Big Data and Process Optimization – The end of many jobs?

Whatever can be done automatically, doesn’t need people, which was the premise on which the IT industry is built. So would it be a surprise if Big Data actually will cost jobs? From the bottom up, we’re seeing a new level of automation in IT that may even break into the business and start to automate business processes in new ways. Let me explain. We all know that IT infrastructure is increasingly self-regulating. Cloud computing platforms are managed by a fraction of the people that traditional platforms needed. Databases are increasingly self optimized and even things like network-configurations can now be done in software instead of in hardware. Up to now, all this automation was for the infrastructure (the execution of the process), the invisible stuff. What will happen if ‘Big Data’ leads to some kind of commodity solution: a system where you can plug in many data sources and ask the system to find the patterns, the correlations, the predictions? Initially, there will be a lot of work for the data scientists to sort through all the false positives, to help ‘ready’ the data etc. Science fiction? Not in this case: read the interesting narrative behind Siri, before it was incorporated by Apple. If Big Data really is embraced by an organization, would it enable you to start automate the content, the design of the process and the visible services that are ultimately delivered to your customer? Could you let the software decide on the risk or profitability of a financial product for a specific customer? We already let software decide where to place an ad or a link to get maximum results, and pricing of airline tickets, hotels and many other services and goods are by now set by software rather than people. Or could we go one step further and let the software design the content of the ad? (This is where spam email senders are pioneering btw). Where will this end? I’m not one for utopian views where technology seamlessly replaces all human beings, but I have a hard time finding the clear line where automation would have to stop. Suggestions anyone?