Scientists have not always had freely accessible high-quality and high-resolution datasets relevant to their study systems. Today, early career researchers routinely confront a deluge of data that is relevant to their research questions. Young scientists face the combined challenges of using accessible yet powerful models, under high publication pressure, and with mixed guidance from scientists trained under an earlier era. There exists a temptation to reach for black-box analytical approaches to offer guidance through this wilderness of data. New complex models consisting of artificial intelligence and machine learning tools are poised to be co-opted by large numbers of early career researchers due to their modelling strength and easy, out-of-the-box usage. Just because we can use these new tools, does not mean we always should. I argue we should reconsider the role of complexity in the construction of our ecological models when we test ideas of our understanding of the natural world.