The AI layer: transforming UX design from tools to intelligence

A practical framework for implementing AI as a foundational layer in digital products

Figure 1: Conceptual representation of AI as a horizontal enabling layer, illustrating its transformative nature similar to electricity in the 20th century. The bright horizontal line represents the AI layer that powers and connects multiple applications and systems, represented by interconnected nodes. Image generated via Krea.ai.

Introduction

“AI is a horizontal enabling layer — it can be used to improve everything. It will be in everything,”

Jeff Bezos declared, comparing AI to the transformative power of electricity.

“These kinds of horizontal layers like electricity, compute, and now artificial intelligence, they go everywhere. I guarantee you there is not a single application that you can think of that is not going to be made better by AI.”

Just as electricity revolutionized every industry by becoming a foundational utility, AI is fundamentally reshaping our digital landscape in ways that go far beyond simple automation. At Amazon alone, teams are working on “literally a thousand applications internally,” demonstrating AI’s potential for widespread integration.

As UX professionals, we’re at the forefront of this transformation, tasked with creating interfaces that make these powerful capabilities accessible and meaningful to users. Through an analysis of pioneering applications, we can establish a framework for designing AI-driven products. These examples serve as both a conceptual model and a practical guide for structuring products where AI acts as a foundational layer. Beyond showcasing technological advancement, this analysis provides a systematic approach to identifying AI opportunities within your own products.

Case 1: Perplexity AI

From manual information assembly to intelligent discovery

The Challenge: Breaking free from traditional search

“AI is the first truly new interaction-design paradigm in 60 years,” observes Jakob Nielsen, and nowhere is this more evident than in information search.

Traditional search engines, despite their sophistication, still required users to master a complex dance: crafting precise queries, scanning multiple results, clicking through various pages, and mentally synthesizing information. This process, while familiar, placed a significant cognitive burden on users.

Figure 2: Traditional search interface. A Google search interface showing the traditional approach where users must manually scan, filter, and synthesize information from multiple sources. Note the cognitive load required to process and connect information.

The transformation: Reimagining search with AI

Perplexity AI embodies this transformation by fundamentally reimagining how humans interact with information. Instead of users adapting to the system’s requirements, the AI adapts to user intent.

Figure 3: Perplexity AI Interface. Perplexity AI’s interface demonstrates the shift from tool-based search to intelligent discovery. Note how it automatically synthesizes information and presents structured comparisons.

How it works: The intelligence layer

The system processes queries through interconnected layers that understand context, verify information in real-time, and present synthesized answers that feel natural and conversational.

Figure 4: System architecture diagram. Perplexity AI’s information discovery system, showing how natural language queries are processed through multiple AI layers for comprehensive, verified answers. Source: Author.

Key Components:

• Query understanding layer: Processes natural language input
• Information synthesis layer: Connects and verifies multiple sources
• Response generation layer: Creates coherent, contextual answers
• Interaction management layer: Maintains conversation flow

Impact on user experience

The result isn’t just a faster search engine — it’s a new paradigm for knowledge discovery that feels more like consulting a knowledgeable colleague than operating a digital tool. Users can:

• Ask questions naturally without worrying about keywords
• Receive comprehensive, synthesized answers
• Follow up with contextual questions
• Verify sources and facts in real-time

This transformation shows how AI can fundamentally change core user interactions, moving beyond mere automation to create truly intelligent systems that adapt to human needs rather than requiring humans to adapt to them.

Case 2: NotebookLM

From document management to knowledge orchestration

The challenge: Beyond digital paper

Traditional research and note-taking tools merely digitized the paper experience while maintaining its fundamental limitations. Knowledge workers and researchers faced persistent challenges:

• Manual information organization
• Limited connections between documents
• Difficulty maintaining coherent management systems
• Cognitive overload when processing multiple sources

Figure 5: Traditional research process. Traditional document management approach showing multiple windows, manual note-taking, and disconnected information sources. Researchers must manually create connections and maintain organization systems. Image generated via Krea.ai.

The transformation: AI as a knowledge partner

NotebookLM transforms this experience by implementing AI as an intelligent collaboration layer. Instead of being a passive repository, the system:

• Actively participates in the knowledge work process
• Automatically maps relationships between documents
• Suggests connections you might have missed
• Adapts its organization to your thinking patterns

Figure 6: NotebookLM Interface. NotebookLM’s dynamic workspace showing AI-powered suggestions, automatic relationship mapping, and contextual document organization.

How it works: The intelligence layer

NotebookLM transforms document management through a three-layer intelligence system. The User interaction layer handles direct user engagement through document uploads, natural queries, and note-taking capabilities. The Knowledge processing layer — the system’s core — analyzes documents, maps connections, and synthesizes information using advanced AI algorithms. Finally, the Intelligent output layer presents this processed information as connected insights, related concepts, and research suggestions, creating a dynamic system that actively enhances the research and learning process.

Figure 7: Knowledge orchestration system. Visualization of NotebookLM’s approach to transforming document management into an intelligent knowledge system. Source: Author.

Impact on research and knowledge work

This transformation creates a dynamic workspace that:

• Feels less like a digital filing cabinet and more like a thinking partner
• Enhances natural workflow while maintaining flexibility
• Reduces the cognitive load of organization
• Facilitates serendipitous discoveries

NotebookLM represents a fundamental shift in how we interact with information: from passive document management to active collaboration with a system that understands and amplifies our thought processes. This transformation demonstrates how AI can serve as more than just a tool — it becomes an intelligent partner in the knowledge work process, while maintaining the user’s autonomy and enhancing their natural cognitive workflows.

Case 3: Genway AI

From individual interviews to scalable intelligence

The challenge: Breaking traditional research boundaries

The traditional UX research process has remained largely unchanged for decades, requiring intensive manual effort at every stage. Researchers face significant limitations:

• Time-consuming participant recruitment
• Limited interview capacity
• Manual analysis of responses
• Labor-intensive synthesis of findings
• Trade-offs between depth and breadth of research

Figure 8: Traditional UX Research Session. A traditional user research session showing researchers conducting a one-on-one interview. This manual approach, while thorough, illustrates the core limitation of traditional UX research: the inability to scale without compromising depth of insights. Image generated via Krea.ai.

The transformation: AI-enabled research at scale

Genway AI revolutionizes this paradigm by implementing AI as a horizontal enabling layer across the entire research workflow. The system:

• Conducts multiple human-like interviews simultaneously
• Processes multiple data streams in real-time
• Analyzes responses across voice, text, and video
• Synthesizes insights automatically while maintaining research integrity

Figure 9: Genway AI Interface. Genway AI’s interface showing parallel interview processing, real-time analysis, and automated insight generation while maintaining human oversight.

How it works: The intelligence layer

Genway AI operates through a comprehensive three-layer system that transforms traditional user research into a scalable intelligence operation. The Data collection layer captures multiple streams of user input through voice, text, and video channels simultaneously. The AI processing core analyzes this data in real-time, recognizing patterns, performing sentiment analysis, and generating insights. The Research insights layer then delivers both quantitative analysis and qualitative insights, automatically detecting trends and providing strategic recommendations.

Figure 10: Research intelligence framework. Visualization of Genway AI’s approach to scaling user research, showing how multiple data streams are processed simultaneously through AI analysis layers. Source: Author.

Impact on UX research

This transformation fundamentally reshapes UX research capabilities through:

Scale with depth: Enables simultaneous processing of hundreds of interviews while maintaining the nuanced understanding essential to quality research. The system can process multiple data streams without sacrificing the depth of analysis traditionally associated with one-on-one interviews.

Real-time intelligence: Combines immediate insight generation during research sessions with sophisticated pattern recognition across large datasets, enabling researchers to adapt and refine their approach during the research process.

Augmented expertise: Creates a balanced synergy between automation and human expertise, where AI handles data processing and pattern identification while researchers focus on strategic interpretation and decision-making.

Genway AI demonstrates how artificial intelligence can transform UX research from a linear, resource-constrained process into a dynamic, scalable system that amplifies rather than replaces human research capabilities. This transformation maintains the nuanced understanding essential to user research while dramatically expanding its scope and efficiency.

The impact on UX design and future implications

Our analysis of these groundbreaking applications reveals fundamental patterns reshaping the future of UX design. Just as electricity transformed every industry it touched, AI is creating new paradigms for how we think about and design digital experiences.

As Jakob Nielsen observes, “the best design for AI will retain some of the old graphical user interface elements, resulting in a hybrid UI, mostly based on user intent. But iterations including tweaks and revisions, specified through GUI commands.” This insight reveals a crucial principle for designing AI-powered products: the need to balance revolutionary capabilities with familiar interaction patterns.

How AI is reshaping digital experiences

AI isn’t simply adding features to existing tools — it’s fundamentally transforming how we design digital experiences. As Henry Modisett, head of design at Perplexity AI, emphasizes, “This technology is just going to be available in everything and everywhere. It’ll just be a way to enable some core product experience. It’ll make some new software that’s amazing, and it’ll accelerate some old software.”

This transformation operates on three integrated levels:

1. Understanding user intent, even when expressed imperfectly
2. Managing complex processing invisibly
3. Adapting interfaces dynamically to user needs

We see this already in action through tools like Perplexity AI, which has transformed information search into natural conversation, and NotebookLM, which actively discovers connections across documents that users might miss. These aren’t merely faster versions of existing tools — they represent entirely new paradigms for human-computer interaction.

Building trust through smart design

The key to successful AI integration lies in balancing transparency and user control. Each of our case studies demonstrates this principle in action; Perplexity AI shows its sources in real-time, NotebookLM visualizes its thought process when connecting ideas across documents, and Genway AI maintains transparency in its research data analysis. These implementations showcase different levels of AI involvement while ensuring users maintain meaningful control.

This balanced approach manifests through specific design choices:

1. Verification mechanisms: Users can verify sources directly in Perplexity AI.
2. Selective adoption: NotebookLM allows users to accept or reject AI-suggested connections.
3. Expert oversight: Genway AI enables researchers to validate AI-generated insights.

By making AI’s role visible and keeping users in control, these systems create the foundation of trust essential for effective human-AI collaboration while maximizing the benefits of AI capabilities.

Conclusion: The path forward with AI in UX design

The transformation we’re witnessing in UX design isn’t just another technological shift — it’s a fundamental reimagining of how humans interact with digital products. We’re moving from an era of digital tools to one of intelligent systems, where AI acts as a horizontal enabling layer that enhances and transforms every aspect of the user experience.

Looking ahead, UX teams need to focus on three key priorities:

1. Designing with intention: Move beyond surface-level AI integration by identifying where AI can most meaningfully transform your user’s experience, shifting from standalone tools to interconnected intelligent systems.
2. Building trust through transparency: Apply frameworks and design patterns that clearly communicate AI’s role and limitations to users, making the intelligence layer visible and understandable.
3. Preserve human autonomy: Ensure users maintain meaningful control and understanding of AI-driven features, creating a symbiotic relationship between human insight and AI capabilities.

The future of UX belongs to designers who can strike the right balance between AI’s capabilities and human needs. Success will come not from maximizing AI usage, but from thoughtfully integrating it as an intelligence layer that augments human capabilities while preserving what makes us uniquely human: our creativity, empathy, and ability to make nuanced judgments based on context and values.

The question isn’t whether AI will transform your product’s user experience — it’s how you’ll lead that transformation. Every UX team now faces the opportunity to evolve their products from collections of tools into intelligent systems that adapt, learn, and grow with their users.

Are you ready to reimagine what’s possible?