Matei Zaharia, co-founder of the $43 billion data company Databricks, just won computing’s most prestigious award and made a bold claim: artificial general intelligence is already here.
Zaharia received the top honor from the Association for Computing Machinery, essentially the Nobel Prize of computer science. But his acceptance speech got attention for a different reason – he believes we’re looking at AI all wrong.
The AGI Debate Gets Personal
Most tech leaders say we’re still years away from AGI – AI that matches human intelligence across all tasks. Zaharia disagrees. He argues that today’s AI systems like GPT-4 and Claude already show human-level reasoning in many areas. The problem isn’t the technology, it’s how we define intelligence.
“We keep moving the goalposts,” Zaharia explained. “First we said AI needed to beat humans at chess. Then it was about understanding language. Then creative writing. AI does all of this now, but we just say it’s not ‘real’ intelligence.”
Zaharia now leads AI research initiatives and believes the focus should shift from building smarter AI to making it more useful for scientific breakthroughs. His team is working on AI systems that can conduct actual research – reading papers, forming hypotheses, and designing experiments.
What This Means for You
If Zaharia is right, we’re not waiting for AGI to arrive – we’re learning to live with it. The next challenge isn’t making AI smarter, but figuring out how to use human-level AI to solve real problems like climate change and disease.
Expected this debate to intensify as more AI leaders pick sides on whether we’ve already crossed the AGI threshold.