Anyone see this Kurzgesagt video? As much as I’ve enjoyed their videos on topics like strange matter in the past, I usually take their work with a huge grain of salt, especially since some of their videos were funded by the Gates Foundation and their ilk. This video comes off like a major tech bro wank fest.

They acknowledge near the end that some people consider the “AI dangerous” framing to be marketing, but quickly move back to the main narrative. As far as I know, this is all OpenAI’s account of the events, or basically, “Sam Altman says…”. There are no details about the “agents”, or what their system prompts look like, etc. Then a few weeks later, the Nvidia HuggingFace acquisition was announced. Not sus at all.

Does anybody believe for a second that this was anything other than OpenAI trying to make a big noise to get their competitors kneecapped with “AI safety” regulations now that LLMs have hit some fundamental limits and open models are rapidly closing the gap? From reading the comments on YouTube for this video, it appears that yes, yes they do. 😬