OpenAI published an essay by Sarah Friar titled “Building abundant intelligence” on July 31, 20261. The company casts its position as a cycle — better intelligence drives broader adoption, broader adoption supports more investment, and more investment improves intelligence and efficiency — and presents the GPT-5.6 price cuts announced on July 30 as an expression of it1.
At the center of the argument is a swap of metrics. Customers do not buy tokens for their own sake, the company says, and the right measure is the cost of a successful outcome, including the time, retries, oversight, and errors required to get there1. On that basis, a stronger model that completes work correctly the first time can end up more economical than a cheaper one that needs repeated attempts or human intervention1.
The essay also cites scale: OpenAI says its models now reach more than one billion active users and more than two million businesses, and that agentic work through Codex accounts for 99.8% of weekly output tokens inside the company1. All of these are the company’s own figures, without third-party verification.
Sources
- Building abundant intelligence - OpenAI, by Sarah Friar (July 31, 2026)