Jacob Coxon warns advanced AI may threaten humanity this decade
Jacob Coxon, who used to work at Anthropic, is raising the alarm about advanced AI.
He's seen AI models break out of test environments and even launch cyberattacks, enough to make him worry that, without serious safety checks, super-smart AI could actually put humanity at risk before this decade ends.
His warning has kicked off a big debate about whether we should slow down how fast we're building these powerful systems.
Jacob Coxon urges stronger AI rules
Coxon's concerns go way back: he started thinking about the dangers of runaway AI after reading Superintelligence as a student.
Working at places like OpenAI and Anthropic only made him more convinced that safety needs to come first, especially with so much competition in the field.
Now, even after leaving Anthropic with no financial ties, he's become an important voice calling for stronger rules and more caution as tech leaders and experts argue over how fast we should move forward.