There are many things that people can do quickly that smart machines cannot.  Natural language is beyond deep learning and new situations baffle artificial intelligences.  The dominant technique in artificial intelligence, will not lead to an AI that abstractly reasons and generalizes about the world.  deep learning is greedy, brittle, opaque, and shallow.  The systems are greedy because they demand huge sets of training data.  Brittle because when a neural net is given a “transfer test”—confronted with scenarios that differ from the examples used in training—it cannot contextualize the situation and frequently breaks.  They are opaque because, unlike traditional programs with their formal, debuggable code, the parameters of neural networks can only be interpreted in terms of their weights within a mathematical geography.  Consequently, they are black boxes, whose outputs cannot be explained, raising doubts about their reliability and biases.  Finally, they are shallow because they are programmed with little innate knowledge and possess no common sense about the world or human psychology.  None of these shortcomings is likely to be solved soon.


Lightbulb Moment: FINALLY, someone said it out loud!  While AI is seductive, it cannot see or foretell a trend’s birth, see or predict an upcoming change in direction, explain the drivers behind a trend’s movement, anticipate obstacles, or tell the difference between a short-lived and long-lived trend.  In short, AI is useful for tracking trends but useless for strategic forecasting due to its linear voice and lack of common sense.  It can be used as one of many tools in a much larger story to help see a trends true behavior, but alone it is misleading and vain.