Eunbee Lee As artificial intelligence continues to evolve, its role in typeface design is shifting from simple automation to deeper contextual understanding. This session explores how AI can extend, rather than replace, the creativity of type designers by intelligently interpreting textual context, design principles, and user intent. Rather than merely following typographic rules, can AI truly understand the contextual background, intent, and meaning behind typography?
Building on my research in AI-generated typefaces and AI-assisted poster design, this session will examine cutting-edge AI models such as FontRNN and DeepFont, which generate typefaces based on learned patterns, as well as the potential of GPT-4 and generative models like DALL-E as a starting point to explore how AI could be trained to generate typography that truly aligns with specific moods, contexts, and purposes. While these models can produce visually compelling results, they often lack the ability to integrate typographic meaning with textual intent. To bridge this gap, I analyzed case studies where AI-assisted type design has been applied in branding, editorial design, and UI/UX applications.
By drawing on existing research and experimental projects, I will investigate the current capabilities of AI in typography, examining its advancements, limitations, and possible solutions.
Additionally, this session will explore how AI-driven tools could enhance creative workflows by automating repetitive tasks and fostering a more interactive, collaborative relationship between designers and machines. Building on these aspects, I will then explore how AI might evolve further to suggest novel typographic compositions, pushing the boundaries of creative experimentation and personalized type design.
Through presentation and discussions, this presentation will offer concrete insights into the future of AI-assisted typography. Attendees will gain a clearer understanding of AI’s current strengths and weaknesses, practical applications in professional workflows, and future directions for AI-integrated type design.