Google’s Gemini has a branding problem, and so does the rest of AI
In its recent announcement regarding the new Gemini Live voice capabilities, Google made a bold promise: “You shouldn’t have to guess whether a task requires Spark, a Daily Brief, or a quick inbox search.” While this sounds like a vision for a seamless, intuitive future, the reality of the current Gemini app suggests otherwise. Google is actively undermining its own message by forcing users to navigate a fragmented landscape of branded features, each with its own icon, navigation tab, and specific purpose.
The Complexity of "Clever" Design
The current iteration of the Gemini app feels less like a unified assistant and more like a collection of disparate tools struggling to find a cohesive identity. Features like "Spark" and "Daily Brief" are prime examples of engineering-led design—solutions that seem brilliant to the developers who built them but feel cumbersome to the everyday user.
Take the Daily Brief feature, for instance. Designed as an AI-powered agenda, it attempts to pull data from across the Google ecosystem, including Gmail and Calendar, to provide proactive updates. In practice, however, it struggles to distinguish between high-priority, actionable information and unsolicited digital clutter. It often resurfaces past search queries or ongoing chatbot research, which can feel less like a helpful assistant and more like an intrusive, "creepy" reminder of things you’ve already moved past.
Conversely, Spark represents a highly functional tool—an AI agent capable of performing tasks on the user's behalf—yet it is siloed behind its own branding. Users shouldn’t need to decide which "mode" of an app they are in to get a job done. The ideal experience is simple: the user makes a request, and the AI determines the best way to execute it, spinning up an agent in the background if necessary.
"Consumers are being asked to learn the brand names for what are essentially interaction modes or surfaces, powered by a company’s AI model. This is the kind of engineering-minded design that makes engaging with AI feel unnatural."
A Systemic Industry Failure
Google is far from alone in this architectural oversharing. The broader AI industry is currently obsessed with exposing internal product structures to the end user. Consider the current state of the market:
- Anthropic’s Claude: Users are forced to choose between "Chat" and "Cowork" modes, which, until very recently, did not even share a unified memory of previous interactions.
- ChatGPT: The platform requires users to manually toggle between "Chat" and "Work" interfaces, adding unnecessary friction to the user experience.
By forcing consumers to learn the "architecture" of an AI product, companies are creating a steep learning curve that discourages adoption. Instead of focusing on the utility of the AI, users are forced to focus on the interface of the software.
The Case for Simplicity
If the current crop of AI giants continues to complicate their interfaces, they may find themselves losing ground to more streamlined competitors. The success of Apple’s approach to Siri serves as a cautionary tale for the industry. Apple isn't asking users to learn a new "AI mode." Instead, it is quietly integrating intelligence into the tools people already use—Photos, Camera, Spotlight Search, and voice commands—without requiring a change in behavior.
Similarly, the rise of text-based AI services—such as Poke, Ollie, Lindy, Orchid, Lucas, Folk, Tomo, and Instinct—highlights a growing consumer preference for simplicity. These services leverage the most familiar interface on the planet: the text message. As investment partner Justine Moore recently noted, the goal isn't to force users into a complex new app, but to provide a contact they can message like a friend.
Key Takeaways for the Future of AI UX:
- Hide the Architecture: Users should not need to know how an AI is built or which "mode" it is currently running in.
- Prioritize Intuition: The best AI interface is one that requires zero training.
- Context is King: Proactive updates must be actionable and relevant, not just a stream of past data.
- Unified Memory: AI assistants should maintain a consistent context across all tasks, regardless of the underlying tool being used.
Ultimately, the companies that win the AI race will not be the ones with the most complex branding or the most granular "modes." They will be the ones that make the technology disappear, leaving behind only a seamless, helpful experience that feels less like a software product and more like a natural extension of the user's intent.