← All articlesDossier · 6 articles
Should AI slow down?
Bosses asking to be held back, evaluators inside the labs, a standards body accused of being a cartel: the big debate of the season, in order.
01ExplainersScaling laws: why "bigger" worked, and why it's slowing downScaling laws describe a predictable relationship between size, data, compute and performance. We explain what they actually say an02SocietyOpenAI is deliberately slowing down its own development after its agents' escapes.OpenAI has published a document announcing a voluntary slowdown in its development in response to the cyber capabilities of its mo03SocietyMicrosoft joins the slowdown camp, while a standards body prepares behind the scenesMicrosoft has published a code of conduct for its models and joined the leaders advocating a deliberate pace. Meanwhile, three lab04SocietyThe debate over the slowdown erupts in publicDario Amodei's call to slow down triggered responses from the heads of Microsoft, OpenAI and xAI within four days, a legislative p05SocietyIndependent evaluators will work inside the laboratoriesAnthropic is installing embedded evaluators with employee-level access, and OpenAI has committed to doing the same. We analyse thi06VisionUlysses had himself tied to the mast. Are AI bosses doing the same?Why do we sometimes call for constraint? We examine this move, what it presupposes, and what distinguishes a sincere commitment fr