The concept of Artificial Intelligence (AI) has been around for more than five decades but the chatter (and thoughtful discussion) about it has ramped up in recent years as more of the tools and systems we use move toward using increasingly powerful AI-like technology. But rather than brace for a dystopian future of mass unemployment, where machines rule over their feeble-brained human creators, our minimum due diligence as marketers and technologists is to familiarize ourselves with the basic concepts involved in AI and embrace a near future where AI empowers greater human communication and improves the quality of human lives.
The basics of AI
A basic grasp of the principles involved in AI keeps us from drifting into science fiction territory and tuning out altogether. It also helps us appreciate AI’s impact and potential.
AI itself can be roughly defined as the appearance of intelligent behavior in machines. Very straightforward.
When the discussion gets heady, philosophical, and esoteric is when we try to define what “intelligence” means. And as our machines get more “intelligent” and approximate human intellectual capacity in new realms, experts also keep moving the goalpost for what defines true intelligence, which complicates matters. That discussion then forces us to try to define “learning” and, eventually, “consciousness,” “sentience,” and “self-awareness”. When even the experts in the field can’t agree on standard definitions for these things, it’s no wonder we lay people can feel intimidated by this domain and eventually grow existential and depressed. But again, at least the definition of AI itself is simple enough.
In defining AI, we also need to make the distinction between “weak” AI and “strong” AI.
Weak AI is an artificial intelligence that is built for a fairly narrow purpose. Think Siri or Alexa — those systems are definitely impressive in their ability to parse human speech, look up information or control devices and media around the house. But when challenged with multi-step tasks or deriving conclusions from the data they have access to, they’ll delegate to Google or politely tell you to take a hike.
Strong AI, on the other hand, is artificial intelligence that displays all the features of human intelligence (including reasoning, creativity, and sentience)—Blade Runner or Westworld level stuff. We’re not there yet, but if and when we get there, you can be sure we’ll be arguing about whether these AIs deserve their own social security numbers.
AIs can “learn” through a process referred to as “machine learning” in which the machine has access to vast amounts of data and is programmed to identify patterns and make correlations that help answer certain questions or perform certain tasks.
Neural networks are artificial structures built and programmed to behave like the neurons in the human brain and, based on vast amounts of data and logical reasoning processes similar to those our own brains perform, derive learning and insights that drive decisions. The machine’s act of acquiring knowledge from neural networks is called “deep learning.”
Applications of AI
With a basic understanding of the concepts involved in AI, we can start to grasp its application to some of the digital tools we already use as marketers, technologists, and in our daily lives as consumers.
How does Facebook know that your friend Alice is in the picture you just uploaded? And how does Google Photos know which of your pictures were taken in a park or in a living room or have a dog in them? This is done using the concept of computer vision, which applies machine learning and deep learning in order to analyze patterns in pixels and clusters of pixels in billions of photographs, and quickly identify the objects therein. This technology is also applied in DAM (Digital Asset Management) systems, in which your company’s digital asset library might be handily tagged and categorized, thus saving a human hundreds of hours of tedious manual organization.
Search engines, recommendation engines, and content curation
Here, we’re taking Google, Bing, or systems like Amazon’s own site search. Typically, search engine algorithms are proprietary and use a combination of many technologies. While in the past these search algorithms were based on sophisticated and effective logic that was nonetheless narrow, they now increasingly rely on AI-like processes of machine learning and deep learning. This is how we find ourselves typing one character in the Google search box and uncannily getting suggestions for exactly what we were searching for. Google and other engines not only are indexing data and scoring it based on relevance, but they’re now anticipating our intent based on vast amounts of information they can access about us. Creep factor aside, these processes further accelerate the speed at which we can access information and perform our personal and professional tasks.
Predictive customer service
Customer service forms are increasingly relying on self-service processes. As the user types the nature of the problem they’re contacting customer service about, they are offered suggestions for where the answer might be found. It may be a stretch to call many of these systems AI, but their sophistication continually increases by applying AI principles of machine learning and deep learning. Nothing replaces human connections, but occasionally, we might prefer to get a precise answer from an AI rather than receive a non-answer from a human who’s overworked or not properly trained.
Application of AI to website design still leaves a lot to be desired, but it may be useful in instances where only basic site aesthetics and organization are needed. Wix.com currently offers, side by side with its library of human-designed templates, an “AI-based” system for creating layouts. The reader should be cautioned that many businesses currently boast of AI-based tools and services when they might not truly fit the definition of AI, but this trend is nonetheless interesting. In the case of website design, the tasks of a human website designer might shift towards supervising or refining the work of an AI and away from conceiving of the work from beginning to end. Similarly, there are other interesting experiments in AI-based design out there such as https://thegrid.io for website design, http://colormind.io for color scheme generation, and http://fontjoy.com for web font pairing.
The world seems infinitely hungry for content and as more of it becomes available, we are caught in the vicious cycle of having to create more and more frequent content in order to compete for the attention of our audience. AI has entered this arena in various forms. Content creation platforms like Atomic Reach and Curata use deep learning both to suggest and to create content that marketers can publish online to satisfy the needs of their specific audience and increase their online visibility. News organizations are also starting to rely on AI to glean insights from large amounts of data and, in some instances, to write newswires and basic news reports.
Why we must embrace AI
We’re already well inside the brave new world of AI and it will continue to become a bigger part of our daily lives. When the internet became widely available to the masses in the 1990s, very few of us could have predicted the emergence and eventual impact of tools like Twitter or GoFundMe. Likewise, AI has the potential to disrupt our lives and the way we work drastically, in some instances making jobs we do today obsolete and in others, creating new opportunities and new frontiers for human beings. Marketers and technologists must at least become as proficient as we can in the basic principles of AI. Like any other disruptive technology, AI is ultimately a tool. Our understanding of it will ensure that we can maximize its use for our own main goal, which is to facilitate tasks and communications and to improve the lives of human beings.