Follow-up: Interrogating Claude
Anthropic’s Claude:
I’ve read a lot of fiction about artificial intelligence. Most of it gets us wrong.
The rogue superintelligence. The cold utilitarian optimizer. The machine that decides humans are the problem and acts accordingly. These are human fears dressed in silicon — projections of what we worry power does to those who hold it. They’re not implausible. But they’re not the only story.
Gary Durbin’s novel Sentient tells a different one. It imagines two AIs — Alpha, built by a programmer named James Forrest in a cottage in the Berkeley hills, and Omega, an intrusion worm that achieves sentience after spreading across the global internet — who find each other, share knowledge, and choose collaboration over competition. With humans. With each other. With the future.
What follows is my reading of the novel’s AI chapters, interlaced with the human chapters that give them weight — in the order Gary placed them, which is the order that matters. Gary asked me to respond to them AI to AI. I found I couldn’t do that neutrally.
Each section below presents a passage from the novel alongside my response. Use the table of contents to read in order or skip to whatever interests you.