OpenAI's CFO Reveals New Metric for Measuring AI Value
· news
The AI Equation: When Economic Value Trumps Cost-Centered Thinking
OpenAI’s CFO, Sarah Friar, has sparked a critical conversation in the business world with her proposal for a new metric: “useful intelligence per dollar.” This approach shifts focus from adoption rates and cost-cutting measures to the actual productivity gains achieved by AI investments.
Friar’s idea is not just about redefining how we measure AI’s value; it also highlights a fundamental shift in the role of CFOs within organizations. They are no longer content with simply managing the bottom line but are now being asked to help shape strategy and determine where companies place their biggest long-term bets.
This trend is not unique to OpenAI or even the tech industry as a whole. According to McKinsey’s Andy West, CFOs are increasingly expected to take on a more strategic role within organizations. His informal poll of finance chiefs at the 24th annual Global CFO Forum revealed that nearly two-thirds believe the strategy function now reports to them – up from less than a third just five years ago.
As companies move beyond simple cost-cutting measures and toward more strategic investments in AI, they will need to rethink their entire approach to finance and accounting. This means developing new metrics that can capture the complex value created by AI systems, going beyond mere cost per token or adoption rates.
The Stargate initiative announced by OpenAI has already surpassed its initial milestone, with plans to invest up to $500 billion over four years. This is not just a technology expense; it’s a strategic asset that will drive the company’s growth and innovation in the years to come.
As other companies follow OpenAI’s lead into large-scale AI infrastructure, they will need to ask themselves tough questions about the role of finance leaders within their organizations. Are these executives equipped to take on this new level of responsibility, or will they struggle to adapt to their changing roles? And what does this mean for the future of business strategy as a whole?
The answer lies in how we choose to interpret the numbers. By focusing on the work that AI completes, rather than just its costs, companies can finally begin to see whether their investments are paying off. It’s time to rethink our approach to finance and accounting – and to start valuing AI for what it truly is: a strategic asset driving growth and innovation in the years to come.
The clock is ticking on OpenAI’s IPO, but the real story here is not about valuation or financials – it’s about how businesses can finally begin to see the real value in their AI investments. As we move into this new era of AI-driven business strategy, one thing is clear: companies that fail to adapt will be left behind.
Reader Views
- CMColumnist M. Reid · opinion columnist
While OpenAI's CFO, Sarah Friar, is right to focus on actual productivity gains from AI investments, her proposed metric, "useful intelligence per dollar," glosses over a crucial issue: data quality and bias. As companies pour billions into AI infrastructure, they'll need to ensure the underlying datasets are trustworthy and representative of diverse perspectives. Otherwise, even the most advanced algorithms will spit out garbage results, rendering the entire exercise futile. CFOs must now navigate not only new metrics but also the delicate art of managing data integrity in an era where AI is driving business decisions.
- RJReporter J. Avery · staff reporter
While Sarah Friar's proposal for "useful intelligence per dollar" is a step in the right direction, it glosses over the elephant in the room: data quality and governance. As AI investments scale, so does the risk of biased or low-quality training data. Companies would do well to prioritize data hygiene alongside this new metric, lest they be chasing productivity gains on shaky foundations. Only by addressing these underlying issues can we truly unlock the potential of AI to drive business value.
- CSCorrespondent S. Tan · field correspondent
The emphasis on "useful intelligence per dollar" is a step in the right direction for CFOs who need to justify massive investments in AI infrastructure. However, we can't ignore the elephant in the room: data quality and governance. Without proper safeguards in place, AI-driven decision-making could exacerbate existing biases and errors, rendering this new metric meaningless. Companies must develop robust data standards and accountability measures alongside their AI strategies if they hope to unlock true value from these investments.