The tricky thing about analytics is that, it doesn’t really generate insights, and insights are largely irrelevant to the goal of analytics.
Insight is a human diet, cooked with human judgement.
The goal of analytics is to support decision making. If the analytics doesn’t support decision making, then it doesn’t have to be.
When the decision maker is a human, she derives insights from analytics, by creating models, algorithms, and, most importantly, points of view. The only reason she needs insights is that otherwise she wouldn’t be able to synthesize what she knows and make decisions like any human does. Human’s ability to synthesize goes beyond logic. While analytics offers only logic, because it’s a human invention based on and only on logic.
Humans need insights to understand and fathom; machines don’t.
When the decision maker is an artificially intelligent machine, insights are irrelevant as long as it can yield the right decisions in a probabilistic manner. Insights only become relevant to analytics when humans intervene and curate the machines. Machines rely on counterintuitive patterns and rules, and they stick to a probabilistic worldview as humans designed them to be.
The Turing Test for analytics is about whether and how much it helps on decision making, not about how many insights it generates. In fact, it doesn’t generate insights. Human does.