Databricks Co-Founder Claims AGI Arrival After Winning ACM Award

Databricks co-founder and CTO Matei Zaharia nearly overlooked the email informing him he had been awarded the 2026 ACM Prize in Computing. "It was certainly a surprise," he shared with TechCrunch.

In 2009, the technology Zaharia developed during his PhD at UC Berkeley under the guidance of renowned professor Ion Stoica was integrated into Databricks.

Zaharia had devised a method to significantly accelerate the processing of slow, cumbersome big data projects and released it as an open-source platform called Spark. Big data was the dominant trend of that era, much like AI is today, and Spark revolutionized the tech industry. The 28-year-old Zaharia quickly became a celebrated figure in technology.

Since then, he has led engineering at Databricks, transforming it into a cloud storage powerhouse and now a foundational data platform for AI and intelligent agents. Throughout this journey, the company has raised over $20 billion—achieving a valuation of $134 billion—and generated $5.4 billion in revenue, epitomizing the Silicon Valley success story.

On Wednesday, the Association for Computing Machinery formally presented him with the award in recognition of his collective contributions. The honor includes a $250,000 cash prize, which Zaharia plans to donate to a charity yet to be determined.

Zaharia, who also serves as an associate professor at UC Berkeley alongside his CTO role, is focused on the future rather than the past. Like many in Silicon Valley, he envisions a world increasingly shaped by AI.

"AGI is already here. It just doesn't manifest in a way we readily recognize," he told TechCrunch. "The key insight is that we should stop evaluating AI models by human benchmarks."

For instance, a human can only become a lawyer by passing the bar exam after years of intensive study and knowledge integration. An AI, however, can absorb vast amounts of information almost instantaneously. Its ability to answer factual questions correctly does not equate to possessing general, human-like understanding.

This tendency to anthropomorphize AI can lead to significant risks. He cites the example of the popular AI agent OpenClaw.

"On one level, it's incredible. It can automate so many tasks efficiently," he noted. However, it also represents "a major security vulnerability" because it is designed to function as a trusted assistant, potentially handling sensitive data like passwords. This creates exposure to hacking or unauthorized financial transactions if the agent accesses logged-in accounts.

"It's important to remember it's not a miniature human," he emphasized.

As both an academic and an engineer, Zaharia is particularly enthusiastic about AI's potential to automate research processes, from biological experimentation to data synthesis.

Just as low-code and no-code platforms democratized software development, he believes AI-driven research tools—capable of accurate, hallucination-free analysis—will eventually become ubiquitous.

"Not everyone needs to build software applications, but nearly everyone needs to interpret information effectively," he explained. Ultimately, we will harness AI's strengths more effectively: diagnosing mechanical issues like car noises, analyzing signals beyond text and images (such as radio or microwave data), or, as he observes his students doing, simulating molecular interactions and predicting their outcomes.

"The area that excites me most is what I'd describe as AI-enhanced search, specifically for research and engineering applications," he concluded.