The key measure of leverage for AI-powered developers is no longer GPU utilization (FLOPs) but the volume of tokens processed by agents. Karpathy feels nervous when his token subscriptions are underutilized, indicating he's the bottleneck, not the system.

Instead of focusing on headcount reduction, Goldman's CIO measures the success of developer AI tools by their ability to consistently help projects finish ahead of schedule. This provides a tangible metric for increased output and organizational capacity.

If your team cannot articulate the specific business outcome of their AI usage in a single sentence, you don't have an AI strategy. You simply have 'token maxing'—usage for the sake of usage. This framework forces a direct link between AI spend and business results.

As AI handles more routine tasks, traditional productivity metrics like 'tasks completed' become obsolete. The focus must shift from output to outcomes. It no longer matters what was done on a given day, but rather how tools were used to achieve a specific business goal.

Demanding a direct, line-item ROI for foundational AI initiatives is like asking for the ROI on Wi-Fi—it's the wrong question. Instead of getting bogged down in impossible calculations, leaders should focus on measuring the business outcomes enabled by the technology, such as innovation speed or new product creation. Obsess on outcomes, not direct financial return.

The AI market has cleared its first ROI hurdle: model revenue has justified massive infrastructure investment. Now it faces a second, harder test. Enterprises spending billions on AI tokens must demonstrate tangible financial benefits, like higher margins or revenue, to sustain the flywheel.

Companies struggle to measure AI's return on investment because its value often materializes as individual productivity gains for employees. These personal efficiencies, like finishing work earlier, don't show up on corporate dashboards, creating a mismatch between perceived value and actual impact.

C-suite conversations have evolved from encouraging broad AI experimentation to demanding measurable ROI. The critical mindset shift is away from fascination with specific models and toward redesigning core, enterprise-grade workflows for tangible business impact, moving from a 'playground' to 'production grade' mode.

The trend is shifting from simply adopting AI to proving its ROI with specific metrics. As industry leaders publicly share their AI-driven gains, it creates a competitive necessity for all other companies to follow suit and quantify their own benefits, making it 'table stakes' for all.