A scorecard for the AI age

TL;DR - OpenAI CFO Sarah Friar introduces a practical "AI scorecard" framework for measuring the return on investment of AI systems. It matters because it offers business and technical leaders concrete metrics to evaluate whether AI deployments deliver real value beyond hype.

• Proposes measuring AI value by useful work delivered rather than raw model benchmarks.
• Introduces cost per successful task as a unit-economics metric for AI workloads.
• Emphasizes dependability (reliability/consistency of outputs) as a key evaluation axis.
• Frames return on compute as the way to justify AI spend against infrastructure costs.

Note: Content is thin (a one-sentence summary), so these bullets reflect the four scorecard dimensions named rather than detailed methodology.