Phillip Maggs on How AI Can Enhance, Not Replace, Your Creativity

TL;DR

• Figma Config 2024 showcased the latest in design and technology.
• Phillip Maggs from Superside discussed AI in design systems.
• Integration of AI tools in Figma has improved cost-efficiency and speed.
• Over a thousand projects have benefited from AI integration at Superside.
• Custom AI models and plugins like Musho in Figma enable efficient and consistent graphic generation.

Last week, the design community gathered for Config 2024, where Figma showed off new features to help designers and developers work faster.

The big highlight was their new AI features, which can perform repetitive tasks like naming layers in bulk or generating images for designs. This means product design teams can spend more time on the creative parts of their projects.

One of the standout speakers was Phillip Maggs, the Director of Generative AI at Superside. He shared how AI is transforming design systems and improving speed and consistency in projects at Superside. By automating routine tasks, their team can now focus on the finer creative details.

Understanding how AI fits into design is crucial for keeping up in today's fast-changing tech world. Phillip’s tips from Config 2024 are super useful for both agencies wanting to use AI in their workflow, and designers looking to improve their own workflows in their current orgs.

Design System as a knowledge base for generative AI

Phillip started by rethinking how we see design systems. Instead of just a set of rules, he suggests we view them as dynamic knowledge bases. This means they're not just static guidelines but active, evolving entities that can generate and articulate on their own.

At Superside, they use knowledge graphs and large language models to open up new possibilities. Here's how they do it:

• Knowledge Graphs: A knowledge graph is a data structure that organizes information into nodes and edges. They map out design system data, making it easier for AI to understand and use.
• Semantic Mapping: Semantic mapping is the process of linking data to meaningful concepts and relationships, making it easier for AI to interpret and generate accurate outputs.

By connecting design system data, brand data, and custom image models, Superside creates richer, more detailed outputs. This approach goes beyond simple automation and actually fully integrates AI to boost creativity while maintaining control and quality.

Figma is central to this process at Superside. They use it as a powerhouse for various creative services, like:

• Digital Ads
• Branding
• Motion Design

Integrating AI plugins directly into Figma gives designers pixel-level control, enhancing their workflows without needing new design tools.

Superside has completed over a thousand projects with AI assistance, achieving significant gains in cost-efficiency and design decisions while maintaining high quality. In turn, their team can now deliver more value to their customers, create a better user experience, and take on more projects.

What came next was putting the AI tools into practice and getting them aligned with the workflow.

Seamlessly integrating AI into design systems

In the last year, Superside used various generative AI tools, models, pipelines, and processes to achieve the impressive results they are producing. Here are the main points he focused on:

• Lots of Experimentation: Superside continues to test hundreds of different AI tools. While many are promising, they found that getting pixel-level control within Figma is where the magic happens.
• Increasing Output: By implementing AI tools, the team was able to allocate resources to more areas, which enabled them to take on a thousand new projects.
• Big Savings and Speed: With AI tools, Superside achieved 40-60% cost savings. They've also sped up their design and development, getting projects done faster while maintaining quality.

This is where Musho came into the picture. Superside uses AI tools like Musho for generative designs and to make their creative process smoother and more efficient.

These AI tools help deliver more value to their customers by allowing designers to generate brand-specific graphics efficiently. Here's how they do it:

• Direct Integration: Custom AI models live in Figma, enabling designers to work seamlessly without switching tools.
• Efficient Graphic Generation: Designers can quickly create brand-specific graphics, reducing the time spent on each project.
• Consistency in Outputs: This method ensures that all outputs maintain a high level of consistency, aligning perfectly with brand guidelines.

One of Phillip's best metaphorical takeaways was that design is a team sport. The success of these integrations comes from the collaborative efforts of designers, technologists, and partners like Musho.

By using these tools, creatives and designers can quickly move from a blank page to a polished first draft. To give you a specific example, Superside uses Musho to kickstart their email designs for Vimeo by generating initial templates from a prompt.

These initial designs act as starting points for the team’s copywriters and designers, who can come in to tweak the copy, colors, and images and make adjustments tailored to the specifics of the projects, ensuring the final design aligns perfectly with Vimeo’s brand.

In short, using AI tools like Musho enables Superside to create more meaningful and brand-specific results in a more productive way.

Planning for the future with AI

Phillip concluded his talk with some forward-looking advice for designers. By grounding AI integrations in your own data, you can let technology work for you and not the other way around.

Here's why this matters:

• Maintain Control: Embracing AI technologies helps designers keep control over their creative processes. It's about using these tools to enhance, not replace, your creativity.
• Future Success: The future of design is at the intersection of human creativity and AI. Those who adapt to these changes will be well-positioned to thrive.

In his closing remarks, Phillip pointed out that AI's true potential in design is in its ability to boost human creativity. AI can handle repetitive tasks, analyze lots of data, and provide insights you might miss otherwise. This frees up designers to focus on the creative parts of their work, exploring new ideas and pushing the limits of what's possible.