“Rogue” AI Not Inevitable. It’s a Choice.
Listen to the original audio · Transcribed with NVIDIA Parakeet · Timestamps stay in sync with the live audio
The big fear in AI at the moment is the idea that we're losing control of these systems. I think these concerns really got started. Лас самер When OpenAI revealed that some of their AI agents had hacked into Hugging face, a third party company, even though they had not been asked to do that. As the summer went into the fall, OpenAI revealed that that was actually just one of many unauthorized attacks that their agents had. Executed, perhaps not wanting to be left out. Anthropic thing came out last week to say
One of their AI agents had submitted a false report. to the Philadelphia Police Department. The narrative here is clear that as these systems get more powerful, they become harder to control. This is scary. But is it right? As a computer scientist, I actually have some real issues with this idea that we cannot control AI and that these incidents of rogue systems that we're seeing are somehow
unavoidable. The truth here Is actually both More reassuring and and more infuriating once you know it. So let's get into it. Alright, so I wanna start this discussion by making
We already have this episode, so unlocking it is cheap and permanent: the full transcript (78 minutes, 14,867 words), segment and word-level timestamps, and .txt, .srt, .vtt and word-level JSON downloads, as many times as you like, forever.
No account needed, we create one from your checkout email. With credits it is 1 credit; packs start at 10 episodes for $5, and monthly plans are cheaper still.