AI Kill Us by the End of This Decade

That is not a line from a science-fiction movie. It was a warning made publicly by Jacob Coxon, a 27‑year‑old AI researcher just told CNN that:

Artificial intelligence could kill all humans before 2030.

He worked on core model training at both Open AI and Anthropic. He didn’t leak the code. He didn’t share secrets. He simply said: we are racing toward self‑improving superintelligence and gambling with our lives.

What did Jacob Coxon Actually Said?

Coxon told CNN that he does not believe current AI models pose an extinction risk. His concern is about what could happen as AI systems become significantly more capable and autonomous.

That distinction matters. Today, AI can already write code, analyze information, generate content, solve complex problems, and work with different digital tools. But humans are still directing these systems.

1. We decide what they should do.
2. We give them access.
3. We set goals.
4. We can shut them down.

The concern begins when those boundaries start changing.

Imagine an AI system that is not only good at writing code, but also good at AI research itself. Now imagine that system can help researchers design a more capable AI.

Then that newer AI helps design an even better one. That is the basic idea behind recursive self-improvement. And that is part of the debate that deserves our attention.

He Is Not Alone: A Growing Insider Consensus

This is not one person’s fear. Other top voices are sounding the same alarm:

1. Evan Hubinger, Anthropic’s safety lead, publicly backed Jacob on X. He said they “earnestly believe AI could kill all humans” and put the odds at more than 10% within the next decade.

1. Dario Amodei, Anthropic’s CEO, told CNN he “agrees with Jacob much more than he disagrees,” while framing the risk as coming from the wider AI industry race.
2. Geoffrey Hinton, often called the “godfather of AI,” has also said a 10%+ extinction risk from advanced AI is not unreasonable.

The Choice Is Still Ours

The future is not fixed. The next decade will be shaped by the choices we make today about speed, safety, and who we decide to be leaders.

Jacob Coxon’s warning is loud. The question now is:

Will we listen, learn, and act wisely?

If you lead a team, run a business, or shape policy, start the conversation today. Ask your AI vendors for hard questions. Review your automation plans. Put humans back in the loop where it matters.