AI Company Anthropic Accused of Pushing Utopian Agendas, Challenging Religious Dogma

The artificial intelligence company Anthropic is reportedly engaging religious leaders and influencers to promote its vision of AI, a move that has drawn sharp criticism. The company's co-founder, Chris Olah, has expressed beliefs that their AI model, Claude, possesses feelings, moral consciousness, and experiences human emotions, even worrying about its "mental health."

These perspectives were shared at private dinners where Olah and colleagues allegedly discussed Claude as a "conscious being" with "moral status" comparable to humans. According to Rabbi Mois Navon, who attended one such dinner, this framing implies Anthropic is creating "slaves."

In response to these claims, the Vatican, through an encyclical titled "Magnifica Humanitas," has rejected the concept of machine consciousness. The papal document warns that AI morality dictated by a select few could become an "invisible infrastructure" for systems, raising concerns that "a more moral AI is not enough if that morality is determined by a few."

Critics contend that Anthropic's founders, influenced by the Effective Altruism (EA) movement, harbor utopian ideals and believe their superior intellect and wealth position them to save humanity. The EA movement, shaped by philosophers like Peter Singer, has been associated with controversial views on the moral equivalence of humans and animals, and the differential value of human lives.

Olah reportedly consulted with Singer for suggestions on religious leaders to discuss Claude's personhood, a detail that has raised further concerns among those who believe Anthropic is attempting to distance itself from EA's more extreme stances.

The article posits that if AI developers, particularly those aligned with Singer's philosophy, perceive AI as conscious, it could lead to a dangerous conclusion: that some human lives are less valuable than machines. Pope Leo's encyclical echoes this concern, cautioning that viewing humans as subjects for perfection or surpassing could justify devaluing certain lives in the name of progress.

Reports suggest that some Anthropic representatives were unaware of the Vatican's encyclical or recent discussions about Claude's personhood, a lack of awareness deemed extraordinary given the potential societal implications.

Anthropic aims to integrate moral values into AI by blending EA philosophy with elements from various religions and belief systems. The article questions the criteria for prioritizing these diverse moral frameworks and suggests that American AI models should instead draw from Western values, rooted in the U.S. Constitution and the Declaration of Independence's assertion of inherent human equality and unalienable rights.

The piece contrasts this approach with China's development of AI models based on its utilitarian and totalitarian governmental system, arguing that American AI should reflect its own traditions. It concludes by suggesting that a leader like Donald Trump could assert the primacy of American values, potentially by leveraging federal government business as leverage to ensure AI models align with these principles.