Every few months the market finds a new excuse to pile into companies that might, someday, produce electricity. The current version of this trade centers on three names tied to AI power demand: Oklo, NuScale, and Bloom Energy. A recent piece comparing the three makes one point that matters more than all the rest combined. Bloom Energy is generating real commercial revenue right now. Oklo and NuScale are not.

That distinction sounds simple enough. You would be amazed how many investors skip right past it.

Bloom makes solid oxide fuel cells. Real products that real customers buy, today, to generate power on-site. The company has been around since 2001, went public a few years back through a SPAC, and has spent the time since burning through plenty of cash while slowly building a legitimate commercial business. They ship working server units to paying customers. Revenue is growing. They are not profitable on a consistent basis yet, but the direction is clear and the numbers are getting bigger.

Oklo is a different animal entirely. The company is trying to build small modular nuclear reactors. Design approvals, regulatory pathways, demonstration projects. All necessary steps if you want to operate a nuclear reactor in this country, and all steps that take years, sometimes decades, before a single watt of electricity reaches the grid. Oklo went public in 2024. It has a compelling story about clean, reliable baseload power for data centers. What it does not have is a commercial reactor operating. There is no revenue from selling electricity. There are designs and partnerships and plans.

NuScale is further along in some ways, having actually received design certification from the Nuclear Regulatory Commission for its SMR concept. But a planned project in Utah fell apart in 2023 after cost overruns, and the company has been restructuring since. NuScale does generate some revenue, roughly $64 million on a trailing basis, but that is not coming from commercial reactor operations. No reactors are operating. No electricity sales from the core business. The company is engineering, planning, and pursuing regulatory approval while burning cash to stay alive.

I have been investing long enough to remember when fuel cell companies were going to change everything in the late 1990s. Plug Power, Ballard Power, FuelCell Energy. The stories were fantastic. The stock charts looked like rockets. Twenty-five years later, some of them are still around, still promising, still not quite there. I am not saying Oklo or NuScale will follow that path. Nuclear SMRs could genuinely work and could be a meaningful part of the energy mix someday. But investing based on what a company might do in five or ten years, versus what it is doing today, requires a very different risk tolerance than most people actually have.

The market is pricing all three of these stocks as if AI power demand is a guaranteed tidal wave that will lift everyone with a connection to electricity generation. Maybe it will. But the company already selling a working product into that demand has a much shorter distance between here and profitability.

If you are drawn to this space, ask yourself one question. Do you want to own a business that is shipping products and booking revenue, or do you want to own a story about what might happen years from now? Both can make you money. One of them lets you sleep at night while you wait.

As always, I am sharing my perspective, not professional financial advice. Do your own research before putting real money into any of these names.