestone that highlighted the company’s commitment to AI integration long before AI became mainstream. Sundar explained:
“The area where we applied AI the most aggressively in the company was in Search. The gaps in search quality were addressed by Transformers. Internally we call it BERT and MUM, and we made search multimodal […] improving the language understanding of search. That’s why we built Transformers in the company.
He further highlighted Google’s recent advancements with AI Overviews, Gemini embedded in search results which now serve over a billion users globally:
“[…] if you look at the last couple of years, with AI overviews, Gemini is being used by over a billion users in search alone.”
Sundar’s response underscored that AI isn’t a new frontier for Google, it’s been at the core of its evolution for years, driving search innovations and enhancing user experiences long before recent developments captured headlines.
AI, Information Overload, and the Future of Search
During the interview, Andrew cited growing concerns from critics who argue that Google’s core business is under threat as more users turn to AI chatbots for answers. Additionally, the proliferation of AI-generated content online has raised questions about whether the value of traditional search could deteriorate in the face of information overload.
Sundar countered these concerns, asserting that Google Search will become even more indispensable as the digital landscape grows increasingly chaotic. He argued that in a world flooded with content, the need for trustworthy, reliable, and well-organized information will only strengthen Google’s relevance. Sundar explained:
“In a world in which you’re inundated with content, you try to find trustworthy content, content that makes sense and can be relied upon, is more important than ever.”
He acknowledged that the way people consume information has evolved, from Facebook to YouTube, TikTok, and beyond, but dismissed the idea that this shift necessarily diminishes search. Instead, he argued that these platforms operate with a zero-sum mentality, assuming users will abandon one source for another. Sundar challenged this view:
“The problem with a lot of those constructs is that they are zero-sum in their inherent outlook. They assume people are consuming information in a limited way, dividing it up. But that’s not the reality of what people are doing.”
Instead, he emphasized that information is fundamental to human behavior and that search, with its focus on quality and trustworthiness, remains the foundation for navigating today’s complex information ecosystem.
This perspective highlights Google’s confidence that its role as a curator of reliable knowledge will only grow in importance as AI-generated content expands, making search engines essential tools for filtering signal from noise.
Political Dynamics and Regulation of AI
Sundar acknowledged that politics will influence the outcome of regulatory debates. With a new president-elect focused on technology leadership, especially in AI, Google’s role in maintaining U.S. competitiveness against China has come under the spotlight. Sundar said:
“[…] in my conversations with him [referring to President-elect Donald Trump] he’s definitely very focused on American competitiveness particularly in technology including AI and I think there’s a real opportunity in this moment that the constraints for AI could be the infrastructure we have in this country including energy - the rate at which we can build things; there are real areas where I think he is thinking about and committed to making a difference so hopefully we can make progress there.”
While regulatory authorities aim to rein in big tech, policymakers also recognize the global AI race, particularly with China rapidly advancing in AI infrastructure and data analytics.
Sundar emphasized the dual challenge regulators face: promoting innovation to maintain U.S. competitiveness while enforcing guardrails against market dominance and misuse. He argued that existing frameworks, such as FDA approvals in healthcare and compliance rules in finance, already provide a foundation for AI governance. Instead, he called for sector-specific policies that regulate how AI is applied rather than the technology itself:
“I would generally take a very pro-innovation approach; there is a lot of regulation already in place so take healthcare, it’s a very regulated industry, it’s not like you can bring a treatment in without going through all the regulatory approvals so just because you’re using AI doesn't change all of that.”
Conclusion: Google as an AI-First Company
The interview highlighted Sundar Pichai’s vision for Google’s AI-first future, balancing innovation, competition, and regulatory challenges. As the AI landscape evolves, Alphabet is well-positioned to lead the next wave of advancements, not just in Search but across cloud computing, enterprise AI, and autonomous technologies like Waymo.
Sundar’s closing remarks echoed this confidence:
“The last decade set the company to be deeply AI-first. The LLM is one moment in time. We are working on so many things in such a deep fundamental way. Pretty much it’s the same piece of technology [referring to AI] which improves Search, which improves YouTube, which improves Cloud. […]
I still think of us as a AI-first company and I think if you’re are at the forefront of making progress with this technology, bringing it in a bold and responsible way I think we’ll do very well as a company.”
Having previously analyzed Alphabet’s strengths and challenges in my earlier articles, this interview further reinforces my conviction in Alphabet as an holding in my portfolio.
Sundar’s forward-looking vision, coupled with Alphabet’s proven track record in AI and continuous innovation, strengthens my belief that Alphabet has a bright future ahead as a leader in the AI-driven economy.