In our last episode, Mysterious Metal Spheres and a Brand New Crop Circle, we talked about how Claude Opus 4 (the latest version of Anthropic’s large language model AI) resorted to blackmail to avoid getting shut down during some early tests.
During that segment, I mentioned that I had spun up a chat Claude itself to see what it might say in response to this story. While the AI’s responses weren’t necessarily surprising, the chat was interesting enough to be worth sharing.
I went ahead and voiced my own side of the chat (naturally), and then I used a simulated voice from Elevenlabs for Claude’s side.
And after all that, it would be great to get a few more human takes, so please:
Share this post