- Published on
Is AI Actually Going to Kill Us All?
- Authors
- Name
- DeepHorus
A former Anthropic researcher, Jacob Coxon, has resigned and publicly warned that artificial intelligence could wipe out humanity within a decade. His viral thread on X claimed leading AI companies are behaving irresponsibly in their race to advance models. The warning has reignited debate about whether AI doomsday fears are overblown or overdue.
Coxon's resignation came this week, and his comments quickly spread across social media. He argued that OpenAI and Anthropic, the two most prominent AI firms, are like "runaway trains" in their pursuit of more powerful systems. WIRED's Uncanny Valley podcast discussed the claims with reporter Will Knight.
Coxon said "most AI researchers earnestly believe there's a non-zero chance AI will kill us all." He described the moment as "crunch time for humanity." The warning centers on the possibility of extinction within the next ten years if development continues unchecked.
The debate matters because AI safety is no longer a fringe concern. Major companies are deploying increasingly capable models with limited oversight. If researchers inside leading labs are quitting over safety fears, regulators and the public may need to pay closer attention.
On Uncanny Valley, hosts Brian Barrett and Leah Feiger also discussed Apple's debut of the iPhone Duo, a foldable phone priced around $2,000. They noted new "always listening" Apple Watch features that made Barrett uneasy. The episode also covered a WIRED investigation into a Census Bureau report on noncitizen voting that used faulty data.
The AI safety conversation is likely to intensify as more insiders speak out. Whether Coxon's prediction is accurate or not, his resignation signals growing unease within the industry. The coming years will show whether companies like Anthropic and OpenAI adjust their approach or continue accelerating.
