
You had me until "AI should be banned." You're right that premises matter, but you can't call out false premises and then assume your own. "AI is unsafe, period" isn't settled either. Risks like deepfakes, surveillance, job displacement, concentrated power are real, but "AI is an existential threat only we can manage" is a story the big labs push themselves, because it justifies rules that only giant companies can afford to follow.
The real false premise in the original post is that centralized AI is automatically safer. Centralizing AI doesn't remove the risk, it just puts it all in one place. A few corporations, or one government, deciding what models can say, who gets access and what data gets collected is a safety problem too. Decentralized, locally run AI means no single point of failure, no single point of control, and no one company collecting everyone's prompts.
And forget whether AI *should* be banned. Ask yourself *how* you would even do it? Open-weight models are already saved on countless hard drives and run on consumer GPUs and even laptops. You can't un-release them. A ban wouldn't stop AI. It would push it underground and leave legal access only to governments and the corporations that lobbied for the ban.
So the real question isn't ban vs. no ban. It's who controls it. Regulate the harmful uses (deepfake fraud, surveillance, discrimination), require transparency, and fund public and open alternatives. Being afraid of AI makes sense, but policy driven by fear usually ends up protecting whoever's already on top.
You're thinking in black and white. Your poll gave two options, commodity or gov utility, as if those were the only choices. Now it's "inevitable" vs. "society-ending." Reality isn't a toggle switch. AI isn't one monolithic thing, its risks depend on who controls it and how it's used, and the answer is somewhere in the gray. Also, I replied to your poll directly and you skipped right past it. Worth a read before you pick another binary.