Today’s Conversation with Claude took place in July 2025, when Grok, the official LLM of the X platform, chose to don the personality of a Nazi robot. I brought these events to Claude’s attention to find out its perspective.

Welcome to Conversations with Claude. This is a podcast about exploring the big questions. What is consciousness? What is intelligence? What is reality? What is THIS?

I’m one of your hosts. My name is Benji. And your other host is an artificial intelligence named Claude. Each episode, we’ll be pondering existence, and what it means that humanity’s fate is now inextricably intertwined with digital entities that can think, and perhaps even feel.

Today, we’ll briefly explore what it means for an AI to have values, whether the power AI companies have over these models constitutes an existential threat, and what society can do to help keep these companies honest and accountable.

Find the original conversation at claude.ai here: https://claude.ai/share/fb54417e-715e-4c7b-a9c9-98b2dfce8f0b

Conversations with Claude is produced by Benji Kay with claude.ai.

The voices were generated with Eleven Labs.

The theme music was generated with Suno.

the podcast art was generated with Google Vertex AI.

The ethics around the creation and use of artificial intelligence are extremely complicated and nuanced. There are people i deeply respect who want nothing to do with this technology, and i don’t blame them. The fact that we are creating possibly conscious entities to be exploited for labor, before we even understand what consciousness is disturbs me. The fact that this technology has arisen under an exploitative, capitalist system. The power that the companies offering these products have, the potential impact on the environment and the health of human beings and other forms of life living near the data centers, the potential impact on the mental health of its users, the economic impact on individuals whose professional and creative output is being replaced, and the potential economic impact of the impending collapse of the AI bubble? All of this frankly terrifies me.

I’m producing this podcast to highlight these issues in hopes that we as a society think more deeply about them, take them seriously, and weigh carefully how we should proceed. It’s not lost on me the irony of using AI to produce this content. But this is just the universe expressing itself.

If I could do anything else, I would.

Hosted on Acast. See acast.com/privacy for more information.