🌀 Complex systems are hard to predict

My wife makes sourdough, which is a multi-day process and at times unpredictable. Recently, the current batch was a little sluggish and not expanding like it should, so she "fed" it and left it on the counter. It took off during the night and was all over the counter in the morning.

Sourdough is a complex living system. Because of this, sourdough is difficult to predict 100% of the time. As computer systems become more complex, they take on some of the same characteristics of living systems. You can't completely predict everything. What worked yesterday may not work today, due to various external factors that cannot be controlled or fully understood. Especially with networked systems -- we cannot control everything else on the network, or involved with the network. We cannot control what new feature customers are going to ask for that might require us to update dependencies. We can't predict what new edge case will trigger a bug we have not seen before.

The best we can do is keep tools and processes flexible, up-to-date, and as automated as possible (YOUR Platform). This enhances the ability to adapt to the external forces that cannot be controlled.