AI is the continued topic of discussion in 2017 and will no doubt remain so for the foreseeable future. Enabling machines to learn, make decisions, and adapt to circumstances without input from people (rather than simply obeying pre-programmed instructions) is the reality of the post-modern world. And while it presents tremendous advantages to society and businesses, there are just as many disadvantages. Being a product of a certain generation, I can’t help but conjure up memories of Terminator and The Matrix with self-aware, self-programming machines running amok.
But there are probably more people who subscribe to the more optimistic view that applying a more restrictive, less autonomous form of machine learning to the wealth of data could help identify correlations and patterns that were impossible for humans to see before. And the potential advantages are limitless – new ways of treating illness, quicker response times for emergency services, etc. From a business standpoint, offers will become more personalized, more relevant, and potentially involve less direct interaction (imagine your home being able to order groceries based on what it has learned about your tastes, habits, or medical needs). Imagine HAL 2000 with a heart of gold.
That said, there was quite a stir last year about customer service chatbots last year, but most of these were actually very limited, merely guessing the most likely answer to fit the question. Impressive to a point, but hardly the breakthrough we’ve all come to expect from SiFi. Real AI, underpinned by natural language processing, neural networks and machine learning, will understand how humans think, talk, and categorize concepts, making it smarter and easier to interact with. It’s simply a matter of time and processing power. And the more we use it, to depend on it, the better it will become. So we will no doubt see a proliferation AI buddies in the year to come, such as Alexa, Google Assistant, Cortana, etc. .
With AI, we have the opportunity to build decision-support systems that see, hear, understand and collaborate with us to help make decisions faster, more relevant and better informed. Which brings up an interesting idea: to whom do we market? Human beings are the obvious, unchanging element in the process, but are we on the verge of having to think about how to market to the machine? And if so, what does that look like?
If AI has the potential to act without our involvement and on our behalf, then we need to be ready to “sell” to the machine. And if AI can learn to make judgements about our personalities and those things to which we have an emotional or culturally grounded response, then our virtual assistants will be targets for marketing. For example, milk is more than a commodity. My assistant will be able to discern that I have a taste preference for glass-bottled, clover-fed milk. But it will also know that consumption aligns with my workout schedule, that I need to reduce my fat intake due to my age, and that I have a dinner party coming up where milk is likely to be used in cooking. It will have to weigh all of these variables, just like I would, and make decisions about what to buy. And that’s just milk. Now apply that to a car, a medication, or a vacation. The implication is that we will need to consider the possibility of marketing to a device that is weighing the same sorts of variables a human being would way, but which has a very different way of conceptualizing, categorizing, and responding to the world. Welcome to the brave new world of marketing to machines.