Learning to Work with AI

One of the projects I worked on involved analyzing public SEC data to better understand which wealth management firms are growing, which appear most operationally efficient, and how leadership structures, particularly technology leadership, show up across the industry. On the surface, this sounds like a perfect AI use case: large datasets, pattern recognition, and comparison over time. And in many ways, it was. But the most interesting insights had less to do with AI's capabilities and more to do with what it exposed about data, decision-making, and leadership.

AI Doesn't Eliminate Work. It Redistributes It.

The first thing that became clear is that AI does not magically remove effort. It moves effort. Before AI added any value, I had to spend time cleaning spreadsheets, standardizing fields, reconciling inconsistencies across reporting periods, and deciding what questions actually mattered. AI was incredibly effective once the inputs were thoughtful, but it was unforgiving when they were not.

This led to an important realization: AI amplifies whatever it is given. Good data and clear thinking produce strong results. Messy data and vague questions produce confident-sounding noise. That is not a flaw of AI. It is a mirror.

Once the foundation was in place, the benefits were real and immediate. Tasks that would normally take hours, such as trend identification, growth comparisons, efficiency ratios, and pattern spotting across time, collapsed into minutes. The work shifted from mechanical analysis to interpretation and judgment. That is a meaningful upgrade, but only if you are prepared for it.

The Real Challenge Isn't the Tool. It's the Organization.

This experience also changed how I think about AI adoption inside organizations. The hardest part is not selecting the right tool. It is everything around it. AI forces uncomfortable questions:

• Who actually owns the data?
• Are our definitions consistent?
• Where does judgment live when a model makes a recommendation?
• Do our workflows support learning, or just execution?

In my own experimentation, AI worked best when I treated it as a collaborator that needed clear constraints, not as an oracle. That same principle applies at scale. Organizations that treat AI as a plug-and-play solution will struggle. Organizations that treat it as a capability, one that requires framing, structure, and evaluation, will get far more value.

Leadership Looks Different in an AI Enabled World

One of the biggest takeaways for me is how leadership roles are likely to evolve as AI becomes more embedded in daily work. When information and analysis become cheaper and faster, leaders add less value by controlling access to information and more value by setting context. That means:

• Communicating to the organization why AI is being used, not just how
• Structuring ownership so accountability does not disappear behind algorithms
• Helping teams understand when to trust AI outputs and when to challenge them
• Investing in data quality and shared definitions, even when it is unglamorous

In many ways, AI exposes organizational weaknesses more than it solves them. It highlights where data is inconsistent, where processes are informal, and where decision-making relies too heavily on tribal knowledge. Used thoughtfully, that exposure is a gift.

Data Culture Is the Advantage

If there is one area where I would invest first, it is data culture. AI rewards organizations that treat data as a shared asset rather than a byproduct of work. That means clearer ownership, better documentation, and an expectation that data quality matters, not because AI needs it, but because good decisions do.

AI does not replace human judgment. It raises the bar for it.

Final Thoughts

What started as an exploration of generative AI tools turned into a broader reflection on how work gets done, how decisions are made, and what leadership looks like when analysis is no longer the bottleneck. AI is powerful, but only in environments that are ready to use it well.