Pod Save the UK Can AI really be your friend? w/Jamie Bartlett
10 snips
Apr 2, 2026 Jamie Bartlett, technology and AI writer and author of How to Talk to AI, joins to unpack AI risks and why chatbots hallucinate. He describes creating an AI alter-ego to show how easy fake-sounding content is. They debate regulation, public-health approaches to AI use, and where specialized AI could help — all in short, surprising bursts.
AI Snips
Chapters
Books
Transcript
Episode notes
Use Chatbots As Idea Assistants
- Use large language models as idea assistants and for feedback rather than authoritative sources.
- Bartlett says prompting them for suggestions or to improve drafts is a useful, low-risk way to apply the tech.
Give Full Context When Asking Medical Questions
- Be skeptical of medical advice from chatbots and include context when asking (age, history, full symptoms).
- Bartlett warns partial inputs + fluent answers lead users to arrive at doctors misdiagnosed and overconfident.
AI Can Make Complex Language Accessible
- Language models can democratize complex text by translating bureaucratic, legal, or academic language into accessible forms.
- Bartlett notes neurodiverse users and the public benefit from models simplifying contracts, government and health sites.
