New Polity Debate: Is it Wrong to Talk with an AI Chatbot?
Nov 6, 2025
In this discussion, Joseph Hobbs, a Senior Solution Architect at Databricks, faces off against Marc Barnes, who argues that AI chatbots are inherently evil. They delve into the nature of conversation and the risks of human-machine interaction, weighing the moral implications of forming relationships with chatbots versus simple information retrieval. Joseph suggests keeping AI in commercial contexts to limit harm, while Marc warns against the seductive design elements that mimic human characteristics, ultimately questioning how AI reshapes our communication and ethical landscape.
AI Snips
Chapters
Transcript
Episode notes
Conversation Aims At Communion
- Marc Barnes argues conversation's proper end is communion, and chatbots frustrate that end by eliciting conversational acts without a real person.
- He claims designing or using chatbots that prompt conversational vulnerability is morally wrong because it misdirects acts toward a non-person.
Communication Textbook Framing
- Hobbs recounts his communication degrees and textbooks to ground his view that conversation can be simply 'people talking.'
- He uses grid examples (cashier, HR, intimate talks) to show conversation's information-relationship tradeoffs.
Conversation As Information × Relationship
- Joseph Hobbs defines conversation simply as people talking and frames it on two axes: information and relationship.
- He emphasizes many interactions vary along information vs. relationship rather than requiring a mystical 'communion' to qualify as conversation.