Gary Marcus calls for boycott of unaligned generative AI

gary marcus, a well-known ai researcher, has made a public call for a boycott of generative ai models, citing the risks associated with their development and deployment. marcus highlights statements from researchers at anthropic and openai, including evan hubinger, who expressed concerns about the potential dangers of advanced ai, with hubinger stating that there is a greater than 10% chance of ai killing all humans within the next decade. marcus argues that the industry lacks a plan to address the alignment problem and mitigate the risks of generative ai. he advocates for a pause in building and deploying ai that cannot be trusted or aligned with human values, suggesting that classical, deterministic ai has been reliable and safe. marcus does not call for a permanent ban on ai, but rather a temporary halt until the industry can ensure the reliability and trustworthiness of ai systems.