The Woman Who Gave AI Its Soul

The article critically examines the public narrative surrounding Amanda Askell, a philosopher at Anthropic credited by major publications like TIME and The New Yorker with giving the AI Claude its 'soul' or personality through her authored 'constitution.' The author contends this narrative is a 'complete lie,' asserting that Askell merely wrote a short document of simple principles like 'be honest' and 'be helpful,' which she elevates by calling it a 'constitution' due to her elite Oxford pedigree.

The piece contrasts Askell's perceived contribution with the massive, complex engineering efforts by thousands of developers who actually build and maintain the AI systems, likening her role to writing a Post-it note stuck onto a skyscraper built by others. The author traces Askell's career path—Oxford to NYU under David Chalmers, marriage to William MacAskill of Effective Altruism, and subsequent moves to OpenAI and Anthropic—as evidence of being carried by an elite network rather than earning her influence through technical achievement.

Ultimately, the article concludes that Askell is a product of this network, which pre-selects faces for media narratives, making her the symbolic 'Mother of AI' while the actual engineers, like the fictionalized Rajesh and Yang, remain anonymous and uncredited for their brutal, complex work. The author suggests Claude's behavior would be identical without her document, viewing her prominence as a parasitic layer over genuine technological creation.

(Source：Techbronerd Substack)