By Azzam Bilal Chamdy 
In an industry currently obsessed with the speed of iteration, the "vibe-coding" phenomenon, and the race to shove generative AI features into every possible software interface, product strategist Anina Botha has chosen a radical path: she stopped building.
For the past six months, while her peers have been sprinting to integrate the latest LLM wrappers and prompt-engineering hacks, Botha has been sequestered in the quiet, dense world of academic research. Her objective was not to keep up with the daily cadence of tech-bro trends, but to answer a fundamental question: How do we build human-technology relationships that aren’t just functional, but inherently trustworthy?
Her findings suggest that the current industry standard—copy-pasting AI capabilities into existing interfaces—is a recipe for failure. Instead, Botha argues that we must move toward a model of "deliberate design," where invisible human behaviors are translated into visible, actionable product architecture.
The Chronology of a Strategic Pivot
The tech industry’s relationship with AI has evolved in a blur. From the initial shock of ChatGPT’s release to the current era of "agentic" workflows, the timeline has been defined by rapid, often reckless, implementation. Botha’s pivot marks a counter-cultural moment in this timeline.
Phase 1: The Six-Month Sabbatical from "Building"
Six months ago, Botha made a conscious decision to step back from the "feature factory" cycle. Her focus shifted from the output of code to the input of behavioral science. She spent this period dissecting academic literature, cataloging cognitive biases, and interpreting human-computer interaction (HCI) research.
Phase 2: Synthesis and Framework Development
Having identified that most product teams were failing to account for how humans actually process machine-generated information, Botha began developing proprietary frameworks. She moved away from the technical "how-to" and toward the philosophical "why-to," creating guidelines for teams struggling with AI adoption, trust, and the nuance of chatbot interactions.
Phase 3: The Integration into Product Strategy
In the final months, Botha began applying these academic-backed frameworks to live product environments. The goal was to prove that the "invisible" layers of human psychology—our tendency to over-rely on automation, our inherent skepticism, and our cognitive load limits—could be mapped to UI/UX components.
The Anatomy of Trust: Decoding Automation Bias
The core of Botha’s argument lies in the concept of "appropriate trust." In the context of AI, users typically fall into two traps: they either under-rely on the system (leading to low adoption) or over-rely on it (leading to dangerous automation bias).
Understanding the Cognitive Trap
Automation bias occurs when users trust a machine’s recommendation simply because it is a machine. They may lack the expertise to verify the AI’s output, or they may have grown complacent due to past successful interactions.
Botha highlights that this isn’t a failure of the user; it is a failure of design. If a user is prompted to make a high-stakes decision by an AI without any friction or "cognitive forcing," the design has failed to respect the gravity of the user’s responsibility.
The Role of "Cognitive Forcing"
To mitigate this, Botha proposes the integration of cognitive forcing functions. These are deliberate design interventions—such as mandatory review screens or clarity-focused summaries—that disrupt the user’s autopilot. By forcing a moment of reflection before a high-stakes action is confirmed, designers can transform the "invisible" state of blind trust into a "visible" state of informed decision-making.
Contextual Design: Why "One Size Fits All" is a Myth
One of the most compelling aspects of Botha’s philosophy is her insistence on context. The industry often treats users as a monolith, assuming that a sleek, high-efficiency AI feature will delight everyone equally.
Botha uses a powerful analogy: the physical architecture of a building. We don’t build stairs for everyone because not everyone can use them—some have mobility needs, some are claustrophobic, and others have strollers. Why, then, do we design software interfaces as if every user has the same cognitive load capacity and environmental context?
The "Baby Clothing" Problem
In her critique of modern product design, Botha points to the absurdity of placing niche items in inaccessible locations. If a store places baby clothing on the top floor, they ignore the fact that their core demographic—parents with strollers—is effectively locked out of that experience.
When applied to software, this logic dictates that:
- Generic design is dead: If an experience is built for "everyone," it is likely optimized for no one.
- The User’s Environment Matters: Are they using your AI tool in a high-stress office environment or a casual, mobile context? The design must reflect that state.
- Complexity vs. Simplicity: The goal is not to remove all complexity, but to ensure that the interface is not so generic that it feels disconnected, nor so complex that it creates a barrier to entry.
Implications for Product Teams
For product managers and engineers, Botha’s approach offers a sobering wake-up call. The era of "shipping AI for the sake of shipping" is coming to a close. As the novelty wears off, users will gravitate toward products that feel intelligent, safe, and contextually aware.
1. Responsibility of Design
Botha posits that if a user is not using an AI feature, the fault lies with the design, not the AI model itself. We "feed" the AI with our design choices; if we feed it a poor user experience, we shouldn’t be surprised when the output is mistrust or rejection.
2. Translating Theory to Action
The challenge for most product teams is that academic research is often dense and inaccessible. Botha’s contribution is bridging this gap. By turning abstract theories of human behavior into actionable, sprint-ready principles, she provides a roadmap for building products that respect human limitations rather than exploiting them.
3. The Future of Human-AI Interaction
The industry is moving toward a more nuanced understanding of AI, shifting from "AI as a magic box" to "AI as a collaborative tool." Botha’s work suggests that the most successful products will be those that intentionally design for the "invisible"—the biases, the emotions, and the cognitive shortcuts that users bring to the table.
Conclusion: Making the Invisible Visible
Anina Botha’s retreat from the trend-chasing culture of modern tech was not an act of withdrawal, but an act of deep engagement. By focusing on the "invisible" human behaviors that occur before the code is even written, she is advocating for a more deliberate, thoughtful future for product design.
As we move forward, the metric of success for AI features will likely shift. It will no longer be enough to measure "time saved" or "tasks automated." We will need to measure the quality of the interaction, the level of appropriate trust, and the extent to which the product respects the user’s specific context.
In the end, Botha’s message is simple but profound: we are responsible for how people perceive, use, and trust the machines we build. It is time to stop copy-pasting, start understanding, and begin designing for the human, not just the algorithm.
Ali Ikhwan 
Dwi Wanna 
Nana Muazin 