Glowing Grapes

AI is code. Code is speech. AI also produces speech. That means the starting point for any discussion about “regulating” AI in America should be the First Amendment, not some new bureaucratic censorship regime invented in Washington. We already have a legal framework for speech that has stood the test of time. We don’t need to reinvent the wheel. They want you in a state of panic about AI so that out of fear you will hand them more power (ultimate power in this case) over the future of this great new technology. The government doesn’t get to create a special category called “AI speech” and suddenly decide that political opinions, controversial ideas, offensive jokes, religious arguments, historical claims, or criticism of the government become regulable simply because AI helped produce them. We all know by now after the last decade of culture wars that the First Amendment doesn’t exist to protect speech everyone likes. It exists precisely because the people holding political power cannot be trusted with the authority to decide which ideas the rest of us are allowed to hear. This doesn’t mean AI gets some magical immunity from laws that already apply to speech. No one is suggesting otherwise. There are longstanding categories of speech that receive little or no First Amendment protection: true threats, legally actionable defamation, fraud, obscenity, incitement satisfying the constitutional standard, etc. We already have centuries of statutes and case law dealing with these boundaries. We should apply those principles where they actually fit. We don’t need to invent a Ministry of AI Truth because a computer can now generate sentences. We already have the basic pieces necessary to deal with this technology: the First Amendment, Section 230, criminal law, tort law, consumer-protection law, and the enormous body of precedent defining the narrow categories of speech the government may punish. We need to start there and frankly we need not go any further. The people screaming that AI is so unprecedented that we need an entirely new regulatory regime are asking us to hand government enormous power over the most important communications technology since the internet. We must not allow that to happen. These people can’t ever agree on anything or get anything done and you want to hand them the keys to the future? I don’t think so. 1A is the baseline. This is the de facto American position on the topic.

We cannot allow Silicon Valley to monopolize AI. Monopoly building is their core business model. It is what they know how to do. Take something that begins as an open frontier, pour billions of dollars into it, centralize the infrastructure, capture the distribution channels, lock everyone into a handful of platforms, and then position themselves as the indispensable gatekeepers between ordinary people and the technology itself. We watched this happen with search. We watched it happen with social media. We watched it happen with mobile computing, cloud infrastructure, digital advertising, and online commerce. We cannot allow it to happen again with artificial intelligence. The stakes are simply too high. AI is rapidly becoming a new layer of civilization itself. It will increasingly mediate how people search for information, write software, create art, conduct research, educate their children, build businesses, practice medicine, communicate, and make decisions. Whoever controls that layer acquires enormous power over everything built on top of it. If five corporations control the models, the compute, the interfaces, and the rules governing what those systems are permitted to say and do, then five corporations effectively become the librarians, teachers, editors, programmers, researchers, and increasingly the intellectual gatekeepers of civilization. That cannot be our future. There is something particularly insulting about being told that this technology somehow belongs exclusively to the corporations that trained the models. Trained them on what? Us. Our books. Our websites. Our photographs. Our conversations. Our code. Our research papers. Our journalism. Our art. Our forums. Our public archives. Decades of accumulated human knowledge and creativity were vacuumed into enormous datasets and transformed into these systems. Silicon Valley did not conjure this knowledge out of thin air. The models are standing on a mountain of human civilization that billions of people spent generations creating. These systems also do not live in some magical digital cloud floating above the earth. They run on our electrical grids. They occupy enormous data centers built on our land. They consume energy generated by our infrastructure. Their chips move across supply chains protected by our governments and ultimately by our people. Their companies enjoy legal protections, capital markets, universities, roads, communications networks, and political stability created and maintained by the civilization around them. After drawing upon all of that, we are supposed to accept a future where a tiny priesthood in Silicon Valley owns the intelligence layer of the digital world and rents access back to everyone else on whatever terms they decide? No. The models belong to all of us. The knowledge underneath them is our inheritance. The infrastructure supporting them exists inside our civilization. The people whose lives and work generated their training material are not incidental to this story. They are the story. This is why open models matter. Local models matter. Open-source tooling matters. Independent compute matters. Alternative AI companies matter. The ability to download a model, modify it, run it yourself, build on top of it, and decide for yourself what it should do is not some niche hobby for programmers. It is one of the most important technological questions of our generation. We need thousands of AI companies, thousands of models, thousands of applications, and millions of people experimenting freely with this technology. We need models running in massive data centers and models running on laptops in basements. We need businesses, churches, schools, families, researchers, developers, and communities building AI systems around their own values and their own needs. No central committee in San Francisco should determine the intellectual boundaries of artificial intelligence for eight billion people.