Like Sabine Hossenfelder, I Was Offered Money to Tell You AI Will Kill Us

YouTube · Watch/Read · (gathered 2026-10-18)

• El critically investigates the scientific evidence for AI’s recursive self-improvement (RSI) leading to human extinction, emphasizing an independent, research-based perspective free from financial influence.
• The video acknowledges that some aspects of RSI are “more real than expected” and concerning, but identifies a “hidden dependency” as a potential bottleneck, questioning if it could vanish or if the self-improvement loop would truly “run away.”
• Despite reviewing twelve research papers, El concludes there are major gaps between what has been demonstrated in research and public claims, specifically challenging the scientific evidence that directly connects recursive self-improvement to “zero humans.”
• The implicit takeaway is to approach AI extinction claims with a critical, evidence-based mindset, distinguishing between genuine research concerns and unsubstantiated public narratives.