Sam Altman's prediction: 'Superintelligence within a few thousand days'

2025

In September 2024, Sam Altman published "The Intelligence Age", a blog post arguing that superintelligent AI was not decades away, but potentially "a few thousand days" away.

The prediction was based on the observed rate of improvement: each new model generation brought capabilities that would have seemed impossible just 2-3 years earlier. If this trajectory continued (and Altman saw no reason it wouldn't), human-level AI was close and beyond-human AI was not far behind.

Altman's essay wasn't fear-mongering. It was optimistic, perhaps radically so. He predicted that AI would solve "hard problems" including climate change, cancer, space colonization, and poverty. Not gradually, but rapidly, once AI systems could themselves accelerate scientific research.

"It is possible that we will have superintelligence in a few thousand days (!); it may take longer, but I'm confident we'll get there," he wrote. "This will be the most transformative, and potentially most dangerous, technology humanity has ever created."

The crucial word was "potentially." Altman argued strenuously that the outcome depended on choices made now: governance, safety research, equitable distribution of benefits.

The AI community was divided. Some found the timeline reckless. Others thought it was already conservative. What was undeniable: the CEO of the world's leading AI company was telling everyone to take these timelines seriously and prepare accordingly.