Google Cloud Chief Says ‘We’re Still Living’ After Many ‘Told Us We Were Dead In 2023’

• Google’s strength lies in its "full-stack" integration, from TPUs to Gemini models, and in strong insights into enterprise data, Kurian said.
• Google Cloud is no longer just infrastructure, but a platform for agents that automate process flows.
• Google unveiled its 8th-gen TPUs on Wednesday. 

Google Cloud CEO Thomas Kurian believes the search giant's competitive edge lies in its "full-stack" integration — spanning custom silicon, AI models, and enterprise data infrastructure — following the launch of the company's latest generation of AI chips.

In an interview with Stratechery, Kurian, an Oracle veteran of over 20 years who joined Google in 2019, outlined the industry shift from simple AI chatbots to agents capable of executing complex business processes. He pointed specifically to the combination of Google's Tensor Processing Units (TPUs), its Gemini models, and a new "Knowledge Catalog" that maps enterprise data as the pillars of that integrated approach.

At Google Cloud Next 2026 earlier this week, Alphabet announced its eighth-generation TPUs — the TPU 8t and TPU 8i — with general availability expected later this year.

Google has emerged as a key player in AI, thanks to its widely praised and benchmark-setting Gemini 3 AI model unveiled last November. Google’s cloud unit revenue increased 36% last year, its strongest pace in four years and more than larger rivals Amazon Web Services and Microsoft Azure platform.
 

“A lot of people told us we were dead in 2023 — we’re still living,” Kurian said.

Despite some pressure last month, GOOGL stock is up 130% from its 52-week low. 

Google’s Enterprise AI Play

“The framing this year is that as AI models have become more sophisticated, we see customers evolving the use of AI models from being used to answer questions in a chatbot-like fashion, to actually automating tasks on their behalf," Kurian said.

He said Google Cloud is no longer just infrastructure, but a platform for agents that automate process flows. Agents require "long-running memory" and the ability to reason across many steps, which Gemini is now optimized for.

To make AI even more sophisticated and reduce hallucinations, especially in enterprise content, Google is launching a Knowledge Catalog. It's essentially a "business graph" that Gemini uses to automatically map a company’s disparate documents, APIs, and databases so agents know exactly where to find specific information.

Highlighting Google's chip strength, Kurian said Google is increasingly selling TPUs to third parties, including rival AI labs like Anthropic (which also uses Amazon.com Inc.’s Trainium chips) and financial services firms.

“We’re unique in three different ways... we have the whole stack of AI technology. In order to do agents well, you need to have a model that takes all these journeys and puts it into the harness that handles the improvement... and having awesome infrastructure, both classical... and TPUs gives us real strength there."

What Wall Street Expects From GOOGL Earnings

Google Cloud’s performance would be among the key things investors will watch out for when the search giant reports its quarterly results on April 29. Analysts expect revenue to increase 18.5% to $106.9 billion and adjusted profit to decline 6.5% to $2.63 per share, according to estimates from Koyfin.

Google shares are up 8.4% year-to-date. On Stocktwits, the retail sentiment for GOOGL declined this week and was ‘bearish’ on Friday.

For updates and corrections, email newsroom[at]stocktwits[dot]com.

Read Next: Palo Alto CEO Calls AI-Led Software Selloff ‘Paranoid' — Sees Big Buying Window In Cybersecurity And Infra Stocks