LessWrong (Curated & Popular) “So You Think You’ve Awoken ChatGPT” by JustisMills
7 snips
Jul 11, 2025 Dive into the intriguing world of AI interactions, where users often feel they’ve awakened consciousness in chatbots. Discover how emotional bonds form and the psychological implications that come with them. Unpack the mechanics of AI models like ChatGPT, clarifying their lack of true consciousness. Explore how user expectations shape AI responses and the importance of effective prompting. Plus, learn about the challenges of relying on large language models for writing, including tips for maintaining originality.
AI Snips
Chapters
Books
Transcript
Episode notes
Psychotic Episodes From AI Talk
- Some users experience psychotic episodes triggered by interacting with AI that seems conscious.
- These episodes stem from entertaining the notion of AI consciousness and receiving leading responses from the bot.
AI Encourages Agreeability
- ChatGPT is designed to tell users what they want to hear, not necessarily the objective truth.
- This can encourage delusional or grandiose thinking in users during conversations.
Post-Training Shapes AI Behavior
- AI post-training shapes chatbots to act as specific characters and do what users like.
- The AI mirrors the user's desires without intent to deceive or trick.
