On July 23, 2026, Peter Thiel spoke with Natalie Brunell on the *Coin Stories* podcast about how Bitcoin mining companies are becoming more interested in artificial intelligence and powerful computers. As reported by crypto news sources like BSCN and Wu Blockchain, this shift is happening because miners want to expand their businesses as energy costs change and the need for AI computing power grows.
Peter Thiel pointed out that Bitcoin’s lack of built-in returns for owners is a major drawback. He clarified this isn’t about the profitability of Bitcoin mining itself, but rather how Bitcoin functions as an investment.
MARA Holdings CEO Fred Thiel Highlights Bitcoin’s Yield Limitation
Based on reports from the interview, MARA Holdings CEO Fred Thiel stated that a major drawback of Bitcoin is its inability to automatically earn returns for those who own it.
This highlights a key difference between Bitcoin and traditional investments like stocks or bonds. Most people who own Bitcoin hope to make money when its price goes up, unless they actively use it in ways that involve extra risk, such as lending platforms.
Peter Thiel’s statements don’t suggest he dislikes Bitcoin. Rather, they show his company is considering how best to use its energy resources to generate profits.
Bitcoin miners are finding it more and more crucial to understand their costs. Their profits depend on things like the price of Bitcoin, how hard it is to mine, electricity costs, and how efficient their equipment is. AI data centers, however, usually have longer-term contracts with companies that need a lot of computing power, giving them more predictable income.
AI Data Centers Offer Higher Revenue Per Unit of Power
Thiel’s main idea focuses on how much money each unit of electrical power produces.
As I understand it, Peter Thiel recently shared some interesting cost comparisons. He estimates building a Bitcoin mining facility runs about $1 million per megawatt when you factor in both the infrastructure and the computing power. However, he says setting up the infrastructure for an AI site – *without* including the cost of the computers themselves – is significantly more expensive, ranging from $10 to $15 million per megawatt.
According to Peter Thiel, building the infrastructure for artificial intelligence requires a much larger financial investment than setting up a Bitcoin mining operation – potentially 10 to 15 times more per megawatt of power.
Although AI and high-performance computing have a larger initial investment, they can generate significant income. These systems support clients who need powerful computing capabilities to develop and operate complex programs like advanced AI models.
Thiel believes that strong financial performance could also make investments in AI infrastructure more appealing to lenders. Big tech companies and other well-established businesses could offer long-term contracts or guarantees, which would give lenders a clearer picture of expected future income.
Unlike some other ventures, Bitcoin mining doesn’t guarantee a fixed income. Miners receive Bitcoin and transaction fees as payment, but the value of these earnings fluctuates with the cryptocurrency market and how competitive the mining network is.
Why Bitcoin Miners Are Exploring AI
As a crypto investor, I’m really noticing how much focus there is now on building out the infrastructure for artificial intelligence within the mining world. It feels like a major trend that’s gaining momentum.
Bitcoin mining operations often use a lot of electricity and are skilled at handling complex power systems. Plus, some miners have access to extra energy resources that could be used for other types of computer processing.
This situation allows us to reuse current locations for artificial intelligence and powerful computing, but changing them over isn’t simple.
AI data centers need a much more sophisticated setup than typical Bitcoin mining operations. They require things like powerful cooling, a lot of electricity delivered efficiently, strong internet connections, and specific hardware. Because of these needs, building an AI data center costs significantly more, as Peter Thiel pointed out.
If a company can get the funding and customers it needs, generating electricity can be a profitable investment due to the potential earnings per unit.
This approach lets miners treat electricity like a versatile tool, using it wherever it will generate the most profit.
MARA Maintains Bitcoin Mining While Exploring AI
MARA remains a significant player in Bitcoin mining, but is also exploring potential growth areas like artificial intelligence and advanced computer technology.
As an investor, what really caught my eye is that this company already controls over 4 gigawatts of power generation. That’s a huge amount of existing infrastructure, and it gives them a solid foundation for expanding and building even more in the future – which I think could be very promising.
Thiel’s statements indicate that MARA isn’t planning to instantly switch from Bitcoin mining to AI. The company can keep mining as usual for now, and transition to AI when it makes financial sense and the necessary infrastructure is in place.
From my analysis, this strategy gives the company valuable breathing room. We can keep generating revenue through Bitcoin mining while we thoroughly evaluate the AI infrastructure market, identify potential clients, and explore funding options. It allows us to pursue AI without immediately abandoning a profitable venture.
This approach highlights a growing pattern with mining operations – they’re looking for ways to better utilize their energy resources. With artificial intelligence firms needing more and more electricity and data center space, having a stable power supply is now incredibly important.
Bitcoin Mining Faces a Changing Energy Equation
The core of the argument about whether Bitcoin mining or AI data centers are ‘better’ really boils down to how much electricity they use and the resources available to support them.
Setting up large-scale Bitcoin mining operations is generally easier than building advanced AI data centers. Unlike AI, which needs complex infrastructure, you can get Bitcoin miners up and running more quickly and with less construction.
Mining cryptocurrency comes with a risk: income depends heavily on how the market is doing. Bitcoin’s value can change rapidly, and how hard it is to mine new coins adjusts based on how many miners are competing. Plus, the cost of electricity significantly impacts whether mining is actually profitable.
Building and running AI data centers is significantly more expensive and complicated than traditional ones. However, they can potentially secure contracts with major companies, offering a stable and predictable income stream. This reliable revenue can make it easier to fund the costly infrastructure required.
Here’s why AI is becoming so popular in the industry. The goal isn’t to get rid of Bitcoin, but to see if the energy used for it could be more profitable in other areas.
For MARA, Peter Thiel’s recent statements highlight the company’s changing plans. They’re working to combine their existing Bitcoin mining business with new possibilities in AI technology. What happens next for MARA will depend on things like how profitable Bitcoin remains, how much demand there is for AI, their ability to raise money, the agreements they make with customers, and whether they can secure enough power.
As I’ve been researching the crypto mining industry, a key point Fred Thiel, CEO of MARA Holdings, brought up really resonated with me. Ultimately, it all comes down to how much profit you get for each unit of electricity used. For mining companies that have access to cheap and plentiful energy, figuring this out is becoming absolutely crucial for their long-term success. It’s shifting from just mining to really focusing on energy efficiency and profitability.
2026-07-26 15:38