Artificial intelligence is a science fiction staple, but the brainy machines haven’t always been benevolent. Look at HAL 9000, Skynet, or the rogue replicants of “Blade Runner” for just a few examples of AI run amok. But when tech experts talk about AI, they’re referring to something far more specific – computers that think and learn as people do, at least in some ways.
We’re already seeing the beginnings of true AI in marketing automation technology. Instead of brute-force methods that provide approximations of data without much interpretation, marketing automation systems can now analyze actual incoming data and offer genuine insights into a market. What’s more, the Internet of Things is making it easier than ever for an AI to gather the knowledge it needs to learn.
What Is a Real AI?
We’ve had computers that were very good at appearing smart for a while. Deep Blue became the first artificial chess champion way back in 1997, for example. But while Deep Blue was exceptional at chess, it was only good at one task, and then only because it went through multiple permutations more quickly than a human mind – even Garry Kasparov’s – could manage. In other words, it was beating a human at a human game, but it was doing it in an entirely mechanistic way.
The trouble with this approach is that it works well for game theory and other applications in which there’s a single optimal solution, but what happens when there are multiple high-yield options? That’s often the case in marketing, where leads take multiple routes to arrive at a buying decision. Learning, not calculation, is the key.
Instead of looking at every possibility and choosing the optimal one, a modern AI analyzes data and develops rule-sets that explain what it sees, refining its knowledge as it learns new information. It’s a bit like an accelerated version of what scientists do, developing hypotheses and then revising or even discarding them as the data suggests. While older systems like IBM’s Deep Blue generated huge volumes of possible results, modern machine learning requires huge volumes of data.
Big Data Produces Big Insight
Neural networks approximate what goes on in human brains. We don’t process information by looking at multiple outcomes but by looking at the knowledge we have, understanding desirable outcomes, and devising ways to get from point A to point B. Processing takes place in layers, and the layers then communicate to produce a nuanced, complex response to a problem. We do that subconsciously. When you drive, you’re processing visual information from your view of the road ahead, your rear-view mirror, auditory signals such as hearing a car passing yours, and so forth to create a complete dynamic map of your surroundings, for example. Machine learning arrives at a similar outcome by using and linking different kinds of input. Self-driving cars are becoming a reality on the road because of this ability to learn.
Marketing Automation and AI
To learn, machines need data; marketing automation systems excel at gathering and storing data. The value of marketing automation technology to applied AI is therefore clear, but that’s only the beginning. Artificial intelligence will soon be everywhere. Your laptop, your desktop computer, and your mobile devices are just the start; soon, your car, your favorite shops, and your workplace’s assembly line will join the party.
When your customer base is giving you feedback via smart technology built into cars, mobile devices, and computer networks as well as their online activity, what can’t you learn about them?
© Reach Marketing LLC 2017 All Rights Reserved.