We’re all interacting with more intelligent machines than ever before — whether we’re asking Siri to suggest the best sushi spots nearby or utilizing Tesla cars’ self-driving features, or even consulting a robot lawyer to get us out of parking ticket fees. Use of artificial intelligence (AI) is only going in one direction. IDC expects worldwide spending on cognitive and AI systems to reach $97.9 billion in 2023 — that’s two and a half times more than its $37.5 billion prediction for 2019.
AI is no longer a new technology at the workplace. It’s now table stakes and is being used across all aspects of our business. As a result, there tends to be a lot of discussion around the fear of robots kidnapping our careers. Yet according to Gartner, the reality is that AI will create more jobs than it will eliminate. That’s because artificial intelligence requires human intelligence. For example, most AI on the market tends to be narrow, which means that it must solve a specific problem that humans choose for it, and operate within the predetermined goals that humans set — at least for now.
Rather than implementing AI for its own sake, businesses will get the most value when employing augmented intelligence, which involves using AI to make humans happier, drive smarter decisions, and create better overall experiences. Here are three ways businesses can begin to foster a human-machine partnership.
1. Empower employees to be more human at work
People often get excited about more advanced AI techniques — it’s easy to imagine a world like Westworld or Ex Machina when we’re already mistaking the chatbot in the online forum of our college class for a real, human teacher’s assistant. But businesses find the most success with AI when they start small, and then accelerate. In fact, one of the most useful benefits of AI currently is simple — automating repetitive, routine tasks in more innovative ways than automation can do without the help of AI. The power in AI-based automation doesn’t just lie in outsourcing our most dreaded assignments to machines, but in enabling us to be more human at work.
By letting machines take over mundane tasks such as data entry or billing hours, employees can focus on developing skills that are central to the human experience, such as creativity and emotional thinking, which can be more complicated to automate. While these skills are at the foundation of what it means to be human, we don’t spend much time exercising those parts of our brain at work, according to McKinsey:
Just 4% of the work activities across the US economy require creativity at a median human level of performance. Similarly, only 29% of work activities require a median human level of performance in sensing emotion.
In combining automation and AI to remove dull work, brands can free employees’ time for projects that allow them to be imaginative, use more interpersonal skills, think critically, and improve overall happiness at work. And having engaged employees pays off — research shows that employee engagement leads to improved customer relationships and increased sales.
Customer service provides one example of this phenomenon. For instance, Dollar Shave Club deployed a chatbot to answer frequent customer questions that members could easily solve on their own. As a result, human agents have more time to focus on issues that require more creative thinking to resolve. Another way companies are giving agents time back for higher-stakes concerns is by using AI to automate the translation of tickets in other languages, thereby fostering a more inclusive customer service.
On the retail front, Amazon is using AI in its Amazon Go stores to automatically check customers out from the moment they grab an item. This first-ever “Just Walk Out Shopping” experience eliminates lines and enables employees to focus on interpersonal, rather than transactional, interactions with customers. Next, we might see robots restocking shelves to remove the burden of manual labor.
Outside of customer service and retail, lawyers might let AI handle copyediting and reviewing contracts, help with conducting research, and even generate defense letters for simple cases so they can spend time developing deeper relationships with clients. Or, insurers and brokers might use AI to follow up with leads, summarize reports about clients’ financial histories, and provide quotes so they can focus on finding solutions that fit their needs in more innovative and personal ways.
2. Redistribute decision making between robots and humans
Many companies are already using AI to uplevel their automation game, but some are taking it a step further — combining AI’s predictive capabilities with human judgment to drive smarter decision making. Mike Rollings, research vice president at Gartner, states:
Rather than have a machine replicating the steps that a human performs to reach a particular judgment, the entire decision process can be refactored to use the relative strengths and weaknesses of both machine and human to maximize value generation and redistribute decision making to increase agility.
One such strength of AI is its ability to spot trends that humans can’t see. Whereas traditional computer algorithms require human developers to set rules that inform its output, machine learning, a type of AI, and a subset, deep learning, are capable of learning on their own. These machines parse data, learn from it, and get progressively better over time. The result is algorithms that can apply what they learn to make strikingly accurate and fast predictions, opening the door to a new kind of machine-aided decision-making process.
For example,customer service agents, such as those at Pinterest, are using machine learning to predict customer satisfaction during an interaction, before the customer takes a satisfaction survey. This helps agents prioritize customer concerns and enables them to make more informed decisions to proactively ensure positive outcomes. Companies are using AI to improve decision making when it comes to self-service, too. For instance, if a rising number of customers are asking questions about a new software update, machines can proactively flag this trend and suggest help article topics that meet customers’ needs.
Recruiters and hiring managers are also taking a bite of the machine-learning cake, using it to reduce unconscious biases and make more diverse and inclusive hiring decisions. That’s because machine learning algorithms can be programmed to predict high-quality candidates without considering demographic information, such as gender, race, or area codes that correlate with income.
Doctors are using machine learning to predict a patient’s length of stay in the hospital, probability of in-hospital death,and to suggest medicines to help them decide the best treatments.
Where humans come into the decision-making process is by ensuring that AI remains transparent, which requires developing an understanding of how AI is making its predictions. For instance, since AI learns from patterns in previous behavior, robot recruiters can still learn to replicate our own biases, even if we attempted to control them. Of course, for AI to truly help us make decisions, we need to trust it. As such, it’s imperative that our understanding of machines scales to ensure that AI remains explainable, auditable, and fair.
3. Use machines to foster customer intimacy
AI, and machine learning in particular, also enable companies to foster personalized brand experiences that result in more intimate customer relationships. And it’s worth it — 91% of customers are more likely to remain loyal to a company that provides them with relevant, personalized offers and recommendations, and 83% are willing to share their data just to get this experience, according to Accenture.
These are the kind of right-thing-right-time moments, like when LinkedIn surfaces a dream job in your area as soon you begin looking for a new role, or when Amazon recommends just the product you need before you even knew it existed. Even your Spotify Discover Weekly playlist introduces you to a new favorite song or artist.
Personalized experiences like these, that help users get better deals, save them time, or provide new information, and that make the purchase process easier or less confusing have the highest impact on customers, according to Gartner.
But personalization is easy to get wrong, which is why consumers get email offers for products they’ve already purchased or continuously see ads on social media for things they’re no longer interested in. Much worse is when personalization starts to feel invasive or offensive, as in the story of when Target accidentally exposed a pregnant teen to her father.
To avoid embarrassing hiccups when it comes to personalization and AI, businesses should foster a human-machine feedback loop. AI isn’t magic — it’s subject to ‘garbage in, garbage out,’ which requires human analysts to interpret data thoughtfully and contextually before feeding it to machines, and to evaluate the implications of the output of that data.
Ask a new question
We should always ask ourselves whether AI is adding to an experience or detracting from it, whether we're talking about the customer or employee experience. But the question is no longer, "Should we use AI?" so much as "Where and how should we use AI?" to capitalize on our respective strengths and weaknesses, and to foster human-machine relationships that work better together.