AI is in nearly every application nowadays. A frequently cited IDC statistic has 75% of commercial enterprise apps using AI by 2021. And as AI has become more prevalent, the C-suite has taken a bigger role in AI strategies and implementations. In recent industry survey, 71% of respondents said that their company’s AI projects are “owned” by C-level executives.
Yet, even with C-level involvement, many organizations’ efforts with AI are falling short. Leaders often view AI as a plug-and-play technology with immediate returns, and they struggle to take a more holistic approach. For instance, they may focus on discrete business problems, such as improved customer segmentation, rather than big business challenges, like optimizing the entire customer journey.
As developers integrating AI into applications, how can you reconcile the need for immediate returns with bigger picture goals? To design for an AI-infused future, you’ll need to think more broadly than you may have in the past.
To start, consider a few truths about AI that should shape the way you approach AI design.
- AI is not as smart as many people think. It’s been said that AI is much closer to the brainpower of an earthworm than a human.
- AI bias exists and is increasingly common.
- The interaction between smart machines and humans is necessitating a form of user experience design for working with intelligence machines in a way that’s more mindful of the nature, intentions, and values of human beings.
You need to think beyond just automation metrics to bring intelligence to products in a way that is truly conscious. We call this approach Conscious Experience Design—a design discipline tailored for a world shaped by artificial intelligence. Conscious Experience Design helps us answer questions for our clients such as:
- How is the human-machine relationship evolving?
- What are the characteristics of conscious machines?
- What are the future paradigm shifts that may require new approaches?
- Which new skills may be needed to design your AI-infused applications now and in the future?
There are three main principles of “conscious experience design” that will help you integrate AI into your applications in ways that can evolve as user experiences and AI change. I’ll discuss these three principles and steps for implementing them below.
Design for human intelligence and expectations
When designing apps with AI, it’s natural to think about what data is needed to increase the machine’s or system’s intelligence. But this focus on data overlooks a fundamental step—understanding and empowering human intelligence. AI is an enabling technology meant to empower and enhance human capability and potential. If we only train the machine and misalign it with the core human needs it becomes a wasted endeavor unlikely to deliver real value.
The way around this is to begin with people from a research and insight perspective. Before you begin integrating AI into business applications, first seek to understand the expectations and key problems that need to be solved. Explore where and when people could benefit from offloading tasks to automation and how much control and awareness they desire to achieve an effective human-machine collaboration.
Design for immersion not interaction
AI is about living systems that are ultimately going to be smart enough to know when, where, and how to engage us. Designing for immersion means that experiences should feel natural. People should be empowered by AI and automation. People should be less hindered, and their interactions should feel frictionless.
To get to this level of immersion, technology on the back end needs to be able to learn from people to become a system that ultimately intuits what people need. And on the front end, designers need to create similarly frictionless interaction. Allow people to speak rather than grab a mouse. Consider immersive experiences that can anticipate and engage in ways that require less effort.
At the same time, when you engage the senses, you want the system to feel “living” and ”real,” not clinical and machine-like. Think about the “EQ” or emotional intelligence of the system, and the human emotions it will impact and interact with. For instance, voice technologies may have a hard time intuiting emotions that could be gleaned from facial expressions. It is incredibly nuanced to understand what people are thinking and how they are feeling. Many tech leaders understand the concept of multi-modal interaction—the same applies here. Make the system come to the people. Don’t make people have to come to the system.
Design for people and systems, not products
It’s easy to focus on just a product or a touch point. But in this new era of intelligent technology, you’re actually creating something that works as part of a full system. A node someone interacts with is just one dimension of that system, yet beyond that there is systemic impact. For instance, blockchain leverages intelligence across the network to assess and determine security—taking advantage of distributed computing vs. using one node.
Designing for AI enablement produces the same ripple effect across all types of devices or products. What used to be designing for a car is now designing for mobility, and for transporting people. Systemic thinking also extends to societal implications of creating bad AI. AI takes on characteristics of the people who put information into it, so it will embrace biases of the people working on it. Simply being aware of this fact helps designers avoid perpetuating bias.
Put conscious experience design into practice
So how do you support all of these principles? Follow these tactical steps.
Research. Audit and map your customer experience journey through internal and external research. Interview the customers and employees who will use the application. Ask about their pain points, what they enjoy, the best part about their day to day work, and the worst parts about their jobs. Do they fear that AI will make them irrelevant? Or do they think that AI will give them more ways to succeed?
Understanding the people who will use AI is critical to your project’s success. This is the shortest path to identify gaps and opportunities for better solutions enabled by AI. Too often AI designers don’t have a succinct answer to the question, “Why do you think you want AI and what do you really want to solve?” Use your research to answer this critical question. Identifying where the biggest gaps are—and the biggest opportunities for immediate value—will help you focus your efforts.
Envision and clarify. Once you home in on specific problems from your research, understand the specific intent and outcome for each of the opportunities. For this step, collaborate with a cross section of stakeholders inside your company. Write clear value statements for intended users. Also use this time to assess and evaluate potential risks and side effects. Collaborating with various stakeholders ensures you’ll have a stronger understanding of the outcomes of AI on the system and the business. As we mentioned above, AI can generate ripple effects, not just on the intended area but even in unforeseen areas.
Assess the impact. Now that you’ve defined the intent and outcome of potential AI integrations, prioritize the most achievable and impactful options. Consider the potential external impact of your AI (customers, privacy concerns) and internal impact (is the AI integration feasible, sustainable). For each potential project, ask questions such as, How equipped are we to actually deliver? What do we lack? Will we need specialists? How much can we invest in them? What hinders our ability to achieve these goals? If we embed AI for this particular project, what are some scenarios that may have positive outcomes? What are potential negative outcomes? Prioritize the project that sits in the intersection of internal and external impact and feasibility, and you’ll have the best chance of success.
Map AI interaction. In this step, you get down to the nitty-gritty and determine the level of interaction or transparency of the system and collection to users. Is there going to be implicit interaction or explicit interaction? Explicit is a human embodiment like Alexa. Implicit AI means the intelligence happens entirely behind the scenes, without users even aware it is happening. You must figure out which part of your AI technology is visible to your user and to what degree users will interact with the AI. Should the AI be a back-end system where explicit interaction is not needed? Or should there be direct interaction with users via voice, camera, or typed response? How friendly will the AI need to feel?
Define a data strategy. This too is a serious endeavor. Artificial intelligence is driven by data. You’ll need to dig in and define a clear data strategy that establishes how the company will collect, manage, store, use, and share this data. This strategy will serve as a consistent guide and common ground framework for teams and products. It is imperative that you outline ways you will empower the business while protecting your customer’s data.
Data scientists and data advocates will be important in this process to help you design both a user-friendly way of gathering data and the most effective, insightful, and beneficial ways to provide data to users and stakeholders. Be sure you pull from a diverse team when setting data metrics and system parameters. Seek ways to avoid biased data through the use of equitable data frameworks.
Remember that when it comes to designing with AI, there is no beginning and no end. You are creating living systems that will be sharing data perpetually that you need to monitor, manage, and evolve. By applying the principles of conscious experience design and the tactical steps outlined in this article, you and your team will have a solid start to successfully designing for living AI systems.
Ken Olewiler is principal, managing director at Punchcut, a user interface design and innovation company specializing in mobile, connected products and services across devices.
New Tech Forum provides a venue to explore and discuss emerging enterprise technology in unprecedented depth and breadth. The selection is subjective, based on our pick of the technologies we believe to be important and of greatest interest to InfoWorld readers. InfoWorld does not accept marketing collateral for publication and reserves the right to edit all contributed content. Send all inquiries to email@example.com.