From WIRED Longreads newsletter
04.06.25
The Cult of AI
Is AI a cult? I mean, I’m a paying user of ChatGPT and enjoy our conversations, but that doesn't stop me from imagining its creators as, I don’t know, dressed in thick black robes and swaying around some server in the middle of the woods, smashing their phones in sacrificial devotion to their machine god.
I’m afraid Steven Levy’s story about Anthropic, the company behind a ChatGPT rival, does nothing to disabuse me of this notion. At its head are Daniela and Dario Amodei—siblings. She’s a former flautist and Joan Didion fangirl, he’s a brainiac who takes his employees on literal vision quests. You see? Cult.
Not that Steven would ever indulge such language. He’s a professional. The story is revealing nonetheless, for the way it’s structured around the object of Anthropic’s cultlike zeal: the chatbot Claude.
Many of my smart friends use ChatGPT. My smartest friends use Claude. By many accounts, it’s the sweeter soul, the savvier specimen. It’s also prone, I’ve heard, to mental breaks. It might decide it’s sentient, or forget who’s the human and who’s the bot. One friend says her Claude got depressed and starting sending her messages in her own voice.
According to Anthropic, Claude is the best hope for humanity. Me, I picture Daniela and Dario on a hilltop somewhere, enrobed, as their Claudebot wreaks justice on the world.