Bezos' Law
Updated 2026-08-08
INTRODUCTION
English translation pending.
CORE DEFINITION
Bezos' Law states that the cost of a unit of cloud compute falls by about fifty percent every three years, a pattern named after Amazon founder Jeff Bezos and observed since the launch of Amazon Web Services. It complements Moore's Law, which describes transistor counts doubling rather than prices falling, by focusing on economics rather than engineering. The trend implies that any business whose costs scale with compute, such as machine learning, faces steadily shrinking marginal costs. Key qualification: the decline applies to compute specifically, while bandwidth, storage, and power costs fall more slowly or even rise.
SCAFFOLDING EFFECT
Reduce cognitive load
- Project cost curves: assume compute-heavy features get cheaper and plan products on that basis. - Time your investment: validate demand before scaling, since waiting buys more capacity per dollar. - Find non-compute bottlenecks: identify the constraints that will not deflate on the same schedule.
Anchor fast decisions
Cloud providers spread enormous fixed investments across growing demand while competing on price, and each generation of hardware delivers more compute per dollar of capital and power. Those gains pass through to customers because the market is competitive and switching costs are low for standard instances. The result is a predictable deflationary trend that shifts the economics of anything whose cost is dominated by computation.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- baike.baidu.comhttps://baike.baidu.com/item/%E8%B4%9D%E7%B4%A2%E6%96%AF%E5%AE%9A%E5%BE%8Bverified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS