I just saw something interesting at the latest conferences on cryptocurrencies and artificial intelligence. It turns out that the real point of convergence is not what many expected. Chappy Asel, who worked on Apple’s Vision Pro and later launched The AI Collective, presents a practical perspective on how crypto and AI will truly intersect.

Most people think of chatbots when mentioning AI, but the argument gaining traction is completely different. Autonomous software agents will need payment systems that operate 24/7, with very small transactions and virtually zero latency. This is not science fiction: stablecoins already enable continuous settlement, and smart contracts offer programmable execution. It’s the only logical way for machines to trade with each other without human intervention.

The problem is that theory is moving faster than reality. AI agents are still in early stages, and most companies continue to use centralized APIs and conventional payment systems. There have been attempts to build infrastructure for these autonomous payments, but commercial activity remains minimal. The narrative is developing faster than actual demand.

But Asel has another angle that’s probably more immediate. He says that the true overlap between crypto and artificial intelligence could come from physical infrastructure: computing power, data centers, and energy. It’s not a technical problem of insufficient models; it’s that computational resources are literally what’s limiting decision-making in AI right now.

There’s already movement in the crypto industry on this. Bitcoin miners have spent the past year repositioning toward AI hosting and high-performance computing. The infrastructure they built for mining can be repurposed for AI workloads. It’s an interesting strategic pivot.

From a practical perspective, the advice is to experiment. When there’s so much uncertainty, the only sensible thing is to try new technologies. And here’s the interesting part: the historic adoption problem in crypto has always been usability. But autonomous agents don’t need tutorials, aren’t scared by complex wallets, and don’t need help with seed phrases. If autonomous software becomes a real economic actor, artificial intelligence could be the user base crypto really needed: users who think in code.

Meanwhile, moves like a Bitcoin whale transferring around 40 billion BTC on Sunday show that significant capital is moving in the space. The move was detected without being associated with any known exchange, leaving speculation open about intentions. These movements always spark conversations about where capital is headed in crypto.