If, like me, you are reading a lot about data science, Artificial Intelligence (AI) and machine learning, and trying to figure out what impact they are going to have on the marketing industry, then this year’s TED promised to be an opportunity to get your questions answered.
Sadly what it did for me was to add to my list of questions, rather than reduce them. Don’t misunderstand me, it wasn’t that the speakers did something wrong, it was simply that they illustrated all too clearly how much impact these technologies are going to have on every aspect of business. They showed how it could change the way just about every service industry could be impacted and raised all manner of larger questions about how we exist in a world where the machines can do so much and the need for people is so greatly reduced. That said, I was left with an overwhelming feeling that we are still at a point where the idea of these technologies becoming ‘human enablers’ rather than replacers, is very much on the table.
While no speaker talked to the specific impact AI and machine learning could have on marketing, it wasn’t hard to see the potential for change. The ability of algorithms to spot patterns in customer data that humans would struggle to see, the capability of bots to automatically generate content from data that was specifically targeted to the needs of certain individuals was at the easy end of the spectrum. In the ‘harder to do, but far from impossible’ column was the idea of machines creating simulations of markets so that you could test the way a marketing campaign would work before it actually ran. Indeed it is clear that such an idea won’t be just an idea for long.
One presentation that did stand out for me was by Ray Dalio, the famed head of hedge fund giant Bridgewater Associates. He is using algorithms, alongside a management model he calls ‘radical transparency,’ to ensure the best ideas win out, rather than just the ones his senior team has espoused. If I understood him correctly, he allows the entire team to score ideas and then uses algorithms to determine which ideas should win, not simply on whose idea scored the most, but also based on past scoring accuracy. In other words if you often give high scores to ideas that failed, your scores will start to count less. Such an approach within marketing departments could be revolutionary.
On the cautious side, Cathy O’Neil made the argument for not just assuming that your algorithm was going to give you the right answer. She made it clear that a poorly written algorithm could be very destructive. So before you hand over the keys to data scientists to make your marketing less of an art and more of a science, you do need to give some thought as to how you test (and retest) the algorithms that are being put to work on your behalf.
Attending TED is a sure fire way to make you feel like the dumbest and least meaningful person in the room, but it also reminds you that as the world faces seemingly larger and more complex challenges, there are some amazing people out there who are already working on these problems and producing some incredible results.