Ben Horowitz says fears of an AI-fueled job apocalypse are based on a flawed assumption

• Ben Horowitz pushed back on fears that AI will create mass unemployment.
• He said automation has always erased jobs while creating new work we couldn't foresee.
• Horowitz pointed to agriculture, where most jobs vanished yet new kinds of work emerged over time.

Ben Horowitz doesn't buy the idea that artificial intelligence is about to wipe out work as we know it.

The cofounder of Andreessen Horowitz says many of the loudest warnings about AI-driven mass unemployment rest on a flawed premise: that the future of work is predictable.

History, he said, suggests the opposite.

"I think people are acting as though it's very predictable when it's not at all predictable," Horowitz said in an interview on the "Invest Like The Best" podcast on Tuesday.

"Why are you so sure it's going to happen next? And why are you so sure no jobs are going to be created? I don't think it's nearly as predictable as people are saying," he added.

A divided AI debate

Horowitz's comments come as tech executives, economists, and policymakers remain split over how many jobs AI could eliminate, how quickly those changes might arrive, and whether displacement is inevitable at all.

Some senior AI researchers, including computer scientist Geoffrey Hinton, often called the "Godfather of AI," UC Berkeley professor Stuart Russell, University of Louisville computer science professor Roman Yampolskiy, and some tech leaders, including Anthropic's CEO Dario Amodei, have warned of AI's potential to replace large swaths of jobs.

Others, including OpenAI's CEO Sam Altman and Nvidia CEO Jensen Huang, have said AI is more likely to reshape jobs and create new roles rather than erase work outright.

Horowitz, by contrast, took a longer historical view, framing AI as the latest chapter in a long arc of automation that has repeatedly destroyed jobs while ultimately expanding opportunity.

The most extreme example, he said, is agriculture. In the early US economy, roughly 95% of jobs were tied to farming, he said, adding that almost all of those roles are gone now.

"We've been automating things since the agricultural days," Horowitz said. "Almost all those jobs have been eliminated. And the jobs we have now, the people doing agriculture wouldn't even consider jobs."

The mistake, he said, is assuming today's job categories are a reliable guide to tomorrow's economy. Each major technology shift created entirely new forms of work that were difficult, if not impossible, to imagine beforehand, he said.

"The idea that we could imagine all the jobs that are going to come, sitting here now, that AI is going to enable, I think is low," he said.

Automation's long arc

Horowitz was also skeptical of the urgency behind job-loss predictions, saying that modern AI didn't appear overnight.

Image recognition dates back more than a decade, and large language models have been advancing for years.

"We've had AI going right — ImageNet was what, 2012 — and then natural language stuff was like 2015, and then ChatGPT was 2022," Horowitz said. "Where's all the job destruction? Why hasn't it happened yet?"

That doesn't mean AI won't eliminate certain roles.

Horowitz said he expects jobs centered on processing information for others to face pressure.

But he believes demand for creativity will rise.

"I don't really think that the door is going to close behind you," he said. "I think the opportunities tend to multiply when you open up a new way of doing things."