Databricks co-founder and CTO Matei Zaharia nearly missed the e-mail telling him that he was the 2026 recipient of the ACM Prize in Computing. “Yeah, it was a shock,” he informed TechCrunch.
Again in 2009, the tech Zaharia developed for his PhD at UC Berkeley, underneath the tutelage of famed professor Ion Stoica, was launched into Databricks.
Zaharia had created a technique to dramatically pace the outcomes of sluggish, clunky, huge information tasks and launched it as an open supply challenge referred to as Spark. Large information was in these days what AI is right now and Spark turned the tech business on its ear. The 28-year-old Zaharia turned a tech celeb.
Since then, he has helmed the engineering at Databricks, rising it right into a cloud storage large and now a knowledge basis for AI and brokers. Alongside the way in which the corporate has raised over $20 billion — valuing it at $134 billion — and hit $5.4 billion in income. The Silicon Valley dream.
On Wednesday, the Affiliation for Computing Equipment issued him the award for his collective contributions. The award comes with a $250,000 money prize that he’s donating to an as-yet-to-be-determined charity.
Zaharia, who along with his CTO duties can also be an affiliate professor at UC Berkeley, is trying ahead, not again. Like everybody else within the Valley, the long run he sees is full of AI.
“AGI is right here already. It’s simply not in a kind that we respect,” he informed TechCrunch. “I believe the larger level of it’s: we should always cease making an attempt to use human requirements to those AI fashions.”
Techcrunch occasion
San Francisco, CA
|
October 13-15, 2026
An individual, as an illustration, can solely move the bar examination to be a lawyer in the event that they’ve built-in huge quantities of information. However an AI can ingest huge quantities of details simply. If it solutions information questions accurately, that doesn’t equate to common information.
This tendancy to deal with AI like a human can have some profoundly destructive impacts. He provides the instance of the favored AI agent OpenClaw.
“On the one hand, it’s superior. You are able to do so many issues with it. It simply does them routinely,” he mentioned. But it surely’s additionally “a safety nightmare” as a result of its designed to imitate a human assistant that you just belief with issues like passwords. That results in the chance of being hacked, or the agent spending unauthorized cash out of your financial institution as a result of your browser is logged in.
“Yeah, it’s not a bit of human there,” he says.
As a professor and product engineer, Zaharia is most enthusiastic about how AI may also help automate analysis on all the things from biology experiments to information compilation.
Identical to how vibe coding made prototyping and programming accessible to anybody, he thinks that correct, no-hallucinations AI-powered analysis will sometime turn out to be common.
“Not that many individuals must construct purposes, however numerous individuals want to grasp data,” he mentioned. Finally we’ll make AI work higher for us by having it lean into its strengths: telling us what each rattle in our automotive means, or scanning past textual content and pictures to incorporate radio and microwaves, or, what he’s seeing college students do now, simulate molecular-level modifications and predict their effectiveness.
“The factor that I’m most enthusiastic about is what I’d name AI for search, however particularly for analysis or engineering,” he mentioned.
