New developments in generative AI have made this technology more emotionally resonant, bringing exciting opportunities to connect with consumers on a deeper level. We hosted an event in Copenhagen to explore ways brands are using emotive AI, why it can help promote brand personality and how to design AI experiences responsibly.
Panellists
- Alisan Atvur, Principal UX Researcher, Novo Nordisk
- Anna Rafferty, SVP Digital Consumer Engagement, LEGO
- Caroline Rossing, Head of Human Interface Design, B&O
- Mårten Jönsson, Director, Android UX, Google
- Marten Tegner, Principle Data Scientist, IKEA
- MODERATOR: Nayan Jain, Executive Director of AI, ustwo
Key takeaways
1. Emotive AI experiences should align with your brand’s voice and values.
AI experiences are extensions of your brand. Consider tone and personality. For instance, LEGO’s AI agent customer service assistant Sophia communicates in the brand’s very defined tone of voice, which is applied across all of its products and IP, including movies.
Likewise, the experiences need to be designed in a way that’s authentic to your brand’s typical interactions with customers. LEGO’s building kit instructions always include text asking for feedback on the products, so LEGO explicitly asks for customer feedback in the experience with Sophia. “It’s very congruent with the way that we have always worked,” Anna Rafferty said.
2. AI brings new ways to reach, delight and educate your customers.
Many brands are selling services alongside their products. AI-driven experiences can help meet expectations and create new levels of personalised and delightful services.
Bang & Olufsen, for instance, can help users adjust complicated equaliser systems to get a better listening experience for their particular situation - like a rainy day in Copenhagen - and learn something along the way, explaining why that was the best adjustment if they’re interested. In addition to the superior listening experience, this also brings another opportunity for branded storytelling and appreciating the value B&O’s products bring, Caroline Rossig said.
3. AI doesn’t always have to get personal to create a personalised experience.
Just because we’re able to make AI agents that are getting more and more “personal” and human-like doesn’t mean we always need to. While people often talk about the future of AI agents that might know a lot about you, understand you on a deeper level or chat like you’re old buddies, this isn’t the only way to create successful products or services. It ultimately comes down to what you’re setting out to accomplish with an AI-driven experience.
For instance, Martin Tegner pointed out that a customer chat functionality doesn’t necessarily need to have a full conversation with the user or have an emotional aspect to fulfil its job-to-be-done. Maybe all a customer needs from a chat in a particular shopping context is a list of products and recommendations. The experience should be able to provide helpful, relevant and personalised information, but tracking the products a customer clicked on or what area of a home they were browsing through is enough.
Likewise, there is a lot of hype and consumer hunger around AI, but the real hunger will ultimately be what the technology can do for someone - getting things done more quickly, making their lives easier, Mårten Jönsson said.
4. Disclose when AI is part of an experience.
As AI experiences become more human-like, it’s important to have moments of friction or labels to distinguish when something is AI-generated. “When we start to add emotions and create identities that people interact with, it starts to get fuzzy,” said Alisan Atvur.
Guardrails are especially important when you have vulnerable users, like children or people with accessibility needs. For instance, LEGO’s Sophia might have a very clearly defined personality - she does yoga, she’s thinking about joining an improv class - but she always mentions that she’s a bot.
If AI can sense and respond to your emotion and make recommendations of content or products based around that, amazing things can be created - but these experiences should be labelled to be safe. People need to understand why and how they are being served up certain recommendations, Marten Tegner said.
5. Users are building a relationship with AI, which means building trust and acknowledging limitations.
Being forthright about limitations and intentions builds trust in relationships with people and technology.
The AI driving experiences can’t be a black box if we want people to trust it, and this concept will continue to play a huge role as we develop new technologies, Alisan Atvur said. He mentioned an AI-driven diagnostic engine which would pause when it was uncertain about something so the doctor could scrutinise it - essentially, the peer review process. “It’s something you would do with another doctor. It’s replicating some of the same ways we build trust and reliability between humans.”
Final Wrap-up
As Emotive AI becomes more and more sophisticated, it’s opening up new ways to reach and serve consumers across all sectors, whether helping a doctor provide better medical care, providing delightful customer service or boosting a luxury listening experience. Like any good technology, it’s most effective when it’s purposeful - meeting a real need, building trust with the user and staying true to the DNA of a brand.
For the full conversation, watch the video above or head to ustwo’s YouTube channel.
Change is coming fast. At ustwo, we’re embracing the potential of AI, while approaching it with a human-first lens that aligns with our values like safety and responsibility.
Want to chat more about the future of AI? Reach out to us at hello@ustwo.com