大多数人认为AI模型的安全漏洞都可能导致严重后果，但作者指出Anthropic发现的所谓'越狱'只能找到'次要'和'相对简单'的软件漏洞，这挑战了政府对模型安全威胁的严重性评估，暗示政府反应过度。

ReconfigBehSci. (2022, March 12). @rwjdingwall @mugecevik @RobFreudenthal 2/2 it’s like comparing how wet you got in a down pour with and without umbrella... The biggest surprise to me in this pandemic hasn’t been the “overreaction” it’s been the constant failure with respect to basic counter-factual reasoning [Tweet]. @SciBeh. https://twitter.com/SciBeh/status/1502681510436130831

Yong, E. (2020, June 4). COVID-19 Can Last for Several Months. The Atlantic. https://www.theatlantic.com/health/archive/2020/06/covid-19-coronavirus-longterm-symptoms-months/612679/