In the 1990s, the HTTP request became the atomic unit of web commerce. Every search query, every page view, every transaction reduced to a standardized request-response pattern. This primitive enabled Google, Amazon, and the entire web economy.
In the 2020s, the AI prompt is becoming the atomic unit of a new economy.
The Historical Parallel framework reveals why this matters: understanding the primitive means understanding the economy it enables.
What HTTP Enabled:
- Search engines (queries as HTTP requests)
- E-commerce (transactions as HTTP requests)
- Social media (interactions as HTTP requests)
- The entire ad-supported web economy
What the Prompt Enables:
- AI Search (natural language queries)
- AI Assistants (conversational interactions)
- Code Generation (prompts to programs)
- Content Creation (prompts to media)
The Critical Difference:
HTTP requests are computationally trivial—fractions of a cent, negligible energy. AI prompts are computationally intensive—0.24 Wh of energy, measurable carbon, real infrastructure requirements.
This resource intensity changes everything. The HTTP economy could scale infinitely with minimal marginal cost. The prompt economy scales with significant marginal cost—energy, water, chips, data centers.
Strategic Insight:
Companies that harness AI profitably will be those that understand this primitive’s true cost structure. Those that treat prompts like HTTP requests—essentially free at the margin—will find their business models broken by infrastructure economics.
This analysis applies The Business Engineer’s framework on economic primitives to understand AI’s foundational economics. Read the full analysis: The Economics of an AI Prompt →









