I’m excited to be speaking at the Business Forecasting and Analytics Summit in Chicago on September 21. My closing keynote is “The Future of Forecasting” and, in it I’m going to explore one of the biggest trends to impact the field: the relationship between man and machine, or more specifically, intelligence analysts and software packages.
I am bullish about the field of strategic intelligence, and also bullish about the future of jobs in the field. Some people are worried about software and artificial intelligence becoming so powerful that executives will no longer need insights from intelligence analysts. I don’t worry. I don’t think that software will ever threat our jobs because computers never more than hardworking idiots. Human intelligence analysts will be doing the thinking and computers will be doing the computing.
Don’t worry, I think that it looks a little scary at times. Check out this genius British kid who made a “robot lawyer” that can appeal parking tickets. This program has written the appeals on more than two million pounds sterling worth of parking tickets. Also, I am told that AI is just around the corner because “look at the advent of self-driving cars.”
Look, there is no question: Moore’s law, the most famous technological forecast in history, is still in operation, and the power of computer chips continues to grow. Sensor technologies are getting smaller and cheaper, and software grows in sophistication year after year. Also, it is very clear, many of the robots upon which we rely for daily life – cars, clothes dryers, and toasters – all involve increasing levels of “smartness.” They interpret data generated from sensor arrays and use software to arrive at a decision: yes, pump the brakes, no the clothing isn’t dry enough, hey the toast is starting to burn. This enables these robots to do tasks for us with fewer negative outcomes.
But intelligence analysis ain’t making toast
In strategic intelligence and forecasting, we have robots to aid us in our tasks, and they are very handy. Google Alerts can tell me when something new has happened on a topic of interest – let’s say a story about my competitors. Digimind software can tell me how much that story has popped up in different parts of the world and display it on a map. Compelligence can plug into Salesforce and make sure my business development people know about that news if they run into those competitors for a certain opportunity.
But which software is going to tell me why Amazon bought The Washington Post? What algorithm can tell my why Wal-Mart’s acquisition of jet.com is important? Which software package is going to explain to me why our brand needs to appeal to Generation Y more, and how to get there?
For quantitative tasks, nobody has been able to beat computers for a very long time. And for predictive modeling, they clearly have an advantage at treating large sets of data that humans will never approximate, much less duplicate. But qualitative analysis – what something means – requires a human element that computers are nowhere near. There are patterns in business and government that computer software has not – to date – been able to identify, because the job simply requires the quantum computing power of the human mind to interpret. Strategic thought requires art and philosophy as much as it does math and statistics. And when it comes to delivering that information to a human executive with a distinct and usually quirky personality – it’s much more about art. We’re a tricky species, and we’re just scratching the surface of understanding neurologically and psychologically how we really make decisions.
In the years to come, computers will help – a lot! There will be new capabilities, new powers, new speed – and it will still come down to the messy, imprecise work done by professional intelligence analysts that will help leaders.
Either that, or we’re closer to a dystopian world when we are slaves to our robot overlords – but I don’t think so.