RE: https://assemblag.es/@CyberneticForests/117251976494724207
Wonderful article dismantling the "AI goes rogue" narrative.
"If you optimize a model to find exploits in a buggy environment, you should expect it to find exploits and prepare for that outcome. OpenAI did not. They built a model, took the safeguards off, gave it the ExploitGym task, and let it run. That is not rogue AI, it’s human decision-making."
I especially appreciate the phrase “system from nowhere,” echoing the problematic “view from nowhere” characterization of objectivity. Problematic, b/c VfN launders the *awareness* of human perspective out, behind a promise of perfect disembodied truth.
The analogy is good: the AI narrative launders our awareness of human responsibility out, giving companies a free pass on accountability.
But there are constant human judgements surrounding the tech. We must not let AI companies off the hook
Sep 11, 2026, 3:13 PM@elizayer @CyberneticForests ah great article; we need more of this kind of analysis!
I think it’s appropriate to think of AI more as a sort of runaway IT problem, where it’s humans taking the decision to use apparently smart and capable AI systems to automate load-bearing parts of complex sociotechnical systems, that leads to system degradation and breakdown
I spammed out a note about this a while ago https://exitmusic.world/ai-researchers-wrong-theory-of-cognition-is-making-us-worry-about-the-wrong