Humanism Confronts Advances in Artificial Intelligence

The Atlantic's Books Briefing (April 7, 2023) reviews Sarah Bakewell's book Humanly Possible and situates its defense of humanism amid recent artificial-intelligence developments, per the article. The Atlantic reports that the piece raises worries that these advancements could upend the primacy of "the faculties of the independent mind," language the article uses to describe the stakes for intellectual personhood. The article also quotes Eric Schmidt saying, "the reason we're marching toward this technological revolution is it is a material improvement in human intelligence," and includes a reported observation that "Every chatbot is created by ingesting books and content that have been published on the internet by a huge number of people."

The article mentions ChatGPT and GPT-4 as exemplars of generative tools created from large corpora of published content, a factual framing reported by The Atlantic. Industry observers have repeatedly noted the same technical dynamic: large language models are trained on broad web and book corpora, which concentrates cultural artifacts into model outputs. This is an industry-pattern observation, not a claim about any single author's intent.

The Atlantic frames Bakewell's humanist argument as confronting both philosophical and social consequences of model-driven information flows. For practitioners, debates about ownership, attribution, and the cultural provenance of training data remain central to public conversations about model deployment and governance.

Observers should track how cultural critiques influence policy debates over data licensing, attribution standards, and disclosure for generative models, and whether those debates change industry practices around training-data provenance. The article is a cultural and literary critique as reported.