Read time 4 min
Algorithmic fatigue arises when machines fail to deliver on the expected user experience. A pioneering new study conducted by Alice Labs in partnership with Reaktor examines how AI can deliver a service that delights rather than disappoints. The solution might be to build something more human.
Endless scrolling. That’s one of the telltale signs of a novel phenomenon our new research identifies as algorithmic fatigue. People who spend ages browsing streaming services, looking for something new to watch, are some of the many examples of the growing numbers of consumers who are now finding that AI systems fall short.
When the algorithm fails to live up to people’s expectations of the user experience, and doesn’t deliver the service its users want, the people using the system end up feeling annoyed, frustrated, and fatigued.
Brands are slowly waking up to the same realization: AI is no longer just about the technology; it’s about how humans experience and interact with the algorithms.
While useful in many instances, algorithms continue to be limited by their machine-ness: they cannot predict when users are having a bad day and need something lighter to watch, nor are they capable of understanding the subtle and varied ways in which users’ tastes evolve and expand over time.
“When the algorithm fails to detect and cater to the individual need in the specific moment, the result is algorithmic fatigue.”
Through a new study conducted by Alice Labs in partnership with Reaktor, we identified three different types of AI interactions: Passive, Guiding, and Collaborative. In other words, we found that there are some instances where users want to remain passive towards the algorithmic system; other moments where they seek to guide the system; and some occasions where they hope to collaborate with the system.
These wants and needs are always shifting, and depend both on the nature of the specific situation those users are in as well as their previous experience with smart technology. When the algorithm fails to detect and cater to the individual need in the specific moment, the result is algorithmic fatigue.
“Consumers now use their best interaction experience in one domain as a baseline expectation in others.”
Customers’ expectations for algorithmic interactions are already exceedingly high: we are at a tipping point for brands, where the ability to create smooth and enjoyable algorithmic interactions is no longer a competitive advantage – it’s a must-have. Consumers now use their best interaction experience in one domain as a baseline expectation in others.
In short, the best-in-class user experience is now the same for everyone, regardless of industry or location. That means that, when it comes to AI, every single business is in competition with the global giants, including Amazon and Netflix.
While it’s true that tech companies have been early adopters of AI and are therefore able to invest enormous amounts of money into its research and development, AI has in recent years become just as viable a proposition for small and mid-size companies. It doesn’t necessarily have to break the bank. Companies of all shapes and sizes should get in the game now, before it’s too late.
“The only thing worse than having poor AI is having no AI at all.”
Addressing algorithmic fatigue starts with taking AI seriously across the entire organization. The only thing worse than having poor AI is having no AI at all.
But beyond investing in and prioritizing AI, firms should also zero in on its actual user experience. Creating another layer on top of the working AI, something between the user and the algorithm, can help create a more responsive and refined interaction for consumers.
Ultimately, it’s about creating parity: granting the users equal agency over decision making by allowing them to choose and change when they want to be actively involved in the algorithm or just passively guided by it.
As always with AI, it’s important to remember that there’s no one-size-fits-all solution. Managers should be mindful that artificial intelligence isn’t something that can be built in a day or bought ready-made at a store: it’s a company-specific capability that’s built over time. And as our research points out, the time to start investing in it is now.
Want to learn how to move from perfect algorithms to perfect user interactions? Read more about the compelling findings of our EverydAI research project.