AI-Generated Websites Plagued by Generic Design, Poor UX

Image credit: Search Engine Journal

Websites designed using artificial intelligence frequently suffer from generic aesthetics and poor user experience due to several common implementation mistakes, according to recent industry discussions.

These issues, highlighted by experts including Aaron Epstein of Y Combinator and Raphael Schaad, founder of Cron, stem from how AI models are trained and subsequently generate design elements, often leading to a lack of originality and unexpected user interactions.

A prevalent problem is the emergence of generic design trends, such as ubiquitous bento-box layouts, purple gradients, and standard icons, which proliferate because AI models are trained on common examples, Epstein and Schaad noted.

AI-driven design tools often produce unoriginal visuals as large language models (LLMs) prioritize widely used design choices, thereby hindering a brand’s ability to establish a unique visual identity, according to the discussion.

User experience (UX) is frequently compromised by unexpected feedback, including superfluous animations or moving user interface elements, which disrupt expected patterns and create a frustrating experience for visitors.

Broken or confusing UX patterns also contribute to website usability problems. These can include scroll jacking, non-standard navigation schemes, and hidden functionalities that make sites difficult to operate, experts said.

The rapid adoption of design trends by LLMs means that once an effective design style emerges, it quickly becomes widespread, subsequently losing its initial originality and distinctiveness, Schaad explained.

While the concept of ‘vibe coding’ with AI can be beneficial for initial ideation, it often results in predictable and problematic designs if not carefully managed and refined by human designers, Epstein added.

Source: Search Engine Journal