JHDD UI UX Report — 2026.10.03

Erika Hall, cofounder of Mule Design, asks what gives designers the right to make interventions on behalf of others, a question directly challenging the industry’s assumed authority.

This question, alongside the re-evaluation of performance thresholds and the fracturing of consensus around AI, reveals a widespread, unaddressed instability in fundamental interaction patterns. Designers are not just integrating new technologies; they are navigating a foundational redefinition of user expectations and professional responsibilities that current frameworks struggle to contain. The articles highlight a sector simultaneously advancing technologically and grappling with ethical and usability principles that are no longer adequate for these advancements.

Walter Doherty’s observation about the 400-millisecond threshold, stating that the fix for slow AI often lies in a week of front-end work rather than cheaper inference, presents a view contradicting mainstream industry opinion. Most teams default to blaming model complexity or computational cost when AI interactions feel sluggish. This perspective sidesteps the designer’s responsibility for managing the user’s perception of time and system responsiveness. The problem is often framed as a technical one, ignoring the interaction pattern design challenge of communicating system state and managing user anticipation. The mainstream focus on optimizing AI models for raw speed misses that the critical design problem is the experience of waiting or the design of collaborative interaction patterns that account for varied latency.

The true problem is not just the model’s speed, but the failure to design robust asynchronous interaction patterns or to define “done” appropriately for AI-driven tasks. By mid-2027, design systems will begin to formally codify AI-specific interaction patterns that incorporate non-deterministic response times, moving beyond simple spinners to more nuanced visual and auditory cues that manage user cognitive load during AI processing. These patterns will become standard components, explicitly detailing how to design for generative output, feedback loops, and varying levels of AI agency.

The primary opposing force to these necessary re-evaluations is the internal fragmentation of understanding within the design community itself, as evidenced by the observation that “Design’s consensus was AI’s first casualty.” Without a shared understanding of AI’s implications for interaction design, usability, and ethics, disparate approaches will continue to undermine the adoption of robust, user-centered patterns. This lack of collective agreement slows the formalization of new best practices.

A working UI UX professional should, this week, integrate an “ethical pause” into their design sprints. Before implementing any new AI-driven interaction, conduct a mini-audit using Erika Hall’s core question: “What gives us the right to make this intervention on behalf of others?” This forces a structured discussion on potential harms, power dynamics, and the true necessity of the proposed interaction pattern, extending beyond standard usability testing to consider broader societal impact.

---

Curated References