So I have been thinking. And the more I think about it, the more I realize that what I wrote about before, as serious as I believe it was and as seriously as I stand behind every word of it, did not go far enough. Because I have had a realization since then that has genuinely unsettled me in a way that I want to share with you. The AI bans are worse than I thought. Not just a little worse. Significantly, fundamentally, potentially catastrophically worse. And the reason comes down to something that I think a lot of people are not fully reckoning with yet, which is just how deeply and how completely AI has been woven into the fabric of the platforms and services and tools that people use every single day, often without even thinking about it or knowing it is there.
Think about your daily life for a second. Think about the apps you open, the services you use, the platforms you depend on. YouTube. Spotify. Google Search. Facebook. Instagram. Healthcare apps and patient portals. Office software. Streaming services. Gaming platforms. Navigation apps. All of it. Every single one of those things has AI baked into it at some level, and in many cases at a very deep level. The recommendation algorithm that tells YouTube what video to show you next is AI. The way Spotify decides what song comes after the one you are listening to is AI. The way Google ranks search results and surfaces information is AI. The way your healthcare provider’s systems flag potential issues in your records or help process your information is AI. This is not a future scenario. This is not something that is coming down the road. This is the present. This is right now. AI is not a separate thing you opt into when you open a specific AI assistant app. It is the invisible infrastructure underneath an enormous amount of what the modern internet actually does and how it actually works.
And here is where it gets really scary. Right now, the United States government is, for the most part, tolerating AI. Not always enthusiastically, not without conflict, as I have written about at length already on this blog, but tolerating it. The interventions we have seen so far, as damaging as they have been, have been targeted. They went after specific models from specific companies. Fable 5. Mythos 5. Claude products from Anthropic. That is a serious thing and I do not want to minimize it, but in the grand scheme of how deeply AI is embedded in our digital lives, what happened to Anthropic was still relatively contained. A significant number of people lost access to tools they valued and depended on. That is real harm. But it is not the same as what I am about to describe, which is what happens when the government decides to go after an AI that is not powering a standalone assistant app but is instead powering something like YouTube’s recommendation engine. Or Google’s search algorithm. Or Spotify’s entire discovery infrastructure. What happens then?
Let me walk you through it, because I think it is important to actually think through the mechanics of this rather than just gesturing at it vaguely. Say the government decides, for whatever reason, political, national security theater, genuine concern, it does not matter for the purposes of this thought experiment, that a particular AI system being used by a major platform is no longer acceptable. They issue a directive. They say this AI cannot be used anymore, effective immediately or within some short window. What does the platform do? Well, if that AI is deeply embedded in the platform’s core functionality, if it is not just a feature you can toggle off but is instead woven into the algorithm that makes the whole thing work, then the platform cannot just flip a switch. They have to go in and surgically remove something that is integrated at a fundamental level. And while they are doing that, while they are trying to figure out how to keep the lights on without the system that has been running underneath everything, the platform probably has to go down. Or degrade significantly. Or both.
And now think about what that actually means for real people. Not in the abstract. Concretely. If YouTube goes down, even for a few days, even for a week, what happens to the people who use YouTube as their primary source of income? The creators who have built their entire livelihoods around that platform, who have audiences and schedules and sponsorship deals and revenue streams that depend on consistent access and consistent algorithmic performance, those people do not just experience an inconvenience. They lose money. Potentially a lot of money. Money they were counting on. Money that in many cases represents their entire household income. And it is not just the big creators with millions of subscribers who would feel that. It is the mid sized creators, the small creators, the people who have been grinding away building something that was just starting to work, the people for whom a week of downtime or a month of degraded algorithmic performance could be genuinely devastating in ways that the platform is not going to compensate them for and the government is certainly not going to compensate them for.
And that is just YouTube. Scale that out. Think about what happens if the AI embedded in a major healthcare platform gets targeted. We are talking about systems that process patient information, that flag potential issues, that help route people to the right care, that are integrated into how medical records get managed and how providers communicate and how prescriptions get processed. If that goes down, if that gets yanked out on the fly because a government directive said it had to go, the consequences are not just financial. They are potentially medical. They are potentially life and death for people who depend on those systems to make sure they get the right care at the right time. And I want you to sit with that for a second. I want you to really think about what it means to live in a world where that is a plausible scenario, where the government has demonstrated that it is willing and able to pull AI systems with relatively little warning and relatively little regard for the downstream consequences, and where AI is so deeply embedded in critical infrastructure that pulling it could cause serious harm to real people.
Now I am not saying this is definitely going to happen. I am not a prophet and I am not trying to be one. What I am saying is that the Fable 5 situation showed us that the government is willing to use these tools. It showed us that the political will exists to intervene in AI deployment in sweeping ways, and that the justifications offered for those interventions do not need to be particularly robust or well evidenced to be acted on. And once you know that, once you have seen that demonstrated in practice, you have to start thinking seriously about what it looks like when that same willingness gets applied to something much more deeply embedded in the infrastructure of daily life. Because the government that banned Fable 5 is the same government that could theoretically go after the AI running Google Search. And the precedent that was set when people largely shrugged about Fable 5 is the same precedent that makes the next intervention easier and the one after that easier still.
This is how these things work. They do not start with the most dramatic possible version of themselves. They start small. They start targeted. They start with something that most people can convince themselves does not really affect them, or that the people it does affect probably had it coming somehow, or that the security concerns being invoked are probably legitimate even if the details are a little murky. And by the time it gets to something that everyone can see affects them directly, the infrastructure for doing it is already built, the legal and political precedents are already set, and the public has already been trained not to push back too hard because they did not push back all the times before when they should have. That is the pattern. And I see it happening here.
I also want to point out something about the economic dimension of this that I do not think gets enough attention. When we talk about AI bans and their consequences, the conversation tends to focus on the technology companies themselves. Will Anthropic survive this. What does this mean for the AI industry. How will the big players adapt. And look, those are not irrelevant questions. But they are not the questions I am most concerned about. The questions I am most concerned about are about the people at the bottom of the economic food chain in the creator economy and the gig economy and the various other economies that have grown up around these platforms. The people who do not have lawyers or lobbyists or venture capital backing or the ability to pivot to a different market if this one gets disrupted. The people who built something real and valuable on top of these platforms because the platforms were there and they were accessible and they seemed stable enough to build on. Those are the people who get destroyed when platforms go down unexpectedly. Those are the people who have no recourse when an algorithm gets yanked and their livelihood goes with it. And those are the people who are most invisible in the conversations happening right now about AI regulation and AI bans and what the government should and should not be allowed to do.
I keep coming back to the same fundamental point that I have been making throughout everything I have written on this topic, which is that the people who make these decisions and the people who cheer for these decisions are not the people who bear the consequences of them. The government officials who issue the directives go home to their salaries and their security. The media commentators who build their brands on AI panic go home to their platforms and their audiences. The activists who call for bans go home to whatever resources and stability they have. And the people who actually get hurt, the content creator who just lost a week of income, the patient whose healthcare system is degraded, the small business owner whose tools stopped working, those people are on their own. They do not have anyone in the room when these decisions get made. They rarely have anyone making the argument on their behalf in public. And increasingly, I think that is exactly the point.
So I am making it here. I am saying out loud what I think needs to be said, which is that AI bans are not just about AI. They are about the infrastructure of daily life for millions of people who never signed up to be collateral damage in a political fight between the government and a tech company. They are about livelihoods and access and the ability of ordinary people to participate in the digital economy that they were told was their opportunity and their future. And they are about whether we are going to be honest with ourselves about where this is headed before it gets to the point where the damage is too big and too widespread to pretend we did not see it coming. I saw it coming. I am telling you it is coming. And I really, genuinely hope more people start paying attention before we find out the hard way what it looks like when an AI ban hits something we all actually depend on.