A veteran technologist recently shared a personal story to highlight the dangers of spreading fear irresponsibly, especially regarding artificial intelligence and its alleged threat to humanity. Reflecting on a prank from his university days—when he and peers falsely warned humanities students of a computer virus, causing widespread panic—he draws a comparison to current alarmist narratives about AI.
This week, former Anthropic employee Jacob Coxon and Anthropic’s Alignment Science lead Evan Hubinger suggested there is over a 10% chance that AI could lead to human extinction within the next decade. Such claims imply that a newborn child today might not survive to middle school age due to AI-related causes. The technologist critiques these assertions as lacking detailed evidence, noting that Coxon’s cited scenarios, such as hacking critical infrastructure or deploying extinction-level bioweapons, are presented without expert analysis or substantiation.
He emphasizes that while the public cannot be expected to understand all technical details of AI, critical infrastructure, or bioweapons, the responsibility lies with experts to provide clear, evidence-based explanations. Fear, once sown by experts, tends to spread rapidly and can overshadow reasoned debate. This phenomenon echoes Jonathan Swift’s observation that falsehood spreads faster than truth.
Drawing from his expertise in computer engineering, the technologist argues that intelligence alone does not guarantee effective action in the physical world. AI systems operate within human-designed frameworks that include accountability and control. Current technological realities do not support the notion that autonomous AI could cause extinction-level events in the near term.
He calls for skepticism toward extraordinary claims about AI risks and urges that such claims be supported by commensurate evidence. The technologist also stresses the ethical obligation of domain experts to avoid abusing public trust by making alarmist statements without solid grounding.
In closing, he offers a message of reassurance: just as the students in his university prank had no real virus to fear, society today should approach claims of AI-driven extinction with caution and demand rigorous proof before accepting such dire predictions.