Last week I attended a talk given by Dharmesh Shah, Founder/CTO of Hubspot who spoke on the impact of machine learning and artificial intelligence has (or will have) on marketing. He mostly discussed the wide array of matters stemming from the ubiquity of advanced technology in marketing. The crux of his position was that A.I. would serve to mostly augment human involvement in the marketing rather than replace it. I found this to be both a highly defensible and platonic view. Nevertheless, I wondered if A.I. was capable of replicating certain human behaviors that were less programmatic and more variable.
How does A.I. Account For Logic-Defying Marketing Problems?
If you’re a digital marketer with a fair amount of experience with A/B testing, you’ve run into the following scenario at least once. You put together a well-designed A version that has the best content, provides the best user experience, and adheres to all of the best marketing practices known to man. A is run against a sloppy B version. You send an equal and statistically significant amount of traffic to each. A version should blow B out of the water, right? But, you are very wrong and B does exceedingly better than A. No obvious reason other than human variability.
The question is how does A.I. account for that? How does A.I. realize that sometimes humans like to deal with a bit of a mess? Surely a logical machine would seek to eliminate anything that doesn’t adhere to best practices before it ever reaches the eyes of an actual human being. Yet, in this generic example, adhering to best practices ended up being the wrong approach.
Dharmesh pointed out in his talk something that artificial intelligence cannot understand. Robotics cannot replicate human perspective and creativity. Technology can be programmed to be random, but it cannot apply emotions or instinct to its activity. While some marketing processes can be farmed out to A.I., humans still provide value in creative contributions. With that said, no one introducing a machine learning system into their marketing activities does so with the intent of replacing or augmenting human creativity. The intent of involving this sort of technology is to maximize efficiency and productivity. This led me to another quandary.
What Happens With Human Desire To Face Challenges?
Given that many human beings enjoy facing challenges (within reason) that lack obvious solutions and sometimes defy logic, how does artificial intelligence operate in a way that doesn’t conflict with this human virtue? For example, there is no logical reason to climb a tall mountain. A mountain peak is cold, dangerous, and lifeless. It lacks resources necessary to support human life. Getting to the top of one exposes one to a drastic increased risk of dying. No A.I. system would choose to go up on a mountain peak, yet humans pursue that kind of challenge all the time.
Let’s say a similar challenge presents itself in the marketing world (to be fair, there’s nothing life-threatening in it). Let’s say that in order to advance beyond the market leader in an industry, a marketer has to try something that defies logic and the path doesn’t appear to make sense. They take a tremendous risk with funding and organizational resources and it pays off with success in the marketplace. They push their organization forward. Would an A.I. system take that risk? Or would it simply pursue the path of least resistance over and over again? There are probably people much smarter than me who could answer that question in less than 2 seconds. Yet, I find it fascinating to wonder about.
Whether you think artificial intelligence is a fascinating frontier beginning to open up in the marketing world or if you think it is a sinister force of evil, it’s clear that we are quickly arriving at a very interesting time in the marketing world. I for one eagerly await to see how AI and machine learning will play a greater role in my world as a marketer.