Six years ago I wrote a blog post called Gaslighting Liberty. In that post I separated my concerns about privacy and government power into two categories: the “now” and the “not yet.”

The “now” was easy. We were already willingly carrying tracking devices in our pockets. We were allowing Google to know where we went, Amazon to know what we bought, Facebook to know who we knew, and Alexa to sit in our kitchens listening for us to ask what the weather was going to be tomorrow. We traded privacy for convenience.

Willingly.

The “not yet” concerned me much more. The technology existed, or was beginning to exist, but the pieces had not yet been completely connected. My concern wasn’t necessarily what the government was doing with all of that information at that exact moment. My concern was what someone could eventually do with it once the infrastructure existed.

Six years later I think I need to revisit that distinction.

Because a lot of the “not yet” is becoming the “now.”

I have had a lot of conversations lately about AI, data centers, automated license plate cameras, digital identification, cryptocurrency, stablecoins, Central Bank Digital Currency, our national debt and the United States Government’s seemingly endless appetite for spending. Most conversations about these subjects immediately become arguments about intent.

Are data centers being built so the government can spy on us?

Is Donald Trump secretly planning a CBDC even though he publicly opposes one?

Are license plate cameras part of some master surveillance plan?

Is Digital ID being created so the government can eventually control us?

I think those questions miss the point.

I don’t care what the system is being built for.

I care what the system is capable of doing once it exists.

There does not have to be some secret meeting taking place in Washington where Donald Trump, Google, Microsoft, the Federal Reserve, Flock Safety, the CIA and JP Morgan sit around a giant conference table designing the future American surveillance state. That is almost certainly not how something like this would happen anyway. Each piece can be created independently for a perfectly reasonable purpose.

A city installs license plate cameras because somebody stole a car. A bank increases financial monitoring because it has to comply with anti-money-laundering laws. The government develops Digital ID because identity fraud is costing taxpayers billions of dollars. Artificial intelligence companies build enormous data centers because AI requires enormous computational power. Stablecoin companies create digital dollars because customers want faster and easier transactions. The federal government encourages stablecoins because they strengthen international demand for the U.S. dollar and create another purchaser of United States Treasury securities.

Every individual decision can make sense.

That doesn’t mean I want all of those systems connected together.

The United States Government currently has an enormous financial problem. We spend too much money. That statement should not be controversial.

Republicans complain about spending when Democrats are president. Democrats complain about the things Republicans spend money on when Republicans are president.

Then everyone spends.

Donald Trump campaigned as someone who was going to change that. Instead we continue running trillion-dollar deficits. We continue spending money on wars. We continue financing an enormous federal bureaucracy. We continue paying interest on the debt we accumulated paying for everything else. According to the Congressional Budget Office, debt held by the public is already approximately the size of the entire American economy and is expected to continue climbing.

This is not sustainable forever.

That doesn’t mean the dollar collapses tomorrow. It doesn’t mean America suddenly becomes Venezuela. It means the United States Government has an enormous incentive to make sure the rest of the world continues wanting dollars and, more importantly, continues financing American debt.

This is where stablecoins get interesting.

Donald Trump has been very clear that he opposes a Central Bank Digital Currency. He actually signed an Executive Order prohibiting federal agencies from establishing or promoting a CBDC.

Good. I completely agree.

But I am starting to wonder if we are concentrating too much on what the currency is called.

The Trump Administration has simultaneously embraced privately issued dollar-backed stablecoins. The government openly sees them as a way to strengthen the international position of the United States dollar. The companies issuing those digital dollars need assets backing them, and one of the most obvious reserve assets is United States Treasury debt.

Think about that.

Someone living in Argentina, Turkey, India or Brazil doesn’t have to decide that he wants to lend the United States Government money. He just wants dollars. He buys a dollar-backed stablecoin. The company issuing that stablecoin takes his money and buys reserve assets that can include short-term United States Treasury securities. The foreign customer gets his digital dollars. The United States Government gets another purchaser for its debt.

Pretty clever.

It also means that an increasingly digital dollar system could become very useful to a government that desperately needs continuing demand for dollars and Treasury securities.

And this is where my concerns about CBDCs have changed.

Maybe we never get a CBDC.

Maybe Donald Trump keeps his promise.

Maybe the Federal Reserve never creates a digital wallet for every American.

Maybe there is never anything officially called a “Digital Dollar.”

Does it matter?

If almost all of our money eventually moves through regulated digital systems connected to our identity, with every transaction electronically recorded and the technical ability to freeze funds when ordered by the government, how important is the distinction?

Technically it isn’t a CBDC.

Wonderful.

Now let’s add Digital ID.

The government and private industry have perfectly legitimate reasons for wanting better digital identification. Identity theft is real. Fraud is real. People lose driver’s licenses. Proving that you are actually you online can be difficult. NIST is already developing standards surrounding digital identity and mobile driver’s licenses.

Again, none of that bothers me by itself.

The problem is what happens when identity and money become inseparable.

A digital identity answers one question:

Who are you?

Digital currency answers another:

What are you buying?

Now we just need to know where you are.

Fortunately, we are working on that too.

Automated license plate readers are appearing everywhere. You have probably seen them without even realizing what they are. They are often installed at intersections or entrances to neighborhoods, and they can identify stolen vehicles, locate wanted criminals and provide police with information following a crime.

That sounds great. If somebody steals my Jeep I would be perfectly happy for the police to use every camera available to find it.

But there is a difference between looking for a stolen Jeep and creating a searchable historical database showing where millions of innocent people have traveled.

The Government Accountability Office has already documented federal agencies obtaining access to third-party automated license plate reader information. Federal agencies do not necessarily have to install their own cameras everywhere.

They can access cameras someone else installed.

Private company owns the technology.

Local police department pays for the cameras.

Federal agency accesses the information.

That distinction matters, and it matters legally, not merely practically.

The objection I hear most often at this point in the conversation is that we still have a Fourth Amendment. In 2018 the Supreme Court decided Carpenter v. United States and held that the government generally needs a warrant to obtain historical cell phone location records, precisely because those records reveal a person’s familial, political, professional and religious associations. That decision should have settled a great deal of this.

It didn’t.

In 2023 the Office of the Director of National Intelligence declassified a report acknowledging that intelligence agencies purchase enormous quantities of commercially available information about Americans, including location data, from private data brokers. The report conceded that the Intelligence Community has no community-wide position on whether Carpenter even applies to purchased information, and that individual agencies have interpreted the decision narrowly enough to keep buying. The report also acknowledged that supposedly anonymized commercial data can frequently be de-anonymized back to specific individuals.

A warrant restricts what the government can take from you.

It says remarkably little about what the government can buy from somebody else.

But there has always been a problem with collecting too much information. Eventually you collect so much information that it becomes useless.

Imagine having billions of license plate photographs. Who is going to look through them? Imagine having billions of financial transactions. Who is going to compare them? Imagine having location information for hundreds of millions of cell phones. Who is going to make sense of it?

Nobody.

At least no human being.

This is where Artificial Intelligence changes everything.

AI isn’t just another method of collecting information. It gives us the ability to analyze quantities of information that would have been completely useless fifteen years ago.

Where does this vehicle normally spend the night?

What vehicles regularly park near it?

Where does the owner work?

What church does he attend?

What stores does he shop at?

What political events has he attended?

Who else attended those events?

Who travels with him?

What purchases were made when he was at those locations?

What has changed recently in his normal behavior?

That is a completely different level of surveillance.

Not because the information didn’t previously exist.

Because computers can now connect the dots.

And connecting those dots requires an enormous amount of computing power.

Enter the data center.

Donald Trump has made increasing American AI capacity a major part of his economic and national security policy. The federal government is actively making it easier to construct enormous data centers and the electrical infrastructure necessary to operate them.

To be clear, I do not think every data center is being built so the federal government can monitor Jonathan Rodebaugh driving to Kroger.

That would be ridiculous.

Data centers are being built because AI is quickly becoming one of the most important technologies in the world. Businesses want it. Scientists want it. Manufacturers want it. The military wants it. The Intelligence Community wants it.

Everyone wants AI.

But again…

I don’t care what the infrastructure is being built for.

I care what it can eventually be used for.

A data center built today to train a commercial Artificial Intelligence model still exists tomorrow when a government agency wants to analyze five trillion pieces of location data.

Technology does not care why we originally built it.

This is why I get frustrated when people demand proof of a conspiracy.

I don’t have proof of a conspiracy. I’m not alleging one. I don’t know what Donald Trump ultimately wants to do with any of this. I don’t know what JD Vance would do with it. I don’t know what the next Democratic President would do with it. I don’t know what some President twenty years from now would do with it.

That is exactly why I don’t want them to have it.

Americans have this bizarre tendency to judge government power entirely by who is currently in charge of the government.

Republicans love executive power when a Republican is President. Democrats love executive power when a Democrat is President.

Then the election happens.

Suddenly everybody discovers the Constitution again.

Maybe instead of trusting the correct politician with enormous power we should stop creating enormous power.

This is not a Donald Trump problem. It is not a Democratic Party problem. It is not a Republican Party problem.

It is a power problem.

And the most dangerous part is that I don’t think the system I am describing would ever be introduced to us as a system of control.

It would be introduced to us as a solution.

That is how these things always work.

Something happens. A terrorist attack. A financial crisis. A cyberattack. Another pandemic. A major war. A currency crisis. Who knows.

Suddenly we have an emergency.

We need Digital ID to stop fraud. We need to monitor financial transactions to stop terrorism. We need to temporarily restrict withdrawals to prevent a run on the banks. We need to know where vehicles are traveling to keep everyone safe. We need Artificial Intelligence to identify threats before they happen. We need to connect databases because investigators don’t have time to use the old system. We need emergency government payments deposited directly into everyone’s digital wallet.

Don’t worry.

It’s temporary.

If that sounds like paranoid speculation, it shouldn’t. It has already happened in a country that looks a great deal like ours.

In February of 2022 the Canadian government invoked the Emergencies Act in response to the trucker convoy protests in Ottawa. Among the emergency measures was an economic order directing banks, credit unions, insurance companies and crowdfunding platforms to freeze the accounts of people connected to the protests. Roughly two hundred accounts were frozen. No court order. No conviction. Some of the people affected had simply donated money.

Canada’s Federal Court later ruled the invocation unreasonable, and in January the Federal Court of Appeal agreed, finding that freezing accounts without prior judicial authorization violated the Charter’s protection against unreasonable search and seizure.

Four years later.

That is the part worth sitting with. The courts eventually said no. The accounts were frozen anyway, within days, by executive order, in a constitutional democracy with an independent judiciary and a free press.

The legal remedy arrived four years after the capability was used.

And look!

The federal government just deposited $5,000 worth of emergency digital tokens into your account. All you have to do is download the wallet.

Most Americans will happily do it.

Of course they will. We have already demonstrated repeatedly that Americans will surrender astonishing amounts of privacy for convenience.

Add free money?

Game over.

If cash becomes increasingly inconvenient, digital identity becomes normal, digital money becomes normal and government payments eventually arrive digitally, there doesn’t have to be a law banning cash.

Cash can simply become irrelevant.

Businesses stop wanting it. Banks make it difficult to withdraw. Young people stop carrying it. Government programs don’t use it.

Eventually you still technically have the freedom to conduct an anonymous cash transaction.

Good luck finding anyone willing to take it.

That is the slow version. A government that decides to move faster does not need decades of cultural drift.

In November of 2016 the Indian government announced with a few hours’ notice that its 500 and 1,000 rupee notes would no longer be legal tender. Roughly eighty-six percent of the country’s currency by value, voided overnight. Ordinary people stood in bank lines for weeks trying to exchange what was in their pockets.

India’s stated reasons were counterfeiting, corruption and untaxed cash.

Perfectly reasonable purposes.

This is how I think an American version of a Social Credit System would eventually develop.

Not with some giant red-white-and-blue website called SocialCredit.gov.

That would be far too obvious.

It would be decentralized. Private companies hold much of the data. Banks know what you purchase. Phones know where you are. License plate cameras know where your vehicle travels. Digital credentials prove who you are. Social media knows who your friends are and what you believe. Artificial Intelligence analyzes everything. Data centers provide the computing power. Government has legal mechanisms to gain access.

Nobody technically created the system.

Everyone just created one piece of it.

That is actually much more frightening to me than a traditional CBDC.

A CBDC is obvious. Congress debates it. People protest it. Politicians campaign against it.

What do you protest when there is no single system?

Your bank? Your cell phone? Your driver’s license? Flock cameras? Artificial Intelligence? Amazon Web Services? Visa? The Department of Homeland Security? The local police department?

None of them individually constitutes the thing you are worried about.

The power exists in the connection.

I think this is the mistake we continue to make when talking about privacy. We look at every individual sacrifice and ask whether it is reasonable.

Can police use cameras to catch murderers? Of course. Should banks identify terrorists financing attacks? Obviously. Should government stop identity thieves from stealing Social Security payments? Yes. Should Artificial Intelligence help us prevent cyberattacks? Absolutely.

Every argument sounds completely reasonable when considered individually.

But liberty is not lost individually.

It is lost cumulatively.

One reasonable compromise at a time.

This is why I increasingly believe that inconvenience is sometimes an essential component of freedom.

Cash is inconvenient. Warrants are inconvenient. Probable cause is inconvenient. Separate databases are inconvenient. Anonymous movement is inconvenient. Due process is incredibly inconvenient.

Good.

Those things are supposed to make exercising power over another human being difficult.

That is a feature.

Not a bug.

I don’t want the government to instantly know everywhere I have traveled. Not because I am doing something illegal. Because it is none of their damn business.

I don’t want every dollar I spend permanently connected to my identity. Not because I am laundering money. Because it is none of their damn business.

I don’t want a Republican administration deciding whether I am responsible enough to use my own money. I don’t want a Democratic administration deciding whether I am environmentally responsible enough to use my own money. I don’t want any government deciding whether my associations, purchases, movements or beliefs make me suspicious.

Liberty doesn’t require the government to trust me.

Liberty requires the government to leave me alone until it has a legitimate constitutional reason not to.

There is a tremendous difference.

I don’t know where all of this ends.

Maybe nowhere. And I want to be specific about what would convince me I was wrong, because I think anyone making an argument like this one owes you that.

If twenty years from now I can still buy something with cash without anyone finding it strange, if the government still needs a warrant to learn where I have been and cannot simply purchase the answer from a vendor, if moving my own money does not require presenting a government credential, and if these databases remain both legally and technically separate — then I was worried about nothing.

I would be delighted to be wrong.

But that isn’t the direction I see us moving.

What I see is financial infrastructure becoming more digital. Identity becoming more digital. Surveillance becoming more comprehensive. Artificial Intelligence becoming exponentially more capable. Computing infrastructure becoming exponentially larger. Government debt becoming increasingly unsustainable. And Americans becoming increasingly comfortable trading privacy for convenience and safety.

I don’t need to know whether someone has a plan.

I can see the pieces.

In 2020 I wrote about the “now” and the “not yet.”

The “now” was frightening enough.

The “not yet” was the infrastructure we were creating that someone might eventually use.

Six years later, I think the distinction is disappearing.

The “not yet” is now.