FTX’s crypto mining venture is now used to advance AI

We are seeing a strange kind of digital reincarnation. The carcass of the FTX estate is still being picked over by lawyers and liquidators, but one of its largest and most controversial bets is currently being rebranded for the artificial intelligence boom. Genesis Digital Assets (GDA), a company that once represented Sam Bankman-Fried’s massive gamble on the future of Bitcoin mining, is now repositioning itself as a powerhouse for AI data centers.

The Pivot from Proof-of-Work to LLMs

Before the collapse in late 2022, FTX funneled roughly $1.1 billion into GDA. At the time, it was one of the largest venture investments in the history of the crypto mining space. The goal was simple: dominate the hash rate. But the economics of mining have changed, and the demands of the AI industry have created a vacuum that traditional data centers can't fill fast enough. GDA isn't just mining blocks anymore; they are pivoting their infrastructure to support the massive compute requirements of large language models.

For builders, this is a lesson in infrastructure flexibility. GDA’s ability to shift from ASIC-heavy crypto mining to high-performance computing (HPC) for AI isn't just a pivot of convenience. It’s a survival strategy. The power contracts and physical cooling systems that were built for Bitcoin are being repurposed to house the GPUs that drive modern AI. This shift tells us that the real value in the "crypto-mining" sector was never just the coins—it was the access to the grid.

The Estate Value and the AI Premium

The timing here is important for anyone tracking the FTX recovery process. The estate's 20% stake in GDA was once considered a distressed asset that might be sold for pennies on the dollar during the bankruptcy fire sale. However, because AI companies are currently desperate for data center space, that stake has become a significant piece of leverage for creditors.

We are seeing a trend where "AI washing" is becoming the standard for any company with a power plug and a cooling fan. But in the case of GDA, the transition is more than just marketing. The infrastructure requirements for AI training and Bitcoin mining share about 70% of the same DNA: high-density power, advanced thermal management, and locations with low-cost energy. By moving toward AI, GDA is effectively de-risking its portfolio against the volatility of Bitcoin prices and the four-year halving cycles.

Why Founders Should Care

If you are building in the decentralized compute space or the AI agent sector, this shift matters for three reasons:

• Capacity Scarcity: The fact that a major mining firm is pivoting to AI suggests that the supply of tier-3 and tier-4 data center space is at a critical low.
• Energy Sovereignty: We are entering an era where energy is the primary currency. GDA’s value isn't in its brand; it's in its power purchase agreements.
• The Infrastructure Bridge: The wall between "Crypto" and "AI" is disappearing at the hardware level. The same physical buildings are now hosting both types of compute.

The Skeptical Take on Rebranding

As a founder, I’ve seen this movie before. When the blockchain bubble was at its peak, every pivot was toward "decentralization." Now, every pivot is toward AI. We have to ask if these facilities are actually ready for the specific requirements of AI workloads. Crypto mining is relatively forgiving—if a miner goes down, you lose a few minutes of uptime. AI training is different. If a cluster of GPUs fails during a training run, you can lose weeks of progress and millions of dollars in compute time.

Repurposing a mining farm into an AI data center isn't as simple as swapping out an Antminer for an H100. It requires massive upgrades to fiber connectivity and redundancy. While GDA is talking a big game about AI, the real test will be whether they can attract enterprise-grade AI clients who demand 99.99% uptime, something that mining operations rarely prioritize.

The real value in the crypto-mining sector was never just the coins—it was the access to the grid.

What This Means for the Future

The FTX bankruptcy has inadvertently created a massive venture fund for the AI revolution. By holding onto these assets while the AI market exploded, the liquidators have found themselves in a much better position than they were a year ago. It’s a bit of irony that Bankman-Fried’s reckless spending might actually be bailed out by a technology he didn't even see coming at this scale.

For those of us on the ground building, the takeaway is clear: stop thinking about crypto and AI as separate silos. They are both just different ways of consuming electricity to generate digital value. If you control the power and the cooling, you control the bottleneck of the next decade. GDA is showing that even the most toxic assets can be cleaned up if they are sitting on enough kilowatts.

The Bottom Line

Don't be distracted by the FTX name or the bankruptcy drama. Focus on the physical layer. The pivot of Genesis Digital Assets toward AI is a signal that the mining industry is maturing into a general-purpose compute industry. The companies that survive won't be the ones who can guess the price of Bitcoin, but the ones who can manage the infrastructure for whatever the world decides to compute next.

Read the original at Cointelegraph →