Bitcoin Mining Unprofitable: AI Offers 10-15x Returns, Says MARA Founder
“Bitcoin mining profits are getting squeezed, and one of the industry’s biggest names thinks AI may be a better use of the electricity.” MARA’s founder is openly discussing a move away from Bitcoin mining. That is not a minor operational tweak. If the company follows through, money, power, and computing equipment now committed to BTC mining could support AI workloads instead. Mining stocks would become harder to assess, while the old fight over Bitcoin’s energy consumption would flare up again. My take: the electricity matters more than the branding.
“Fred, MARA’s founder, says Bitcoin mining is no longer profitable and the company is moving into AI.” In a recent interview translated into Russian and shared by Crypto Headlines, Fred puts it bluntly: “Bitcoin mining is no longer profitable. We are moving into AI.” His headline claim is even sharper: AI can earn 10 to 15 times more than BTC mining from the same electron. Sounds decisive, right? Not yet. Investors still need MARA’s costs and projected AI revenue before treating that comparison as proven. Fred also points to what he considers a weakness in Bitcoin as an asset: it does not produce a return. I’ll be honest: that criticism lands differently when it comes from the person credited here with building the world’s largest mining company.
“MARA’s new direction could pull money and computing capacity out of Bitcoin mining.” Capital follows returns. If an AI data center can generate 10 to 15 times the revenue from the same electricity, miners have a strong reason to reconsider where the next dollar goes. Most commentary makes that choice sound automatic. That is only half right. An AI conversion may require different chips and faster network connections. It can also mean new cooling systems, plus customers prepared to sign contracts. The spreadsheet is easy. The conversion is not.
If MARA and other big operators redirect their infrastructure, Bitcoin’s total hash rate could fall. That would not automatically pull down BTC’s price. A prolonged decline, however, might indicate that miners are losing confidence or shutting off machines, and markets could react badly if it coincides with a broader retreat from risky assets. We saw a similar mood in 2022: as the Federal Reserve raised interest rates, BTC fell from more than $48,000 in March to less than $16,000 by November. The cause of an AI shift would be different. The market experience could still feel familiar—capital leaving crypto infrastructure because another corner of technology pays better.
“Fred’s claim that Bitcoin ‘does not generate returns’ raises an uncomfortable question about why institutions own it.” BTC supporters call it digital gold or a store of value. Others frame it as a hedge against inflation. None of those arguments requires Bitcoin to pay interest. So what is Fred really asking? What the capital earns while an investor waits. That is a narrower question, but arguably a tougher one.
The debate itself is old. Its source is not. A corporate treasury comparing Bitcoin with income-producing assets may hesitate over BTC’s lack of yield. MicroStrategy, for example, has accumulated more than 214,000 BTC, according to public filings cited in the original account. The company is betting on long-term price growth, not cash flow from the coins. When Bitcoin rises, that can look brilliant. In a long downturn? Much harder to defend. I would not pretend those two conditions deserve the same analysis.
If companies start favoring assets that produce income, future corporate BTC purchases could slow. That shift may also sharpen comparisons with cryptocurrencies such as ETH, whose holders can earn staking rewards. Counter to the usual yield-first argument, though, the asset with a payout is not automatically the more useful one. Bitcoin and Ethereum work differently, so yield alone cannot settle the comparison. Investors still notice what pays them. One interview will not rewrite Bitcoin’s story overnight, but criticism from inside the mining industry stings more than the same line from a career skeptic. That part is hard to wave away.
What this means
“The mining business may be approaching a point where electricity earns more from AI customers than from producing Bitcoin.” Fred’s estimate is specific: AI offers 10 to 15 times the earnings for the same energy input. Why does this matter? Because electricity is the shared constraint in both businesses. If MARA’s results support his estimate, miners may steer new investment toward AI facilities and reduce spending on proof-of-work expansion. My take: reported allocation will matter more than another bullish interview.
That would change how investors price mining stocks such as MARA. These companies have often traded like leveraged bets on Bitcoin’s price and the economics of mining it. Add a substantial AI operation and the model gets messier fast. Investors would need to weigh data center contracts and conversion expenses against electricity prices. Mining difficulty and BTC output remain in the equation too. Yes, that cuts against the tempting “diversification makes valuation safer” story. More revenue sources can also create more ways to misread the company. Anyone offering a simple valuation may be selling more confidence than clarity.
Bitcoin’s security depends on miners continuing to provide computing power. A flat or falling hash rate would therefore be worth watching, especially if several large operators announce AI conversions at roughly the same time. Is one weak reading enough? No. More efficient hardware can move the figure, as can changing electricity costs. Routine adjustments in mining difficulty matter as well. A sustained decline carries more information than a handful of soft readings.
“The next useful evidence will come from company accounts, not interview clips.” Upcoming reports from MARA, RIOT, and other large miners should reveal more about their AI plans. I would go straight to the actual numbers: capital spending and power capacity allocated to AI; then conversion costs, signed contracts, and revenue forecasts. Earnings calls should clarify whether Fred described a funded project or merely an option MARA is still considering. Skip the guesswork.
The energy debate could move with the industry. If miners switch to AI, Bitcoin may consume less electricity, but the underlying demand will not disappear. AI data centers use plenty of power too. Most guides frame this as a crypto-energy story. That framing is too narrow. Regulators may simply redirect their attention from crypto mines to AI facilities, particularly where both industries are competing for limited grid capacity. Same grid. New argument.
Technical analysts are also watching BTC’s $60,000 support level. A break below it could deepen concerns about market weakness, particularly if mining companies cut spending or announce conversions at the same time. Still, I would not connect one price level too tightly to a company’s business plans. Bitcoin has broken supposedly decisive levels before and then reversed as though the line had never existed. Charts help. They do not run MARA.
FAQ
Q: Who is Fred, and why do his comments matter?
A: Fred is identified here as MARA’s founder, and MARA is described as the world’s largest Bitcoin mining company. His comments matter because they come from direct experience running a major mining operation, not from an outside critic.
Q: What does Fred say about Bitcoin mining?
A: He says it is no longer profitable and claims AI can earn 10 to 15 times more from the same electricity. Bold claim. MARA’s financial reports should show whether the comparison survives contact with actual costs and revenue.
Q: What is his criticism of Bitcoin as an asset?
A: Fred says Bitcoin has a basic problem because it does not produce returns. Holders profit when its price rises rather than through interest or another form of yield.
Q: Could a move into AI hurt the Bitcoin network?
A: It could reduce Bitcoin’s hash rate if several major miners redirect substantial computing capacity. One short dip would reveal little. A prolonged decline might show that miners are pulling back.
Q: What should investors watch now?
A: Start with company filings and earnings calls. They should show whether miners are committing real money and power capacity to AI. Bitcoin’s hash rate deserves attention as well, along with new energy rules affecting crypto mines and AI data centers.