SOCAP International

3 Reasons the AI Revolution Is Actually Human

A Successful RevolutionAs companies transition into what the World Economic Forum calls the Fourth Industrial Revolution, they are being challenged to leverage the most advanced technology ever available. It might seem on the surface that a technology revolution would not require much human consideration. However, the biggest shift—and why it is truly a revolution—is that we are asking machines to do work that used to be the jobs of humans.

Humans are limited by their capacity: physiological limits like speed, strength, a need for rest. They are also limited by capability: mental limits like the ability to retain large volumes of information, and recognize patterns in large data sets. 

A human can only work so long and remember so much. New technologies—chat bots, automation, machine learning and artificial intelligence (AI)—do not have these limitations. (See Fig. 1)

The transition to these new technologies is not as simple as the typical technology implementation—that’s where the human component comes to the forefront, specifically in respect to these three areas:

1. Cultural change is a prerequisite for adoption.

The current perceptions of this technology are mixed with excitement at the new possibilities, as well as fear of what those capabilities mean. At a very logistical level, there is fear of job displacement. According to the survey “AI-Ready or Not: Artificial Intelligence Here We Come!” conducted by Weber Shandwick and KRC Research (as reported in Harvard Business Review, Oct. 24, 2016) “When asked whether AI is more likely to create jobs or lead to job loss, more consumers said job loss (82%) than job creation (18%).” At a cultural level, concerns arise around the moral and ethical use of this technology—even down to the question of whether or not robots should have rights. Most people are not fully educated on what is possible and what is not likely (at least for now), so there is a tenor of a fear of the unknown and the uncontrolled.

Fig. 1The same survey reveals: “Two-thirds of those surveyed say they know something about AI, although only about two in 10 (18%) say that they know a lot. One-third acknowledged knowing nothing about AI. We found that by far the most common first impression of AI is ‘robots,’ as 22% of respondents said.”

Additionally, and this shift is already occurring, the cultural norm of using this technology in our day-today activities has to become more commonly accepted. The more people interact with bots and experience the value these tools can have as enabling partners, the more adoption will become commonplace. It is a matter of time before relying on these technologies becomes the normative experience.

What to do: Accelerate cultural shifts in your organization by embedding behavioral change and targeted communication specifically addressing the current perceptions into your implementation strategy. Companies can have the best tech in the world, but it only has impact if they can convince humans to use it.

2. Experience differentiation is a relationship game.

AI RevolutionMost organizations acknowledge the value of customer experience delivery, though many still struggle to execute on it. Brands that excel at CX stand out among their competitors. Yet as good CX becomes the norm, and consumers rely more than ever on digital connection, there is a shift in expectation on what makes a good experience.

Consumers are now demanding a more human connection—they want to not just have a great interaction with organizations, but they want their experience to be personalized and curated to them at the individual level—they crave meaningful connection, evolving CX into human experience (HX).

Research into customer emotions, conducted by Motista, a predictive intelligence company, and reported in “The New Science of Customer Emotions” (Harvard Business Review, Nov. 2015), reveals that “customers become more valuable at each step of a predictable ‘emotional connection pathway’ as they transition from (1) being unconnected to (2) being highly satisfied to (3) perceiving brand differentiation to (4) being fully connected.” CX is number three on this pathway, whereas HX is number four.

This demand for human connection seems to be in direct conflict with our new-found technological capabilities—after all, how can an organization deliver a human experience with machines and algorithms?

Fig. 2The common thinking right now is a binary choice between humans or machines, when the real answer is partnership—leveraging machines to free humans to do high-value work—arguably relational in nature. (See Fig. 2.) Companies that differentiate on HX are going to intentionally decide which parts of experience delivery should be optimized with machines for easy, orchestrated delivery and which parts of the experience demands relational connection and a human touch.

What to do: Map out your customer journey(s) and designate which moments in the journey are best delivered by humans, or machines or both. Conduct consumer insights work to ask customers about their preferences and their perceptions. Avoid mindlessly applying a technology solution simply because the business has the technology/capability.

3. Conversation is a human social activity and there is no room for error.

Note: This applies specifically to chat bots. Let’s say all the appropriate customer research has been done, and it’s the right decision to deploy technology that enables your business to connect with consumers via conversational commerce. There are very human-centric design considerations for driving maximum adoption and consumer impact, so much so that Google has a division dedicated to it—humancentered machine learning (HCML).

First, accept that it is human nature to ascribe human qualities to things we talk to—even if we know they can’t talk back. Just think of how most people engage wit their pets, for example. Humans are inherently social beings with an inherent need for connection as a means of survival. (“Social Connection Makes a Better Brain” The Atlantic, Oct. 29, 2013) So, even the simplest chat bot should have some level of personhood—basic qualities like a human name, an origin story, and saying “please” and “thank you” when interacting with users will make it feel more social. Doing so meets an unspoken but very real human expectation of interaction.

Second, there is very little patience for bot error—which is any time the user expectations are not met by the bot. Essentially there is a “one and done” phenomenon as many users will abandon a bot after one to three times of experiencing bot error. Essentially the bot experience needs to deliver the expected outcome quickly and hit the exact right balance of “humanness” right out of the gate or consumers will quickly disengage.

What to do: Build a clear narrative about your bot to set the right expectations—what it does for the human user (and what it can’t). Design and test the level of “personality” with live users to get feedback on the right balance of human qualities before full launch and then build a feedback loop into your bot so it can consistently capture consumer insights for continuous improvement.

Like with anything focused on creating a differentiated, innovative consumer experience, the deployment of even the most technologically advanced tools requires human consideration, insights and ultimately the judgment on where to deploy tools and where to be purely human.

Joni RoylanceJoni Roylance, aka “The Heart,” is global lead for AI readiness. She is hyper connected to the emotional world of humans and how those emotions impact everything in business. She has more than 10 years’ experience in employee experience design, CX enablement, culture definition, talent management, and organizational design.
Brian ClementsBrian Clements, aka “The Robot,” is director of technology at Sparks Grove, helping grow the world’s largest brands. His brain lives at the intersection of marketing, data and technology, so much so that his ability to perform basic life functions often eludes him. He and his team research, design and develop analytics solutions and products in the fields of cognitive services, information retrieval, social network analysis and text mining.