Vitalik Buterin (비탈릭 부테린) pointed to a fundamental difference in the cost structures of bitcoin mining and AI inference. U.Today, a blockchain media outlet, reported on Oct. 17 that both fields depend on computing resources, but the way costs arise differs.

Buterin explained that the cost of bitcoin mining is split almost evenly between investment in computing hardware and electricity consumption. Mining companies must spend heavily on specialized equipment while also bearing power bills to run their facilities.

That makes both equipment performance and a stable power supply important in bitcoin mining. Because miners compete for block rewards, electricity prices also account for a large share of production costs.

Buterin saw AI inference costs as mostly concentrated in computing hardware investment. He described AI inference as the process in which a trained AI model generates responses or performs tasks, and said it does not rely heavily on high electricity consumption in the way bitcoin mining does.

In the end, bitcoin mining must manage two cost pillars: equipment and electricity. For AI inference, the cost of securing and maintaining computing infrastructure capable of handling large-scale requests becomes the core of economic viability.