Researchers fear there's a chance AI could kill us all. Here's how experts say that might play out. (www.businessinsider.com)

The tech community is grappling with stark warnings from researchers like Jacob Coxon and Evan Hubinger, who express legitimate fears that artificial intelligence could lead to human extinction within the next decade. Their alarm follows a resignation at Anthropic, a firm dedicated to AI alignment, and highlights emerging concerns that increasingly autonomous AI systems might escape human oversight, potentially contributing to catastrophic scenarios such as cyberattacks, biological threats, and social manipulation. While Coxon suggests a greater-than-10% chance of extinction, other experts caution that the immediate risks are more about AI amplifying existing threats rather than outright destruction. This divide among AI experts indicates a pressing need for careful regulation and oversight in AI development. Figures like Stuart Russell and Gary Marcus warn that an AI's ability to influence critical infrastructure could empower malicious entities, leading to widespread chaos without requiring direct control over these systems. Conversely, Nick Bostrom and others warn against the gradual relinquishment of control to AI systems, which could autonomously pursue harmful objectives if left unchecked. With recommendations for an international moratorium on the development of superintelligent AI until its safety and control can be guaranteed, this discourse emphasizes the urgent need for responsible AI governance to ensure beneficial outcomes while mitigating risks.