Agents alongside humans — what we got wrong twice

When AI models began accelerating in capability, we saw an opportunity—not to replace people, but to help our team spend less time on repetitive work and more time on the parts of their jobs that require creativity, experience, and sound judgment. Like many businesses, we explored the technology with high expectations, believing it could take on large portions of technical work while our team focused on higher-value responsibilities. Our first mistake was expecting too much from the models themselves. AI is remarkably capable at computation and pattern recognition, but it does not reason, understand context, or make thoughtful business decisions the way people do. Left to solve complex problems on its own, it often produced work that was close, but not close enough.

Our next step was to keep people firmly in the loop. We broke larger projects into smaller tasks, reviewed every output, and guided the models through implementation. The quality improved considerably, but we uncovered our second mistake. We were still treating AI as if it were another teammate that needed constant direction. While it completed individual tasks quickly, the continuous prompting, reviewing, and repeated model usage introduced enough time and cost that the overall gains were smaller than we expected. Faster execution on paper did not always translate into a more efficient process.

The breakthrough came when we stopped asking AI to work like a person and started using it according to its strengths. Instead of replacing human thinking, we use AI to build technical tools, automate repetitive computation, process information, and generate structured outputs that our team can confidently review and build upon. Our people remain responsible for strategy, design, critical thinking, customer relationships, and every important decision. AI simply removes friction from the work that does not require uniquely human judgment.

That shift has shaped how we approach every AI project today. We believe the best businesses are not those that replace people with technology, but those that use technology to make people more effective, improve work-life balance, deliver better outcomes for customers, and create sustainable growth for everyone involved. When humans and AI each contribute where they are strongest, everyone benefits.