The International Energy Agency predicts that global electricity consumption from cryptocurrencies, data centers, and the AI sector could double by 2026. Toward this goal, Oklo is looking to create clean energy from nuclear fission reactors and has the backing of OpenAI founder Sam Altman. Oklo CEO Jake DeWitte appears on Catalysts to talk about the company's work and the outlook for energy in the age of artificial intelligence.
DeWitt noted that Oklo is deploying “next-generation” nuclear technology with “enormous” potential, capable of powering the planet for over a billion years. He noted that the energy demands are “staggering” as the AI sector is already building on the drivers of the clean energy transition.
Oklo designs, owns and operates its own nuclear plants and sells the power through long-term contracts, DeWitt said, adding that Oklo sometimes works with data centers to build power plants next to them.
To learn more about expert insights and the latest market trends, click here to watch this full episode of Catalysts.
This article was written by Gabriel Roy
Video Transcript
AI requires a lot of power.
The International Energy Agency predicts that global energy demand will double by 2026.
But where does that energy come from?
Olo produces clean energy from nuclear fission reactors.
This is where atoms split in the reactor.
And that may ultimately be what fuels AI and cryptocurrency.
Well, at least that's what's expected.
The company is also backed by Open AI CEO Sam Altman, who joined from the New York Stock Exchange, as chairman.
We're joined by Jake DeWitt, co-founder and CEO of Olo.
Jake, thank you so much for your time.
First, please give us a brief explanation.
Why does this happen?
Yeah, well, thanks for the invite.
Well, I mean, it's pretty cool to have next-generation nuclear technology, it holds so much promise and potential, and with proven reserves of materials and technologies that can, frankly, power the planet for over a billion years.
It has been proven.
It's what I was passionate about as a kid, what really inspired me to want to work in this field, and what inspired us to start the company.
Uh, you know, 11 years ago, um, and that's exactly what we're doing now.
Sorry, I didn't mean to jump in there, but it just occurred to me.
Pause for a moment and tell us about the magnitude of this opportunity.
This is especially important when you consider the energy required to power an I-deployment.
These products are also not yet online.
Yeah.
So the opportunity right now is huge.
You have a few things going on too, right?
We were obviously talking about a large-scale AI piece that builds on the impetus of the energy transition.
That's exactly what's happening right now.
And that has been going on for several years.
As we move towards a cleaner, more sustainable future and everything from transportation to electrification becomes more widespread, there are high expectations for what that will look like.
How we heat our homes.
And now it's getting a lot of traction on the AI side, with more energy being added with each use.
In many cases, they are still being solved, but at a scale several times larger than simple queries.
And I think that as we start to use “I” more, it becomes more accessible in general.
It just keeps increasing.
Yes, it's a labor force, it's a productivity multiplier, it requires a lot of energy input.
And what we've seen on the data center side has been massive deployments basically trying to grab all the capacity they can.
And there's a lot of interest in what we can create together.
For example, if we build nuclear reactors next to or near where data centers are being developed, or build reactors where we already have them, or build data centers around us, the great thing about that is, of course, it not only provides the power to help us scale and grow and bring more reactors, but it also paves the way for us to bring them to other markets.
But the demand for the AI part is surprisingly high.
Frankly, the numbers are eye-watering and they just keep getting bigger.
With that in mind, what is the revenue model?
How can you achieve profitability and sustain it over the long term?
Yes, this was a big focus for us when we started the company.
It wasn't like the traditional model of nuclear power.
It was about providing something that would make it much easier for customers to buy the power they wanted from a nuclear system.
In other words, nuclear power produces natural, clean, reliable and safe electricity.
We just wanted to make it easy for them to buy.
So we design and operate these power plants ourselves and sell the electricity through long-term offtake agreements.
Essentially, power purchase agreements are very similar to those we see with many other types of energy generation technologies, particularly renewable energy, and allow for a data center company, for example, to enter into a long-term power purchase agreement and buy power from our system.
We then build the power plant and produce the electricity that our customers buy for 20 years – sometimes more, sometimes less – to power their data centers, and then we can add more facilities to support our customers' growth.
Now, recurring revenue is great because you can fund it, which is a big thing.
Plus, we have access to some great investment tax credits. All of these things are great pluses for us, but they're not necessary for us to be successful.
But this is very different from what nuclear power typically does, not just in terms of meeting customer demand, but also in terms of the business benefits of continued revenue from selling electricity.
And that's been a big differentiator for us so far.
We are very unique in our field and offer such a dynamic approach.
Jake, what are you hearing in terms of interest and how are your conversations with AI and data centers progressing at this point?
Yes, we're excited. Yesterday afternoon we announced a partnership with HYPERSCALE Data Center Partners in Wyoming.
This is an exciting opportunity as we look at driving higher computing density with things like advanced liquid cooling technologies.
One of the big benefits of I is that it allows for much higher compute density per rack at the data center level, which opens up a lot of advanced cooling technologies and really opens up interesting integration possibilities, not just in terms of the power that we can produce.
So it has some cool features that utilize heat to help cool it down.
It may sound strange, but it is actually doable and can be combined.
These are very promising capabilities for the next generation data center.
So we're excited about that.
We are working on it further.
The pipeline has been very busy for us.
It feels like there is an even greater demand for the electricity we can bring online.
So I think we're going to have our hands full for quite a while.
Not only will it meet our energy needs, but it will also create a whole range of other opportunities in the context of the wider energy transition.
The data is certainly very interesting, but as we move forward with reindustrializing America's industrial base, we also need to look at the factories.
By the way, power is needed to run these things, including manufacturing the chips for the I.
We're looking at bringing chip manufacturing back to the United States.
So there's a tremendous opportunity, especially in highly industrialized areas like Texas and Ohio.
So, we're very excited about what the opportunity looks like and very happy that we're on track to basically try to build as much as we can as quickly as we can.
