Vitalik Buterin: Bitcoin Mining and AI Inference Differ in Cost Structure

Vitalik Buterin has pointed out that the cost structure of Bitcoin mining fundamentally differs from that of AI inference. According to a report by blockchain media U.Today on July 17 (local time), both fields rely on computational resources, but the way costs are incurred varies significantly.
Buterin explained that Bitcoin mining costs are nearly equally split between computing hardware investment and power consumption. Mining companies must invest heavily in specialized equipment while also bearing electricity costs for operating their facilities.
This means stable power supply and hardware performance are both critical in Bitcoin mining, as the block reward competition makes electricity prices a major component of production costs.
In contrast, Buterin noted that AI inference costs are predominantly focused on computing hardware investment. AI inference involves generating responses or performing tasks using trained models, which does not rely heavily on high power consumption like Bitcoin mining.
Ultimately, Bitcoin mining requires managing both equipment and electricity as dual cost factors, while AI inference centers economic efficiency around securing and maintaining large-scale computing infrastructure capable of handling high volumes of requests.
Korean Source
This article is an English localization of a Korean-language crypto news report. Original headline: 비탈릭 부테린 "비트코인 채굴과 AI 추론의 차이는 비용 구조"