A graphic designer's experiment with integrating Perplexity AI into their creative workflow has revealed both promising benefits and notable limitations. The designer tested the AI tool as a design guide across popular software platforms Krita and Figma, discovering that while AI can streamline certain aspects of the design process, it still requires human oversight for optimal results.
Traditionally, the designer's approach involved piecing together information from various sources including tutorials, forums, and trial-and-error experimentation within graphics tools. While this discovery process was enjoyable and often led to satisfactory results, it frequently became scattered and inefficient. Inspired by colleagues who had successfully paired AI with productivity tools, the designer decided to test whether Perplexity could serve as a structured design guide rather than a creative replacement.
The initial setup process proved crucial to the experiment's success. Rather than treating Perplexity as a simple search engine, the designer learned that proper configuration and thread structuring significantly impacts the quality of responses. Since Perplexity possesses contextual memory and can personalize answers based on user information and previous queries, careful preparation was essential. Using the free version of the platform, the designer revisited personalization settings to ensure all objectives and information were clearly defined.
To optimize the AI's performance, separate threads were created for each design tool, plus an additional thread for general workflow guidance. Before beginning actual design work, the designer asked basic questions in each thread to train the AI, covering fundamental concepts like drawing basics in Krita. This preliminary training helped attune the AI to the specific types of assistance that would be needed throughout the design process.
The practical application began with workflow-focused prompts that explored design principles and style considerations. Questions like "What are the design differences between a modern editorial layout and a minimalist one?" and "What visual cues make something feel trustworthy in meditation apps?" helped establish a design landscape. One particularly useful response highlighted rounded edges and natural imagery as key elements, providing an excellent starting point for the creative work.
For the illustration component, the designer used Krita to create a fish design, representing the natural imagery element. Rather than seeking basic instruction, Perplexity served as a creative sparring partner for more nuanced decisions. When asked "What's a stylized way to depict fish fins without going hyperrealistic?" the AI provided valuable guidance that informed the artistic choices. After completing the fish illustration, the challenge became integrating this asset into a mobile UI design using Figma.
The transition from illustration to interface design presented new challenges, prompting another strategic query: "Suggest ways to integrate a fish illustration into a mobile UI so it complements the layout without overwhelming usability, including placement, size, and color." Perplexity delivered comprehensive recommendations that addressed positioning, sizing, and color considerations. This guidance proved particularly valuable in implementing design elements like rounded corners and determining optimal placement for the illustration within the mobile layout.
Throughout the iterative design process, the designer consistently returned to the established threads for advice and direction. This approach significantly reduced decision fatigue and enabled smoother transitions between design phases. The Images section within Perplexity's results proved especially valuable for design-related queries, providing visual references and inspiration that complemented the text-based guidance.
However, the experiment revealed significant limitations in AI-assisted design work. Despite following Perplexity's recommendations for calming colors and rounded corners, the AI failed to recognize that the complex fish illustration with numerous lines and edges contradicted the intended minimalist and relaxing aesthetic. Even when the designer provided detailed prompts with attached image files describing the specific issues, the AI continued recommending minimalist techniques that clashed with the illustration's inherent complexity.
This disconnect highlighted a fundamental limitation of current AI technology in creative applications. While AI can interpret prompts and reference established patterns, it cannot visually analyze designs to provide contextual feedback. The AI's inability to "see" the actual design work means it relies entirely on user descriptions and defaults to known patterns, often missing crucial visual elements that human designers would immediately recognize as problematic.
The designer ultimately resolved the conflict by creating a more minimal fish illustration from the beginning, demonstrating that human judgment remains essential in creative decision-making. This solution required recognizing that the original design approach was fundamentally incompatible with the intended aesthetic goals, something the AI could not identify despite detailed explanations.
The overall assessment of using Perplexity for graphic design work reveals a mixed but ultimately positive outcome. The AI excels at providing foundational design principles, workflow structure, inspiration, and tool-specific guidance. Since this information already exists in various forms online, Perplexity's strength lies in aggregating data from real-time web searches and proprietary indexes, then presenting it in an organized, contextually relevant manner.
The designer concluded that Perplexity offers significant value for learning design tools, exploring color palette ideas, discovering keyboard shortcuts, and understanding design principles. However, successful implementation requires applying human visual judgment to achieve quality results. The AI serves best as a knowledgeable assistant for information gathering and initial guidance, rather than as a replacement for human creative insight and visual assessment skills that remain essential for effective design work.
