Could artificial intelligence surpass human intelligence, achieving with its feared self-sufficiency the capabilities to dominate humanity and bypass vital and critical processes across the planet? It sounds like a movie script, but it is the real warning from Daniel Kokotajlo, a former OpenAI researcher who formalized in 2025 a serious report on the risks that technology can bring to humanity. In this episode, you will learn more about the details of this warning and other worrying cases of the apparently uncontrolled evolution of AI, detected in recent simulations and experiments. Do you like these topics, Open Mind? So get the Universo Mente Aberta book and have 600 pages of these subjects always at hand: https://bit.ly/livroUMA
Books mentioned in this episode:
“Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI”, book by Karen Hao: https://amzn.to/4kFfeI5
Universo Mente Aberta episode about old AI warnings (April/2023): • Alertas de riscos e cuidados com a I.A. re...
“AI 2027” REPORT (pdf): https://ai-2027.com/?ending=race
コメント