AI Unleashed: Transforming UX for Tomorrow
The realm of AI is an ever-shifting landscape, catalyzing change across industries. At Ekohe, we don’t just play alone — with our powerful development team as the backbone, our design wizards are weaving our own brand of magic into the design realm to enhance user experience.
And here’s the secret sauce: we don’t wield AI just because it’s the hottest ticket in town. We adhere to the mantra of “Don’t use AI just because you can.” We’ve got the latest tech in our arsenal, but we wield it with surgical precision, ensuring it adds maximum value for both business and users.
The digital cosmos is shifting, and AI is at the helm, steering user behaviour and expectations into uncharted territories. Ekohe’s mission is not just here to ride the AI wave — we’re shaping it, and moulding it to transform the UX of tomorrow. Drawing inspiration from the AI Guidebook from Google PAIR, in this article, I’ll walk you through some real project examples and share tips on designing a product in an AI-powered world.
Bringing AI into the Fold: Adding Value with Purpose
In the cacophony of AI chatter, designers must exercise discernment. Not every field is ripe for AI integration. Before embracing this tech revolution, we advocate a deep dive into user needs and desired outcomes.
Take one of our podcast projects for instance, with a mature platform as our canvas, we set out to elevate content promotion and precision this year. Podcast (or other content-creation) aficionados relish personalized content recommendations, an arena where Natural Language Processing (NLP) shines. In the following paragraphs, I will reference various design examples from this podcast project (as well as another PE/VC project) to elaborate.
Setting Realistic Expectations
AI is a Pandora’s box of unpredictability. Designers play the crucial role of guiding users through its capricious terrain. Transparency is our guiding principle — make it clear to users what our product can and cannot do.
In high-stakes situations, like aerospace or healthcare, it’s our ethical duty to underline that AI’s crystal ball isn’t always crystal clear.
Let’s imagine designing the integration of a neural search function into the existing podcast/episode search system. Picture this: users input a keyword, and like a virtuoso, the machine interprets the essence of their query, conjuring related results. Yet, here’s where the plot thickens: it’s not always about hitting the exact keyword bullseye. After fully understanding, the AI system gathers all relevant content searching by associative keywords.
So, when designing we’ve thoughtfully placed a gentle reminder to help fine-tune our users’ expectations, that they probably won’t be able to see any matched keywords that are artfully highlighted in the results, like in the past.
Focus on Benefits, Not Technical Wizardry
In our intimate dance with developers and data scientists, we’ve gleaned intricate insights into the AI machinery. Yet, our users crave clarity, not a backstage pass to the tech circus.
The art is in illuminating the product’s prowess, not dissecting its inner workings. Conducting user research will help to gauge the right balance: What do users need to grasp about AI’s mechanics to wield our product effectively? Why does this knowledge empower them?
Striking a Balance: Precision vs. Recall
In the game of AI, precision (higher quality) and recall (more results) are the two ends of a seesaw.
In some high-stakes domains, like our other products designed for PE/VC companies or the FIFA World Cup, precision reigns supreme.
But in the podcast example here, the platform’s low-stakes environment allows us to prioritize recall, casting a wider net of recommendations. It’s the thrill of serendipitous discoveries, though some may be a tad off the mark.
Anchoring in the Familiar: A Trust-Building Endeavor
The allure of AI’s novelty can be intoxicating, but it’s a double-edged sword. Sometimes I find myself easily inclined to convey the magic of our predictive system through the interface, but gradually realize that unfamiliar interfaces can shatter trust, irrespective of AI’s quality. There is a balance to find between fresh and familiar.
Hence, we embark on a journey of user onboarding imbued with comfort and trust. Mystery breeds fear, and guidance dispels it.
A Human Touch in a Digital World
AI recommendations shine brighter with a dash of human context. Trust blooms when recommendations align with third-party endorsements. Human connections are the currency of trust, which can be built through community reviews or friends’ input: For instance, providing some contextual information from other people can help users determine the value of the “Best Match” recommendation of our podcast platform.
Feedback: The Two-Way Street of Improvement
In the grand exchange of information, our AI models not only recommend but also welcome feedback and corrections. The feedback avenue boasts various lanes — a simple thumbs-up or down, flagging troublesome suggestions, or banishing unwanted ones, etc.
But don’t forget, the journey doesn’t end at feedback collection — it begins with a response. Users deserve acknowledgement, clarity on follow-up actions, and an understanding of how their feedback will in the end benefit themselves.
In the age of AI, design is the compass that steers users through the digital wilderness. By heeding these principles, designers are expected to wield AI as a tool to illuminate rather than mystify, fostering a future where technology serves users and businesses alike.
