Teams and organisations grow on incremental improvement. That’s a given.
The best build stable foundations through prudent talent identification and recruitment, then drive compounding improvement through a feedback loop—built by coaches, fed by players, through performance and outcomes. When built and used effectively, it’s a powerful mechanism.
AI does something similar. Retrieval-Augmented Generation (RAG) improves large language models (LLM) by allowing them to search an external knowledge base before answering. When an LLM remembers and learns from its own outputs and past interactions, it engages in continual learning—compounding gains. The feedback loop, just faster.
In my /now space, we’re building an app that looks beyond the page, in our case, a cricket scorecard—using data to validate or disprove the subjective view. We hope that PlayStats will be the feedback loop for recreational and professional cricket.
But, here’s the danger.
Pat Riley flagged it in an earlier post, when he talked about the the disease of more. When the foundation shifts, for any number of reasons, all of that incremental improvement breaks.
Hard to build, easy to bust.
If you need the benefits of compounding gains, do the work, build the feedback loop, but don’t forget to stress-test and maintain the foundations.
Nick
P.S. Happy first day of Spring from this corner of the world. With a little more time, I might have swapped this post for something more appropriately spring-like :)