Reimagining Perplexity: A more accessible AI search experience
Exploring a new Perplexity AI interface that breaks down accessibility barriers.
In the rapidly evolving landscape of AI-powered search engines, Perplexity AI has emerged as a frontrunner, offering users a conversational and source-cited approach to information retrieval.
However, as we push the boundaries of AI technology, it’s crucial to ensure that these advancements are accessible to all users, including those with reading and writing barriers.
Jakob Nielsen (Co-founder of Nielsen Norman Group) expects that in countries like the United States, Northern Europe, and East Asia, less than 20% of the population is sufficiently articulate in written prose to make advanced use of prompt-driven generative AI systems.
“For sure half the population is insufficiently articulate in writing to use ChatGPT well.“
(Jakob Nielsen, The Articulation Barrier: Prompt-Driven AI UX Hurts Usability)
Today, I’m excited to share a new prototype that aims to make Perplexity more inclusive and user-friendly for everyone. 🎉
The Challenge: Bridging the Accessibility Gap
While Perplexity AI has revolutionized how we interact with search engines, users with reading or writing difficulties may still find it challenging to formulate precise queries or navigate through lengthy responses. This accessibility gap can significantly limit their ability to harness the full potential of AI-powered search tools.
A Solution in Sight: An Interactive, Assistive Interface
To address this challenge, I’ve developed a prototype using Figma that introduces a new user interface for Perplexity. This interface is designed to make intelligent suggestions and ask users for more detailed information about their initial prompt, especially when the input is brief or vague.
This is an example of a hybrid user interface that Jakob Nielsen mentions in his article “AI: First New UI Paradigm in 60 Years”. He says: “Future AI systems will likely have a hybrid user interface that combines elements of both intent-based and command-based interfaces while still retaining many GUI elements.”
How It Works
1. Smart Prompt Analysis: The system quickly analyzes the user’s initial input.
2. Guided Refinement: If the prompt is short or lacks specificity, the interface presents a series of simple follow-up questions to help refine the query. The step-by-step questions help to reduce cognitive load during interaction and bring up potentially forgotten details. A highly specific, tailored prompt is generated automatically behind the scenes.
3. Visual Aids: The interface incorporates icons, color-coded states and crisp formulations to assist users who may struggle with text-based interactions.
4. Concise Prompt Summary: To not overwhelm the user with its lengthy, tailored prompt, an AI-generated summary is shown instead of the prompt itself.
Figma Prototype and Design File
I’ve made the Figma prototype and design file publicly available for anyone who wants to take a closer look.
Prototype
You can view and interact with the full prototype. Simply click 👉 here 👈 to experience the new interface design firsthand.
Figma Community File
Interested in exploring the new hybrid user interface in detail? Click 👉 here 👈 to open the full design file directly in Figma.
Benefits of the New Design
This reimagined interface offers several key advantages:
- Enhanced Accessibility: Users with reading or writing difficulties can more easily interact with the AI, breaking down barriers to information access.
- Improved Query Accuracy: By guiding users to provide more context, the system can generate more precise and relevant responses.
- Reduced Computational Costs: The user-guided AI-optimized prompts can be achieved using smaller language models (SLM). They minimize user interaction and lead to fewer iterations with large language models (LLM), enabling precise, efficient results at reduced computational costs.
(See also: OpenAI is losing money) - Reduced Cognitive Load: The step-by-step refinement process helps users articulate their needs without feeling overwhelmed.
- Increased User Confidence: With a more supportive interface, users may feel more empowered to explore and learn.
Looking Ahead: Implementation and Impact
While this prototype represents an exciting step forward, implementing such a system would require careful consideration of several factors:
- Seamless integration with Perplexity’s existing AI technology
- Optimizing performance to maintain quick response times for follow-up questions
Despite these challenges, the potential impact of this interface is significant. By making AI-powered search more accessible, we can democratize access to information and empower a wider range of users to benefit from cutting-edge AI technology.
As we continue to push the boundaries of what’s possible with AI, let’s ensure that we are creating tools that are not just powerful, but also inclusive and accessible to all. This prototype for Perplexity is just one step in that direction, but it’s a step that could make a world of difference for many users.
What are your thoughts on making AI search more accessible? I’d love to hear your feedback and ideas in the comments below!
Disclaimer: This prototype is a personal project created independently. I am not affiliated with Perplexity AI, nor am I receiving any compensation. As a passionate user of the platform, I developed this concept purely out of my own interest in improving AI accessibility.
