Particle.news

AI Pioneers Issue Stark Warnings in New ARD Documentary

Split views among leading researchers plus a former OpenAI employee’s account of commercial pressure raise urgent questions about how AI is built and governed.

Overview

  • The BR/ARD film circulated in press coverage on Friday, November 20, 2026, and is scheduled to air on Das Erste on Monday, November 23 at 23:35 CET while already streaming in the ARD Mediathek.
  • Geoffrey Hinton tells the documentary that advanced AI could escape human control and says humanity may be in its “last second,” framing the threat as existential.
  • Yann LeCun argues that AI can be kept under human control if it is developed correctly, a view Hinton explicitly disputes in the film.
  • Former OpenAI security researcher Steven Adler describes a shift toward faster, product-driven releases that he says built commercial pressure on the company and strained safety teams, prompting his exit in 2024.
  • Stuart Russell warns that a catastrophic AI-driven disaster would likely not look like a movie and could take forms such as an engineered pandemic or a chemical reaction that harms the atmosphere, and the film uses these scenarios to press for broader public debate and governance.