Realize that the computer's advantages can also lead to weaknesses. By this, I mean that a computer's powerful and precise memory means that it is better able to work off raw correlations, without as much of a need to abstract or seek out causal models. While this might turn out okay for detailed large but ultimately simple (stationary) patterns, it will not be so advantageous in more dynamic settings or scenarios with multiple levels of organization with differing dynamics.