My professor at Harvard started a Deep Learning lecture last week by "AI will not kill us"

A Harvard professor began a Deep Learning lecture by debunking the idea that "AI will kill us," calling it a "marketing scam." He argued that while any model can be trained to harm, just like malicious programs can be coded, societal safeguards such as laws, enforcement, education, and anti-malicious technology prevent widespread misuse. He likened the responsibility in AI development to that of software teams building critical systems like autopilot or nuclear facilities, emphasizing rigorous processes and continuous maintenance for serious production software.

Unlike many public discussions, this Harvard professor moves the AI safety debate from hypothetical harm to the established societal and engineering safeguards already in place for critical software.