Leading Before the Map Exists
Summary: Traditional investment models assume a degree of certainty that AI no longer affords. CFOs and business leaders need a different approach - one that balances financial discipline with the value of learning as technology evolves. Turning Point’s own experimentation with AI has shown that even when something you invest in today is obsolete tomorrow, the learning carries forward.
Two years ago, a complex valuation and analysis might have taken our team days to complete. Today, one of our analysts finished the work in hours. A multiyear, $1 billion cash reconciliation that once required a team of people and several weeks can now be produced by one person in a few days. A Claude skill can analyze hundreds of documents and uncover something that was buried deep that no one noticed. We are using software to write code in software.
Everyone I talk to sees the opportunity. But the harder question is: where is this going? I don't think anyone knows.
Just in the last few weeks, we have been reading about AI systems attacking other companies’ AI systems. Wait, it can do that? The AI tools we use can't do that. It makes me wonder how far ahead the tools we don't have access to already are, and how much we still don’t know. Understanding where all of this is going may be one of the most important leadership challenges of the next decade.
Business leaders are accustomed to uncertainty. Markets change. Competitors emerge. Economies cycle. A pandemic happens. Plans rarely survive contact with reality exactly as written. We know forecasts are wrong the moment we complete them. Entire management disciplines and strategic planning processes have been built around navigating that uncertainty.
But the current pace of change is challenging our ability to understand what’s happening in real time. Capabilities that seemed years away arrive in months. New competitors are emerging with what appears to be fewer people and less capital. Jobs are beginning to change before companies have had time to decide how those jobs should change.
For leaders, the natural response is to want more information. We want the study. The strategy. The business case. The three-year plan. We want to understand the destination before fully committing ourselves to the journey.
There is just one problem. The map does not exist yet. Waiting for someone to draw it may be the riskiest strategy of all.
For CFOs, this creates a particular tension. Finance has traditionally been the function that brings certainty to the organization. We forecast. We measure. We model. We establish controls. We help develop the business case and challenge the assumptions behind it.
There is a very rational reason to hesitate. Every significant AI investment comes with the uncomfortable possibility that the next model, tool, or application could make part of that investment obsolete. Something we spend a year implementing could be available as a feature in somebody else's software six months later. A capability that seems proprietary today may become a commodity tomorrow. That makes the traditional investment case difficult to build.
Organizations investing in AI are not simply buying technology. They are learning how to use it. They are discovering where it creates value, where it disappoints, how employees respond, and which processes are worth rethinking entirely. They are building judgment about a technology that none of us fully understands yet.
We experienced this firsthand at Turning Point. In 2025, we invested more than $200,000 developing AI agents. We were proud of them. We used them internally and took them into the marketplace.
By March 2026, they were obsolete.
It would be easy to look at that investment and conclude we got it wrong. I don't see it that way. The technology may have become obsolete, but the learning did not.
What we gained from building and using those agents allowed us to adopt the next generation of AI models and tools much faster than we otherwise could have. We understood more about what AI could do, where it struggled, and how our people could actually use it. That experience changed how I think about ROI for AI.
In 2026, we are doing it again, and in 2027, we plan to double down on our AI strategy. A pilot may never scale. A tool may be replaced. A model may become outdated. The organization can still come out of that experience better prepared for whatever comes next.
The question is no longer just whether a particular application will generate a financial return. We should also ask: What might we learn after making this investment that we don’t know today? In some cases, the learning may be the only return on an early investment.
Financial discipline still matters. The goal is to make investments small enough to survive being wrong and meaningful enough to teach us something. Put real people against real problems, measure what happens, and invest more aggressively in what works.
People inside our organizations know change is coming. They are using AI at work and at home, reading the headlines, and watching technology perform tasks that look remarkably similar to parts of their jobs. What remains unclear is exactly what that change will look like. Pretending to have the answers does not create confidence. It erodes it.
Here is the confidence I believe we can offer. We can be clear about what we know. We can be equally clear about what we do not know. Most importantly, we can be clear about what we are going to do next.
That requires curiosity without chasing every new idea. Urgency without panic. Discipline without rigidity. Conviction to make a decision, and humility to admit when we are wrong and change course as the evidence changes.
Leadership has never really been about predicting the future perfectly. It is about moving an organization forward with confidence even when the future is unclear.
