The Left’s AI Narrative is Wrong - And Dangerous

If you listen to Bernie Sanders, Elizabeth Warren, and others on the left, AI is going to make a handful of tech oligarchs super rich while the rest of us pay higher electricity bills and live next door to dangerous data centers. They are supporting legislation all the way from bans on new data center construction to nationalizing 50% of certain AI companies through a novel “equity tax.” One of their principal arguments is that AI companies are profiting from knowledge that all of us created together. The implication is that AI is something that is being done to the American public rather than for it.

We are all used to the tension between Sanders and Warren and the business community. The risk in this case, however, is that if their narrative of the AI marketplace becomes widely accepted, we could as a country mismanage one of the biggest developments of our time. Doing so could have terrible consequences for our economic competitiveness in the world, as well as our fundamental national security.

A pro-growth agenda requires that we reject the Sanders/Warren argument and harness AI for the betterment of all of us.

It isn't the knowledge - it's what you do with it

The Sanders/Warren argument goes like this: AI systems are trained on data, books, articles, and the accumulated knowledge of humanity. That collective knowledge belongs to the public. Therefore, the companies building AI are essentially mining a public resource for private gain.

The first flaw in this logic is obvious. It is like saying that because Encyclopedia Britannica built a business by organizing centuries of human knowledge into books, it was not entitled to the profits.

Every technological development in history has been built on the accumulated knowledge that preceded it. Steve Jobs didn't invent the transistor, the cathode ray tube, or the computer mouse. He assembled them into something people wanted to use. Alexander Fleming didn't create the biological processes that enabled him to discover penicillin. He observed them and figured out what they meant. The Wright brothers may have invented the airplane, but they didn't invent the internal combustion engine that propelled it.

Oil companies don't create the oil in the ground, but nobody argues that they don’t have the right to profit from extracting, refining, and selling it. Similarly, steel companies don't create the iron ore that they extract and convert to steel. Drug companies are built on centuries of chemistry and biology developed in laboratories and public universities around the world.

The recurring theme of all of these developments is that it isn’t the input that generates the value added. It is the processing, reshaping, and delivery of the inputs that does so. In the case of AI, it isn’t the accumulated store of human knowledge that is of value, it is the AI’s ability to make it useful to humans that creates the potential value.

And one final note I can’t resist. Warren went to law school. For a short period of time she practiced law. She didn’t create the law; the law was developed over centuries by those that came before her. Nonetheless, she benefited from applying it to the needs of her clients. I doubt that she ever felt like she was unfairly profiting from the knowledge that came before her.

This only works if people find it useful

The second flaw in the Sanders/Warren framing is the assumption that AI wealth is automatic or inevitable. In fact, it is far from it.

First, if consumers don't find AI tools useful, they won't buy them. If companies don't find that AI makes their employees and processes more productive, they won't pay for it. History is littered with failed technology that originally had promise. When was the last time you used a BlackBerry or 8-track? Or a land-line telephone?

Furthermore, even if the technology itself lives up to the current hype, it doesn’t guarantee that any particular provider, or related founder, will survive. People are still risking their capital and their time to compete in our free market with no guarantee of the outcome.

If AI succeeds, it will do so because it genuinely improves people's lives by giving them access to productivity tools and information that they didn’t have previously. Many of us already benefit from it day to day. We are also already seeing the benefit to businesses that implement it.

Radiologists are using AI to detect cancers in scans that human eyes miss. Farmers are using AI-driven precision agriculture to improve crop yields while reducing water and fertilizer use. Drug researchers are using AI to compress the discovery timeline for new medicines from decades to years. The FAA is building AI tools to predict traffic conflicts hours ahead and cut delays and fuel burn. Utilities are using it to predict equipment failures before they cause outages. Weather services are using it to forecast storms earlier and more accurately, giving people more time to get out of the way. Translators built into a phone now let a traveler hold a conversation in a language they have never studied.

These productivity gains won’t be contained at the corporate level. They will flow through to the consumer in lower prices, better services, and products that don't exist yet. That is how technology has always worked.

Yes, some people will get rich

That is one of the attributes of capitalism. Some people get rich, even very rich. And although those on the left view this as a flaw, they have yet to find a system that is better at generating and then converting innovation into broadly shared prosperity.

While we are accustomed to the relentless campaign against “millionaires and billionaires”, the problem here is that the failure to distinguish between the challenge of AI and the wealth it creates for a few, will cloud the serious debate that the AI revolution demands.

The correct narrative 

The right framing is not that AI is something nefarious that is being done to us. It is that AI is a technological development that has the potential to dramatically increase productivity across the economy and thereby continue the historical increase in the standard of living of the average person.

Although it may be larger in size, scope, and speed than the many technological revolutions that have come before it, it is in essence the same.

Because of the magnitude of its potential impact, it merits serious attention and debate by policymakers and the industry. In my view, the appropriate agenda is fairly straightforward.

Education: It starts with education. Our schools still prepare students for a world that is disappearing. We need to develop the judgment, creativity, and adaptability that AI cannot replicate, and we need to teach people how to work alongside these tools and leverage them, rather than compete with them. A student who graduates today without knowing how to use AI is like a student who graduated in 1995 without knowing how to use a computer.

Workforce Retraining: Next, we need to plan for aggressive workforce retraining. Major technological developments throughout history have temporarily displaced workers. This displacement is real, it is painful, and it is the legitimate concern. The answer, however, is not to fight the change, but to invest in retraining pathways for people of all ages and all stages of their careers. It may also be necessary to provide a temporary safety net for those that struggle with the transition.

Infrastructure: AI runs on electricity and data centers. Detractors want to block both. But you cannot lead the world in a technology while refusing to build the thing it runs on. We need to expand power generation, modernize the grid, and permit data centers in a way that protects communities without strangling the industry. China is not debating where to put its data centers. It is building them.

National Security: The national security implications should worry us the most, and they get almost no attention in the Sanders/Warren narrative.

AI is already remaking warfare. In Ukraine, AI-guided drones now identify and strike targets on their own, and a single operator can direct a swarm of dozens of them at once. The economics are brutal, and they favor whoever has the better software: a cheap autonomous drone can destroy a tank or a fuel depot that cost a thousand times as much. Earlier this year, China demonstrated one soldier controlling 200 autonomous drones in a single coordinated formation. The same capability that lets AI read a medical scan lets it pick a target out of a video feed.

The advantage extends well beyond the battlefield. AI is becoming central to cyber defense, to intelligence analysis, and to the race to control space, where satellite networks now decide whether a country can communicate, navigate, and fight at all. The nation that leads in AI will hold the kind of decisive edge that air power conferred in the last century.

Other countries understand this. They are not going to stand still while we argue about the location of data centers and how much AI executives are paid. If we fall behind here, no equity tax will save us. This is the strongest reason of all to get the narrative right, and the one Sanders and Warren ignore entirely.

Data Privacy: AI systems are trained on, and increasingly fed, enormous volumes of personal information: what we write, what we search, what we buy, where we go, and what we say to a chatbot at two in the morning. We should require clear disclosure of what data is collected and how it is used. People should have the right to know when they are interacting with an AI system, the right to see what it has on them, and the right to have it deleted. Sensitive categories such as health, financial, and biometric data deserve heightened protection. And companies that profit from our data should bear real liability when they fail to safeguard it.

None of this will be easy. All of it will require work and compromise. But it can and must be done if America is to continue to prosper and compete on the world stage.

The historical verdict is already in

In my previous piece on this topic, I wrote about the long arc of technological change and why the fear of AI, while understandable, is not new. Farmers became factory workers. Factory workers became software engineers. Software engineers will become something we don't have a name for yet.

Every generation in the middle of a technological transition believed it was different, that this time the disruption was too big and too fast to survive. Every one of them was partly right about the magnitude of the change. Every one of them was wrong about the conclusion.

It is no coincidence that the leaders of the AI revolution are American companies. It is the result of America’s entrepreneurial spirit and healthy capitalistic system that much of the world’s technological innovation begins here.

Now it is up to those that make our laws and regulate our commerce to do their part. The question for them is not whether AI happens, it is whether we will manage the transition well or badly. Stoking fear and resentment of the people building it is not management. It is the political equivalent of smashing the looms.

Get the narrative right, and the right governing agenda will follow. That governing agenda is a pro-growth agenda. There will be bumps along the way, but ultimately, we will all prosper. Get it wrong, and we will preside over a decline in American competitiveness and security unlike any in our modern history.