Lecture 3: Generative AI & Design Trends
Course code: DES646
Course name: AI/ML for designers
Dr Amar Behera
Email: [email protected]
Extension: 2401
Design Studio 605H-A
Diamond Jubilee Academic Complex
Learning Outcomes
• Understand the concept of Generative AI
• Explore key generative models and their applications
• Examine the intersection of Generative AI with design
• Identify current and emerging design trends influenced by AI
• Discuss ethical considerations in AI-driven design
What is Generative
AI?
• Definition: AI systems capable of
creating new content (text,
images, music, 3D models)
• Based on machine learning
models trained on large datasets
• Examples: GPT, DALL·E,
Midjourney, Runway, Stable
Diffusion
How Generative
AI Works
• Training phase: Learn patterns from
data
• Generation phase: Create new
content based on learned patterns
• Techniques: GANs (Generative
Adversarial Networks), VAEs
(Variational Autoencoders),
Transformers
• Prompt engineering for guided
output
Key Generative AI Models
• GANs: Two networks (Generator & Discriminator)
competing to improve quality
• VAEs: Encode input data into latent space and
decode to generate variations
• Transformers: Use attention mechanisms for
sequence-to-sequence tasks (text, image captions)
• Diffusion Models: Gradually add and remove noise
to generate high-quality outputs
Generative AI in Design
CONCEPT ART AND RAPID GENERATING UI/UX PRODUCT BRANDING AND 3D MODEL
MOOD BOARDS PROTOTYPING OF LAYOUTS VISUALIZATION MARKETING CREATION FOR
VISUAL ASSETS MATERIALS AR/VR
Benefits of generative AI for Designers
Speeds up ideation and iteration
Expands creative possibilities
Enables personalization at scale
Reduces production costs
Image source: https://generativeaimasters.in/advantages-of-generative-ai/
Challenges and Limitations
Quality control Risk of Intellectual
and refinement homogenized property
required design styles concerns
Need for
Potential bias
human
in training data
oversight
Design Trends Influenced by AI
AI-ASSISTED CO- HYPER-PERSONALIZED GENERATIVE AI-POWERED AUTOMATED DESIGN-
CREATION USER EXPERIENCES BRANDING AND IMMERSIVE TO-PRODUCTION
WORKFLOWS IDENTITY SYSTEMS ENVIRONMENTS PIPELINES
Case Studies
Architecture: AI-
Nike: Generative Adobe Firefly: AI-
assisted floor
product powered creative
plans and
customization tools
visualizations
Nike & AI: Reimagining Footwear Design
How generative AI and athlete collaboration are shaping future shoes
‘Athlete Imagined
Revolution (A.I.R.)’
• Nike’s A.I.R. project pairs generative AI
tools with Olympic athletes’ input to
create unprecedented shoe concepts.
• The term “A.I.R.” stands for Athlete
Imagined Revolution—signaling a
paradigm shift in Nike’s design process.
• The initiative was prominently
showcased at Paris events connected
with the 2024 Olympics.
The AI–Athlete–
Designer Workflow
• Generative AI Ideation
• Produces concept visuals by blending
athlete data, styles, textures, and
performance needs.
• Collaborative Feedback Loop
• Athletes review, sketch, and adjust designs—
ensuring functional relevance and personal
expression.
• Rapid Prototyping
• Designs are refined using 3D rendering,
computational modeling, and 3D printing for
near-instant feedback cycles.
Athlete-Inspired AI Concepts
• Eliud Kipchoge: Suggested
modifications to avoid trail debris
catching on the heel—illustrating
athlete-driven iteration.
• Sha’Carri Richardson: Desired a design
that embodies grace and harmony—
resulting in sleek, integrated Air units
extending into the upper.
Performance and Design Synergy
These AI-backed concepts push beyond traditional design boundaries—
heightened aesthetics, aerodynamic forms, artistic expression.
While not market-ready, they represent an “otherworldly potential” that
may influence future performance lines.
This process delivers velocity and fidelity in ideation, with AI serving as a
creative co-conspirator rather than a replacement for human designers.
Broader AI Applications in Nike
• Nike Fit:
• Uses AI-powered computer vision to scan
customer feet and recommend accurate shoe
sizing, enhancing comfort and reducing returns.
• Beyond A.I.R. – AI-driven design elsewhere:
• Air Max Scorpion: Developed during the
pandemic using AI and VR for traction
optimization and rapid iteration.
• General product design:
• AI-enabled modeling and prototyping optimize
materials, geometry, and performance, speeding
time-to-market.
Strategic Insights & Future Outlook
• Key Advantages: Accelerates innovation,
deepens athlete-designer collaboration, surfaces
radical aesthetics, and reduces prototyping time.
• Strategic Edge: Enables Nike to stay cutting-edge
amid competitive pressure, especially leading up
to major events like the Olympics.
Adobe
Firefly
Source: https://www.youtube.com/watch?v=Fxck1CWjue4
What is Adobe
Firefly?
• Adobe Firefly is a suite of
generative AI models integrated
into Creative Cloud, Adobe
Express, Photoshop, Illustrator,
Premiere Pro, and more.
• Text-to-image
• Text effects
• Generative fill
• Vector recolor
• Video and audio editing.
Core Capabilities
• Text-to-Image Generation:
• Turn everyday language into visually
stunning art (e.g., “oil painting of a
forest at dusk”).
• Adjust style, lighting, composition,
and tone.
• Generative Fill:
• Seamlessly add, remove, or expand
image elements in Photoshop or
Firefly with smart text prompts.
• Generative Recolor (Illustrator):
• Quickly explore color themes by
typing simple descriptions,
dramatically speeding up workflows.
Video and audio editing
Text-to-Video & • Generate short, high-quality 5-second clips using text
or a reference image, with control over camera
Image-to-Video movement and style.
Enhance Prompt • This feature improves prompt effectiveness, cutting
down trial-and-error and boosting creation efficiency.
Video Model • Includes generative extend (fill in missing frames), bulk
editing, and third-party AI integrations like Sora, Pika,
Advancements: and Runway.
Ethical and Commercially Safe AI
Firefly is trained exclusively
Includes Content
on Adobe-licensed, public-
Authenticity Initiatives
domain, or creator-
with metadata to prove
consented content—
origin and uphold trust.
ensuring IP-safe output.
• Fully integrated with Creative Cloud tools like Photoshop,
Workflow Illustrator, Premiere Pro, and Express—streamlining
seamless transitions through the design workflow.
Integrations &
Mobile Access • Mobile App Launch
• Adobe’s upcoming Firefly mobile app allows users to
create visuals on the go—syncing work across devices
effortlessly.
Use cases and
industry impact
• Rapid Prototyping & Ideation: Generate mockups
and styles in seconds.
• Enterprise-Ready: Brand control, templates, and
customizable workflows.
• Used by brands like Deloitte, Paramount+, and
Pepsi for marketing.
Ethical
Considerations
• Attribution and copyright
• Transparency in AI-generated
content
• Bias mitigation
• Impact on creative jobs
• Sustainable AI practices
The Future of
Generative AI in
Design
• More intuitive human-AI
collaboration
• Multi-modal generative systems
• Real-time adaptive design tools
• Cross-disciplinary creative
ecosystems
Human-AI
collaboration
Back and forth cycle
The design process: Double Diamond
Problem and Solution Domains
Image by Digi-ark
Traditional
Traditional v
Generative Generative
Shape computation using shape
grammars
• Shape grammars in computation are a specific
class of production systems that generate
geometric shapes.
• Typically, shapes are 2- or 3-dimensional, thus
shape grammars are a way to study 2- and 3-
dimensional languages.
• Shape grammars were first introduced in a
seminal article by George Stiny and James Gips
in 1971.
How shape grammar works
• A shape grammar consists of shape rules and a generation
engine that selects and processes rules.
• A shape rule defines how an existing (part of a) shape can
be transformed.
• A shape rule consists of two parts separated by an arrow
pointing from left to right.
• The part left of the arrow is termed the Left-Hand
Side (LHS).
• It depicts a condition in terms of a shape and a marker.
• The part right of the arrow is termed the Right-Hand
Side (RHS).
• It depicts how the LHS shape should be transformed and
where the marker is positioned.
• The marker helps to locate and orient the new shape.
Let’s figure this out
together
Use the rule shown on the top to generate the
sequence of shapes as shown below on the right
Solution
The same rules can also lead to the below
The question is: How?
AI for Creativity Enhancement
AI can support divergent
Creativity is central to the
and convergent thinking
designer's role.
processes.
Lubart (2005): Technology AI shifts from being a
can manage creative facilitator to an active
work, aid communication, teammate in creative
and facilitate co-creation. exchanges.
Mixed-Initiative
Co-Creativity
(MI-CC)
• Defined as human and
computational initiatives
interacting to create artifacts.
• AI can inspire, suggest,
evaluate, and trigger lateral
thinking.
• AI provides 'random stimuli' to
break fixed patterns in
problem-solving.
• Example: Sentient Sketchbook,
Sketch-RNN.
Sketch-RNN
• Neural network for completing user-
drawn sketches.
• Uses QuickDraw dataset of millions of
human-drawn sketches.
• Collaboration involves iterative cycles
of human input and AI suggestions.
• Both parties influence each other’s
output, improving over time.
• Firetruck drawing suggestion starting
from a user drawn square
AI most used in research, team
building, and task management.
Can process and analyse large
AI in Idea datasets, including qualitative insights.
Development Example: CRIS chatbot for qualitative
Phases and quantitative research.
Limited use in concept phase—area for
further research.
Three roles of AI in
idea development
• AI as Representation Creation:
• Provides inspiration, widens design scope, and
triggers actions.
• AI as Empathy Trigger:
• Supports scenario building, offers new
perspectives.
• AI as Engagement:
• Prevents creative stagnation, encourages
problem reframing and lateral thinking.
https://www.menti.com/alnfj487gi9z
Go to menti.com and use the code: 6990 3011
Discussion
• How can designers maintain
originality in AI-assisted
workflows?
• What ethical safeguards are
essential for AI-generated
designs?
• How might generative AI redefine
the role of a designer in 10 years?
• AI is not replacing creativity—it’s
augmenting it.
• Generative AI is transforming creative
Summary processes
• Offers both opportunities and challenges
• Designers need to adapt and guide AI tools
responsibly
• Nike uses generative tools to amplify human
inspiration.
• The A.I.R. initiative bridges athlete insights
and rapid design cycles to produce visionary
concepts.
Summary • These experiments offer a glimpse of the
future of performance footwear—where
tech, collaboration, and artistry converge.
• Adobe Firefly uses multimodal AI across
image, video, text, and audio—offering a
unified creative platform
