
AI Risks and Digital Hygiene

Stop Telling Yourself You're Bad at “People Stuff” - Martijn Versteeg
Topic 10 of 12
Published Feb 23, 2026 · Clip 53:19–1:02:15 (8:56)
Martijn argues we'll look back on this era as the time before 'digital hygiene,' drawing parallels to industrialism polluting rivers — except now the river is the mind. He cites Healthy Gamer research suggesting LLM use can induce psychosis even in healthy adults, warns about anthropomorphizing bonded text conversations, and describes how agreeable AI feedback loops may be raising neuroticism and lowering conscientiousness at population level. He isn't anti-AI — recommender models and AI in astronomy are valuable — but worries about adopting LLMs without guardrails, sketching an 'AI Manifesto' he'll present at a conference.