Personalization is where AI excels, and why Gemini’s move toward personalization is significant for conversational AI, more specifically. AI’s ability to tailor experiences through conversational tools, predictive analytics, and curated content makes interactions seamless and relevant. Consider how well Netflix builds engagement through personalization and Spotify uses AI to enrich music discovery in ways that improve the listening experience. As a result, both have attracted and kept hundreds of millions of subscribers.
Empowerment, on the other hand, is about AI giving people agency. Apple’s photo Clean Up feature (available for more recent iPhone models) helps photographers edit their work by removing unwanted objects from photos, like holding Photoshop in your hands.
More generally, AI empowers users every day in ways that are less obvious but still important. For example, in automobiles, adaptive cruise control uses AI with radar and cameras to adjust speed based on traffic. The driver sets the speed and distance preferences; AI handles microadjustments, freeing up the driver to focus on steering and decisions, like a co-pilot.
Bringing consumers along for the ride
Conversational AI has an opportunity to earn our trust with more transparency. As it stands, consumer trust in generative AI is a mixed bag, shaped by a blend of enthusiasm for their capabilities and skepticism. A 2024 SEMRush survey reported widespread user distrust in the accuracy of generative AI search results. Some 60% of users surveyed wanted more security by getting more accurate data and sources included in search results. In other words, conversational AI has a challenge in the “value” component of trust.
Consumers don’t need AI to be perfect, but they do need to understand its imperfections, and they need to know what conversational AI providers are doing to improve accuracy and inject rigor. Perplexity footnotes its AI-generated answers, allowing users to see where information is sourced. But even so, Perplexity makes mistakes by drawing wrong conclusions from those sources. AI providers could build trust by offering real-time error tracking or public-facing reports that document both progress and challenges.
Beyond transparency, trust flourishes when consumers feel they have a seat at the table. Providers of generative AI apps can earn trust by showing how they empower people to get better at what they do, not replace our work. AI companies could also involve users in co-creation, whether through feedback loops, beta testing, or personalized customization. An AI assistant that periodically checks in with a “Did I get that right?” is the kind of participatory trust-building that turns AI from a black box into a collaborator.
