JHDD UI UX Report — 2026.09.11

A shopper at Bossa Nova Robotics attempted to scan blueberries against an autonomous inventory robot, illustrating a fundamental challenge in current interaction design.

This scenario, alongside Anthropic’s realization that its Claude Code engineers, shipping at triple their headcount, created a bottleneck not in coding but in deciding what to build, points to a pattern of systemic interaction ambiguity. The industry is not merely shifting from explicit interfaces to intent-driven experiences; it is confronting the deep, unarticulated human intent that underlies all interaction, whether with a physical robot or an AI system. The problem is consistently less about the machine’s capability and more about the human context, expectations, and the machine’s ability to interpret or respond to those nuances.

The prevailing narrative often suggests that “the best interface is no interface,” implying a simplification of the user experience and, by extension, the design process. This perspective is dangerously misdirected. While the visible buttons and forms may recede, the complexity of designing for user interaction intensifies dramatically. The challenge is not removed; it is transmuted into a more profound need for understanding and designing transparent feedback, robust error states, and clear pathways for user clarification when AI misinterprets intent. Consider Anthropic’s experience: the sheer efficiency of Claude Code in generating output exposed the prior lack of clarity in intent and planning. This suggests that “no interface” does not eliminate design work, but rather shifts it upstream to the critical tasks of user research and definition of human-system communication protocols. Within two years, leading design systems will begin to integrate robust “intent clarification patterns” as standard components, moving beyond basic alert messages to truly help users diagnose and correct AI misinterpretations.

The widespread perception of AI as a magical solution often overlooks the messy realities of human-centered automation. The push for seamless, invisible interaction, championed as “the death of the button,” frequently underestimates the human need for control, predictability, and discoverability. As the “Most AI problems are really human problems” article notes, issues stem from “unread intent, missing oversight, thin context, loose language, unset expectations.” These are not technical AI failures; they are human interaction design failures at a fundamental level. Achieving a truly “no interface” experience that is also usable and accessible demands an increase in design effort, not a reduction, focused on anticipating user mental models and creating reliable implicit feedback mechanisms.

The primary opposing force to effective interaction design in this evolving landscape is the pervasive expectation of “magic” by product leadership and engineering teams. This expectation often prioritizes the perceived seamlessness of an invisible interface over the critical human need for understanding, agency, and clear feedback. This leads to deploying systems like the Bossa Nova Robotics inventory robot without sufficient attention to how humans will inevitably interact with and misinterpret their functionality, because the focus was on the robot’s primary task, not the ambient human interaction.

A working UI/UX professional should dedicate focused time each week to auditing a “smart” system—whether a voice assistant, an automated checkout, or a personalized feed—specifically to identify points of ambiguity where the system’s intent is unclear, or where a human’s intent might be misinterpreted. Document these ambiguities and prototype micro-interactions or feedback cues that clarify intent without adding explicit buttons or steps, prioritizing the design of expectation management and error recovery over purely frictionless interaction.

---

Curated References