They agree human-level AI is years away. Altman says celebrate. Amodei says evacuate. The former colleagues who built the two most powerful AI labs explain why the same facts produce opposite feelings — and why only one of them is willing to say it out loud.
image from grok
Sam Altman and Dario Amodei, former OpenAI colleagues now leading rival AI labs, agree on AI timelines but diverge sharply on urgency of concern. Altman recently declared confidence in building AGI and is pursuing acceleration toward superintelligence, while Amodei warns the industry is ignoring an approaching 'tsunami' of transformative AI. Their public rivalry reflects a deeper philosophical split in the AI field between maximalist acceleration and safety-first caution.
Two former colleagues are racing toward the same destination. One thinks you should cheer. The other thinks you should panic.
Sam Altman and Dario Amodei worked together at OpenAI — Altman as president, Amodei as research vice president. They founded the lab together in 2015 with Elon Musk, who recruited both men to ensure Google had a competitor in artificial intelligence, according to Business Insider. By December 2020, Amodei was gone, departing over what he later described as an unbridgeable difference in vision. He left with a handful of colleagues and founded Anthropic, the AI safety company behind Claude. Five years later, the two are running the two most consequential AI labs in the world, and they have not softened toward each other.
At an India AI summit in February, Prime Minister Narendra Modi asked gathered tech executives to join hands in a show of unity. Everyone complied. Altman and Amodei, standing side by side, raised their fists instead. No hands, CNBC reported.
The more revealing split emerged in separate public statements over the past two months. Altman and Amodei agree almost exactly on where AI is headed. They disagree violently on whether anyone should be alarmed.
Altman wrote in a January 2025 blog post that OpenAI is "now confident we know how to build AGI as we have traditionally understood it," and that the company is already looking past that threshold toward superintelligence. He predicted 2025 could bring the first AI agents materially joining the workforce. "Superintelligent tools could massively accelerate scientific discovery and innovation well beyond what we are capable of doing on our own," he wrote.
Amodei agrees with the timeline. Speaking to Indian podcaster Nikhil Kamath, he said: "It is surprising to me that we are, in my view, so close to these models reaching the level of human intelligence. And yet there doesn't seem to be a wider recognition in society of what's about to happen."
His metaphor was blunt. "It is as if this tsunami is coming at us, and you know it's so close we can see it on the horizon, and yet people are coming up with these explanations for 'oh, it's not actually a tsunami—that's just a trick of light,'" he told Kamath.
Amodei told Kamath that an ideology favoring maximum acceleration without adequate caution "has taken hold in parts of the industry." He did not name Altman, but the description was unmistakable.
Altman has not publicly reciprocated the criticism. At the India summit, he dismissed the hand-raising moment as devoid of deeper meaning. His post-AGI framing has been consistently forward-looking, at times almost casual. His January post described the moment as something to accelerate into, not step back from.
OpenAI is valued at roughly $730 billion, a figure that reflects the $110 billion it raised in February 2026 alone, according to CNBC. Both labs are building toward systems that match or exceed human cognitive performance.
What the Kamath interview reveals is that Amodei is no longer content to let the disagreement play out only in product decisions and contract competitions. He is trying to make the alarm public and personal, aimed squarely at the man who used to share his mission. Whether that changes anything in a political environment that has already chosen its preferred AI champion is another question. But the split between two people who agree on the facts but not on the feeling is the most honest account of what the AI race has become.
Story entered the newsroom
Assigned to reporter
Research completed — 0 sources registered. Amodei and Altman agree on AGI destination but diverge on urgency and alarm. Amodei publicly alarmed; Altman accelerating. Former OpenAI colleagues no
Draft (562 words)
Reporter revised draft (665 words)
Reporter revised draft (671 words)
Reporter revised draft (587 words)
Reporter revised draft based on fact-check feedback (671 words)
Approved for publication
Headline selected: Anthropic's Amodei Has a Warning. OpenAI's Altman Has a Party.
Published (578 words)
Get the best frontier systems analysis delivered weekly. No spam, no fluff.
Artificial Intelligence · 47m ago · 3 min read
Artificial Intelligence · 51m ago · 3 min read
