LLMs as a Cognitive Virus: New Paper Models AI Dependence Tipping Points
serrjoa · x · 2026-09-07
A new arXiv paper by Ricard Solé, Michael Levin and 7 others models LLM adoption through a viral analogy. Users transition among uncoupled, coupled and persistently dependent states; once adoption crosses a critical threshold, runaway dynamics can trigger population-level dependence and abrupt losses of cognitive competence, plus technological lock-in. The same framework identifies conditions for "cognitive immunization" — reducing transmission and preserving reversibility.
More from AGI Musings
- Gary Marcus mocks the recurring 'this model is AGI' cycle, citing Tyler Cowen's o3 claim — GaryMarcus · 2026-09-07
- Do code quality norms built for human maintainers still hold as models approach superhuman coding? — burny_tech · 2026-09-07
- OpenAI Chief Scientist Expects Progress Speed to Sustain Into Recursive Self-Improvement — Neurogence · 2026-09-07
- AI safety essay calls for third-party audits and mandated safety bars on scaling — dgrobinson · 2026-09-07
- Researchers worry automated alignment work could enable 'scaling at maximum speed' — dhadfieldmenell · 2026-09-07
- AI's power use looks normal once you divide by users, argues Andy Masley — AndyMasley · 2026-09-07