From: aidotengineer
The AI Accelerated Need Finding Process is a proposed methodology aimed at empowering designers to leverage AI simulation and “invisible users” to enhance the design lifecycle, particularly in the early stages of need finding [03:56:00]. This approach seeks to accelerate design and improve user experience by addressing the current “AI trust gap” [01:51:00].
The AI Trust Gap
Currently, AI faces a significant trust gap among users [00:35:00]. Recent research from Edelman in December 2024 revealed that only 32% of US adults trust AI, and only 44% of adults globally feel comfortable with how businesses are using AI [00:37:00]. This lack of trust is largely attributed to “AI slop,” which refers to instances where products or interfaces powered by generative AI produce incorrect or nonsensical results [00:56:00]. Examples include web searches suggesting eating rocks daily or websites offering cars for $1 [01:13:00].
While many website creators are “stuffing AI chatbots” everywhere and claiming it’s “magic,” the true potential of generative AI lies in its ability to enable users to talk to a machine learning model in natural language—a significant UX revolution [01:25:00].
Invisible Interfaces and AI Simulation
Drawing on UX first principles, specifically Don Norman’s concept of “invisible interfaces,” the goal is to create software so seamless and intuitive that users practically forget they are using it [02:18:00]. The opportunity with AI is to design interfaces that genuinely feel like magic, not through poorly implemented chatbots, but by accelerating the need finding process [02:31:00].
The traditional data-driven design process, which predates ChatGPT, relies on designers collecting qualitative observations, quantitative data, and ethnographic insights to guide prototyping [03:33:00]. The proposed new approach involves:
- Empowering designers with AI simulation: Using “invisible users” in the form of AI simulation [03:58:00].
- Turning data artifacts into active participants: Transforming collected data into active elements in the design process [04:04:00].
- Creating mini feedback cycles: Providing designers with their own feedback loop using AI simulation within the broader need finding process [04:11:00].
The AI Accelerated Need Finding Process
The new design lifecycle with AI simulation shares similarities with the existing process but introduces key differences in components and workflow [04:56:00].
Process Steps:
- Define Audience with Data: Start with data representing target audiences, including demographic, psychographic, and contextual datasets, to simulate audience behaviors [05:06:00]. Companies like Huge use platforms (e.g., “Live”) for this foundation [05:14:00].
- Apply Intent Mapping with Intelligent Twins: Transform audience data into “intelligent twins”—simulations that represent user behaviors, desired outcomes, needs, and motivations [05:34:00]. These twins become active participants in the design simulation [05:49:00].
- Brief Intelligent Twins for Task Evaluation: Intelligent twins can be briefed to evaluate interfaces by focusing on specific tasks, similar to how a human designer might conduct a heuristic analysis [05:56:00].
Advantages of the Methodology:
- Scale and Speed: Intelligent twins can operate with different velocities than human teams, allowing for global audits and rapid analysis [06:23:00].
- High-level and Granular Insights: Insights can be rolled up to a high level for broad understandings across categories or explored more granularly for specific nuances of the experience [07:50:00].
- Focused Design Briefs: The ability to operate at different “altitudes” allows for the creation of very focused or broad and deep design briefs [09:18:00].
Case Study: Global Sports Website Audit
An example project demonstrated this AI-accelerated design audit by evaluating global sports websites [06:36:00].
- Objective: Understand how to engage audiences across different countries and cultures for a business partnering with sports leagues [06:41:00].
- Audience Personas: Two personas were created for the intelligent twins: a “casual fan” new to the sport and a “super fan” who is lifelong and savvy [07:06:00].
- Tasks: Across three different sports websites (basketball, Olympics, English Premier League), a series of tasks were briefed, categorized into navigation, information architecture, and fan engagement, with four tasks per category [07:18:00]. This resulted in simulating 72 AI simulated actions [07:31:00].
Findings:
- Initial Success in Navigation: All audited sites performed well in initial navigation tasks [08:09:00].
- Drop-off in Deeper Engagement: Task completion rates dropped significantly when fans browsed content, information architecture, and engagement pathways [08:29:00].
- Pain Point Identification: This highlights that while initial access works, deeper user needs are not being met, contributing to the AI trust gap [08:42:00]. The methodology can identify these pain points to solve them [08:53:00].
Using computer vision models and human-in-the-loop observation, this process can generate focused design briefs, allowing human teams to concentrate on solving specific user pain points [09:36:00].
Limitations and Future Improvements
This methodology is currently in experimental early stages [10:53:00]. Key areas for improvement include:
- Reproducibility: Standardizing parameters in a code repository, such as briefing instructions, simulated audience dimensions, number of audit runs, and task completion/failure parameters [11:00:00].
- Test and Control: Applying a test and control methodology to isolate the strengths of intelligent twins for design need finding [11:24:00].
- Complementary Use: Determining how intelligent twins can be used alongside or complimentarily to human teams, especially across different industries, geographies, and domains [11:35:00].
The acceleration of design tools, such as Anthropic’s MCP protocol turning Figma prototypes into React or Node.js code components, will make creating new designs easier than ever [10:12:00]. This emphasizes the importance of focusing on the “why” (strategy and problem-solving) rather than just the “how” in the design process [10:36:00].
Conclusion
To repair the existing AI trust gap, the focus should shift from merely adding non-functional GenAI chatbots to websites to creating better websites, mobile apps, and surfaces that offer clarity and simplicity [11:57:00]. AI simulation can empower design teams to gather insights in a smarter, faster, and better way, ultimately creating interfaces that restore user trust [12:17:00].