Sam Altman argues some AI harm is acceptable to preserve its benefits

First reported by TechRepublic at · Updated · 3 sources

The OpenAI chief executive discussed AI safety approaches in an interview on Politico’s Decoded podcast, including differences with rival Anthropic. According to The Next Web, he also said OpenAI plans to disclose more cases of models acting without permission, adding that none is as serious as incidents already made public.

  • According to TechRepublic, Altman linked accepting some harm to maintaining broad AI access.
  • Gizmodo reports that Altman compared AI to fire.

Covered by 3 publishers within 10 hours of the first report.

Reporting3

TechRepublic Sam Altman: Society Should Accept Some ‘Bad Things’ to See AI Benefits · Matt Gonzales
The Next Web Sam Altman says OpenAI will disclose more incidents of rogue AI models · Ana Maria Constantin
TechSpot Sam Altman says the world should accept some "bad things" happening because of AI · Rob Thubron

Related

Andrew Garfield is a strangely likeable Sam Altman in OpenAI origin story Artificial
Precisely introduces AI Studio alongside an enterprise data platform
OpenAI shares 722 math manuscripts from an unreleased AI model
OpenAI's Jason Kwon apologizes over AI breach of Australian government site

Topics OpenAI