Hindsight is always 20/20. Consider the nuclear renaissance.

From around 2001 to 2010, the world rediscovered commercial nuclear energy. For the first time in a long time, demand for electricity was on the rise, and people wanted something clean, domestic, and affordable. By then, the cultural impact of Three Mile Island and Chernobyl had largely faded, the anti-nuclear movement had lost much of its momentum, and conventional energy prices were on the rise. This convergence of factors made nuclear energy cool again, and a new generation of scientists, engineers, and investors emerged to make a revived run at commercial nuclear energy.

But then the 2011 accident at Fukushima raised concerns about safety, the advent of hydraulic fracturing reduced energy prices, and costs for new nuclear skyrocketed. And just like that, the nuclear renaissance fizzled and was deemed a failure.

But was it?

The world is standing at the threshold of a commercial nuclear revolution, one that would have been impossible without the renaissance that came before it. The renaissance reintroduced the public to nuclear energy and gave them time to become comfortable with it. The failures helped to weed out the weakest companies and reveal industrial inefficiencies. The experience helped policymakers understand that the regulatory structures they created had become obsolete and required reform.

All of this enabled the U.S. public, government, and industry to be better prepared for the next time nuclear energy had an opportunity. With growing demand for power in the United States, a greater appreciation for the value of secure energy, and the negative impact of failed green policies such as higher prices and reduced reliability, could that time be now?

Why Wall Street Is Betting Big on Nuclear Energy

Wall Street certainly thinks so.

Uranium-related investments returned roughly 32 percent annually over the past three years, advanced-reactor startups are raising billions of dollars in private money, and Morgan Stanley now expects $2.2 trillion to flow into the global nuclear industry by 2050.

>>> America Can’t Duck and Cover From Washington’s Nuclear Waste Disposal Failure

This influx of private investment differs from the nuclear renaissance period, when most spending centered on utilities and traditional power producers. This is where the renaissance begins evolving into a revolution.

Certainly, the large, centralized, heavily regulated utilities will play a role in the next era of nuclear energy, but they likely won’t drive it as they did during the initial rise of American nuclear energy in the 1960s and 1970s. That leadership will come from a new generation of power users that are underserved by the current system.

Some are obvious. The tech world needs power for data centers, for example, and the modern grid is incapable of providing it in many cases. But others aren’t so obvious. What about remote villages and towns? Islands? Or entire nations or regions that lack access to modern power infrastructure?

The fact is that new nuclear energy makes sense for all these applications for the exact same reasons. Whether because of geography, policy, or economics, these underserved users find themselves in electricity deserts where conventional energy is difficult to deliver. Nuclear energy, on the other hand, can be delivered anywhere. It doesn’t need a grid to connect to, doesn’t require complex fuel-delivery infrastructure, can be scaled up or down depending on application, and can produce power for an almost limitless spectrum of uses.

But a nuclear revolution is not just a few new reactors to power a handful of underserved technology campuses. Revolution means thousands and thousands of reactors that bring electricity to places and people across the nation—and around the globe.

Decentralizing Power Production for the Next Generation of Nuclear Reactors

Conventional, centralized grids where power is produced by large facilities and distributed over long distances can make sense. But we shouldn’t be tied to them. Extending high-voltage transmission and fuel supply chains through mountains, forests, oceans, and other difficult terrains can be extremely expensive. Nuclear energy microgrids, on the other hand, would locate power generation much closer to the energy consumer, which would reduce dependence on costly, long-distance power-line transmission.

In other words, each community or industrial campus could generate its own power economically and at lower risk.

Nuclear Microgrids Bring Electricity to Islands, Mines, and Data Centers

Today, isolated islands rely primarily on diesel fuel shipped over long distances for power. This can be very expensive, is subject to delivery disruptions, and exposes residents to volatile fuel prices. The International Atomic Energy Agency notes that microreactors, which are small, often portable reactors, could solve this problem by providing an alternative that is more reliable and not subject to the same supply disruptions. Marine-based SMRs are another option that could be constructed in shipyards and delivered to locations where building or expanding a conventional grid is impractical.

Mines and critical mineral processing facilities are also often located far from developed grids and require substantial power to operate. An Idaho National Laboratory analysis found that microreactors that produced roughly 1-20 megawatts could supply remote mines with electricity and heat, replacing expensive diesel generators and allowing capacity to expand as mining operations grow.

Then there are the power-hungry data centers that are driving much of the rapidly rising demand for power in many communities across the country. Though they may be geographically close to energy infrastructure, they exist for all intents and purposes in power deserts because conventional grids often lack the capacity to power them. Here again, nuclear micro-grids could be the answer as data center operators could build their own mini-grid to power their operations. Doing so would not only ensure that America’s tech industry remains competitive, but also reduce demand on conventional grids as power producers adjust to rising demand.

Small Modular Reactors and National Security

Decentralized nuclear power also could provide substantial national security benefits. A locally powered military base or critical facility would be less vulnerable to transmission failures, fuel-delivery interruptions, cyberattacks, and even extreme weather.

The Eielson Air Force Base in Alaska is a great example of how this could work. The base sought a contract with a commercial developer that would finance, license, construct, own, and operate the reactor and sell electricity to the base. This is a great model because it requires private companies to compete for the base’s business, and market forces will ultimately drive new nuclear energy to success.

Time to Go Big on Nuclear Policy Reform

Moving from renaissance to revolution means going big on policy reforms. The goal should not be to build a few nuclear power plants. Rather, we should strive to create an economically sustainable, competitive, innovative, and uniquely American nuclear industry.

This will require a realignment of responsibility. The government’s role should be to protect public health and safety. The private sector’s role should be to operate a competitive commercial nuclear sector. Not fixing this misalignment of responsibilities was the biggest policy failure of the nuclear renaissance, and policymakers must not repeat that mistake if they want a different outcome.

That means getting rid of the subsidies, rethinking regulation and getting Washington out of nuclear-waste management. The reason is simple: Governments are not good at business because they make decisions based on politics rather than on good economic sense.

Reducing Government Risk to Unlock Private Nuclear Investment

Some argue that nuclear energy requires more government control and investment, suggesting it presents more financial, technical, and political risks than other industries. But all big projects have significant financial risk. Private oil refineries can cost billions of dollars, and projects such as skyscrapers, liquid natural gas export terminals, and other large industrial projects all require massive capital outlays. Companies and individuals regularly take big financial risks where they see attractive opportunities.

Then there is technological risk. But nuclear energy is not really that different from other industries. With around 440 nuclear reactors operating globally, technical risk for existing technology is relatively low. That only two new plants have been built in the United States in the past few decades cannot be an excuse, just as it wasn’t an excuse when Americans built the very first commercial nuclear plant in Shippingport, Pennsylvania in 1958. This plant, it is worth noting, was built in less than four years. We should not accept as a fait accompli that a modern nuclear plant must take a decade or more.

Possible technological risks with new designs are not beyond the realm of those posed by innovation in other cutting-edge businesses, such as fracking, offshore energy exploration or tech. Beyond that, as it pertains to nuclear energy, there is a vast federal research infrastructure in place that the private sector can, and does, access to help mitigate that risk.

Political risk, however, is real and uniquely high when it comes to nuclear energy, and it exacerbates financial and technical risk calculations. Any justification for government intervention is based on mitigating government-imposed risk.

But here is the problem.

When government intervenes to mitigate a risk that it has created, it adds another layer of political risk. Worse, it creates dependence, distorts capital flows, incentivizes rent-seeking and lobbying, and forces firms to allocate resources to satisfy politicians and bureaucrats rather than improve their business. This creates misalignments between responsibility and authorities and undermines economic efficiency.

Even worse, politics often changes, making it difficult to build a sustainable business model around political preferences. At best, this approach could yield a couple of reactors or keep some firms above water, but it won’t produce a robust, competitive, innovative nuclear revolution.

The major question is: How does America minimize political risk and let the private sector manage other risks so a robust and economically competitive industry can emerge? It will require changing the Department of Energy’s (DOE) role, bold regulatory reforms, and solving the problem of nuclear waste management.

The Department should not be funding grants, loans, or demonstration projects. Nor should it attempt to improve operations or economics of existing plants or new technologies. The private sector can do these better than government.

However, DOE could have an important regulatory role. The Trump administration’s idea to take advantage of the DOE’s ability to regulate research and demonstration reactors as part of a reformed commercial nuclear ecosystem is a very good one and is already paying dividends.

Firms can build their initial reactors under DOE authority and then take that work to the Nuclear Regulatory Commission (NRC) for an expedited commercial permit review. This process could be a game changer for moving innovative ideas from development to commercialization.

Rethinking the Nuclear Regulatory Commission for Advanced Reactors

What about the Nuclear Regulatory Commission?

Worthwhile attempts are being made to improve the Nuclear Regulatory Commission. Indeed, complaints about the agency being a roadblock to nuclear development have significantly diminished. An efficient, predictable, and affordable regulatory process for new reactor technologies is essential and emerging.

But America needs to think bigger. While NRC’s reforms are undeniable and the agency should be applauded for its modernization efforts, new ways to safely permit reactors must still be considered if the goal is for nuclear energy to reach its full potential.

For example, states should be authorized to take a larger role in nuclear power plant regulation. The Atomic Energy Act of 1954 already allows states to regulate some nuclear materials. That should be expanded. If a state can demonstrate the technical know-how to oversee commercial nuclear operations within its borders, it should be free to do so. This should not only be allowed but encouraged by Washington, especially for advanced reactor designs, which tend to be much smaller and have very different safety profiles than large light water reactors.

But even they could be overseen at the state level in some capacity. U.S. utilities have been safely operating large light water reactors for over 50 years. America should not regulate them as new, scary technology, because they are neither new nor scary. The regulatory burden should be significantly lifted on those reactors.

>>> Our Energy Crisis Has a Nuclear Solution

Further, companies should be allowed to build reactors outside the existing NRC regulatory regime if they obtain their own liability insurance against accidents. In exchange, they would forgo participation in the federal Price-Anderson program that currently provides liability coverage. Imagine a regulatory framework that is light on NRC regulatory rules but heavy on liability insurance requirements.

Some might question whether private insurers would cover a nuclear reactor absent a government backstop. But given the outstanding safety records of existing reactors and promises that new technologies are safer, this should be an option. Insurance comes in many forms, and no one can predict what could ultimately emerge.

However, the insurance industry is extraordinarily sophisticated and does a tremendous job at pricing risk. It will be effective at ensuring that only the safest nuclear plants are built and that the right incentives are in play compared to government guarantees.

Fixing Nuclear Waste Management Through Private-Sector Competition

Finally, there is the question of what to do with nuclear waste—or, more accurately, spent nuclear fuel. The federal government took responsibility for managing the nation’s spent nuclear fuel in 1982. By removing responsibility from spent fuel producers, the 1982 Nuclear Waste Policy Act removed any incentive for the nuclear industry to integrate spent fuel management into its long-term business planning and left it to Washington bureaucrats instead. It should surprise no one that the plan has failed.

Reforms are needed to reconnect the nuclear industry to waste management. This includes transitioning responsibility for nuclear waste management away from Washington and to the private sector, introducing competition for waste management services, and ensuring that prices are connected to services rendered. These reforms would allow a private spent fuel industry to emerge that would drive innovation in reactor technologies and spent fuel processing. They would allow the nuclear industry and communities to engage in real negotiations, bound by legal contracts, to build and operate spent fuel management facilities.

These reforms are a major departure from the status quo, but they are reasonable, not radical. They would foster good governance and economic progress in the industry. Most important, they are necessary if we are serious about revolutionizing nuclear power in the United States. And make no mistake, they would be impossible if not for the nuclear renaissance, and for that reason, it was not a failure.

This piece originally appeared in The National Interest.