You can feel it, can’t you? That creeping existential dread that a handful of Silicon Valley elites are building a god in a server farm, and nobody asked for your permission.
It’s a valid fear. When a few private companies make civilization-level decisions behind closed doors, resentment boils over. Enter Bernie Sanders, riding to the rescue with a bill to ban artificial superintelligence and pause advanced AI development. It feels good. It feels like someone is finally fighting back against the tech bros.
We are so terrified of creating a machine that outsmarts us, we’re ready to pass laws that make us stupid.
But look closer. The Sanders bill isn’t really about protecting you from Skynet. It’s a political forcing move designed to shift the AI debate from “how should we regulate capabilities?” to “what should be outright forbidden?” It exposes the real fracture in our politics today: not left versus right, but accelerationist versus precautionist.
The irony here is thick. Some of the loudest voices sounding the alarm on AI are the same people who see it as our best shot at a post-scarcity utopia. They want a world where machines do all the labor, prices drop to zero, and everyone lives a middle-class lifestyle without lifting a finger. Yet, in a moment of panic, they’re ready to burn the bridge to that future.
Here is where the well-intentioned utopian plan collides with ruthless corporate reality. A preemptive ban requires sharp, enforceable definitions. But the more concrete you make the definition of “superintelligence,” the easier it is for savvy engineers to evade or exploit it.
You cannot legislate away the apocalypse, but you can absolutely legislate away your competition.
Think about the mechanics of this ban. If you impose strict compliance costs and regulatory hurdles on AI development, who survives? Not the open-source rebels. Not the garage startups. The only entities with the legal arsenals, compliance teams, and lobbying budgets to navigate a preemptive superintelligence ban are the incumbent frontier labs.
OpenAI, Google, Anthropic—they get to sit behind a wall of federal regulations, safe from the upstarts nipping at their heels. “Protecting humanity” legislation quietly becomes an incumbent moat.
A ban on superintelligence isn’t a shield for humanity; it’s a tollbooth for the oligarchs.
This isn’t just about one bill. This is the test case for whether democratic institutions can assert preemptive control over technologies that could reshape labor, global power, and daily life. If our only response to a paradigm-shifting technology is to panic and ban it, we aren’t governing the future. We are just handing the keys to the future over to the very tech elites we’re trying to stop.
FAQ
Q: If a ban stops AGI, isn't that worth the cost to startups?
A: No, because AGI isn't a light switch you can just turn off with a bill. Vague definitions just push research underground or into the hands of state actors who don't care about US compliance.
Q: How does this actually entrench Big Tech?
A: Compliance costs. Only frontier labs have the legal and compute resources to prove they aren't crossing the 'superintelligence' threshold. Startups and open-source devs will be priced out of the market.
Q: So we should just let AI run wild?
A: No, we need open, democratic oversight of capabilities, not preemptive bans that act as regulatory moats. We need to regulate the deployment of AI, not criminalize the math behind it.