In my life experience, speed usually wins. There seems to be two general approaches in life, whether it be business, technology, or even sporting events. You can be careful and steady or you can be fast and more reckless. This is note a fast beats slow observation, but I can certainly say with absolutely certainty that speed usually wins. The question is why?
The system that can act, receive feedback, correct errors, and act again faster tends to outperform the system that tries to avoid errors in the first place.
I think there are several reasons. One of which concerns the speed of the feedback loop for errors. John Boyd, a military strategist, created his own feedback loop mechanism called Observe, Orient, Decide, Act, and then Observe Again (OODA). I learned about this in a leadership training series at West Point a few years ago. The key point he made was less about the loop itself and more about how quickly an organization learns to do it. Boyd argued that competitive advantage comes from cycling through this loop faster than an opponent. It isn’t necessarily the side with the best original decision that wins. It’s the side that can Decide → Act → Learn → Adjust faster.
This pattern repeats everywhere and whenever you have a loop. In my world (Building AI Systems), the concept is Train → Evaluate → Adjust → Train. Same concept, the structure is the same, and the outcome is predictable. The faster you can do this, the more likely you are to get better results.
There is a well-known speed–accuracy tradeoff: moving faster generally increases the number of mistakes. The slower competitor initially looks superior because they make fewer errors. But this becomes misleading when errors are recoverable. When I was young, I was in college studying contention algorithms for communication systems. Essentially, how do machines negotiate for a shared medium? Well, there were many methods, but ultimately Ethernet won because, even though it made more mistakes, it was faster and other protocols could easily recover from them. Thus, the marriage of TCP/IP and Ethernet became the standard for modern business networks. Interesting that only a few decades ago, computer scientists thought the opposite.
Having said all of this, the single contributing and crowning factor in speed’s favor is that of compunded learning. Speed creates more learning cycles per unit of time. Suppose two organizations start at roughly the same level.
Company A improves 10% every year.
Company B improves only 3% each iteration but runs an iteration every month.
The individual improvement might be smaller and messier, but the repeated learning begins to compound.
Iteration speed overwhelms initial precision.
So, while the allure of the perfectly crafted, meticulously planned approach is strong, there’s a compelling argument for embracing the iterative, slightly imperfect, but undeniably swift path. It’s not about recklessness, but about recognizing that in many endeavors, the ability to move, learn, and adapt quickly often creates a more robust and ultimately more successful outcome. Perhaps the real art lies in discerning when to lean into that speed, trusting that the compounding effect of rapid learning will, in time, refine the journey.
