Study
ModellingHigh ImpactModerate effect

AI sketch similarity influences creative design shifts

The degree of visual and semantic similarity between a user's sketch and an AI-generated sketch significantly impacts the type of creative thinking employed by the user.

Academic Publication · 2019

01

Key Findings

  • 01High similarity between user and AI sketches is associated with combinatorial creativity.
  • 02Medium similarity is linked to exploratory creativity.
  • 03Low similarity is related to transformational creativity.
02

Application

Design takeaway

Designers of AI-assisted creative tools should consider tuning the similarity of AI-generated suggestions to intentionally steer users towards combinatorial, exploratory, or transformational creative thinking.

How to apply

When developing AI tools for brainstorming or ideation, implement algorithms that can adjust the novelty or relatedness of AI-generated suggestions based on the desired creative outcome.

Project actions

  • 01When using AI for ideation, consider how the AI's suggestions might influence your thinking.
  • 02Experiment with AI tools that allow you to control the 'creativity' or 'novelty' of their output.
03

Method & Evidence

AimTo investigate how varying levels of similarity between user sketches and AI-generated sketches affect different types of design creativity (combinatorial, exploratory, transformational).
MethodEmpirical study involving a co-creative sketching tool and computational models of conceptual shift.
ProcedureParticipants sketched in response to AI-generated sketches. The similarity between the user's and AI's sketches was manipulated (high, medium, low visual and semantic similarity). User responses were analyzed to identify the type of creativity exhibited.
ContextHuman-computer interaction in design sketching, computational creativity.

Variables

IVSimilarity level between user sketch and AI sketch (high, medium, low).
DVType of design creativity exhibited by the user (combinatorial, exploratory, transformational).
CVSketching interface, task instructions, time constraints, user's design domain.
04

Strengths & Limitations

Strengths

  • +Investigates a novel aspect of human-AI co-creation.
  • +Provides empirical evidence for the link between AI suggestion similarity and creative modes.

Limitations

The specific AI model and dataset used might not be generalizable. Measuring creativity can be challenging and subjective.

Reliability & validity

Reliability could be improved by using multiple coders for creativity types. Validity is supported by the clear link between similarity levels and distinct creative outcomes, though the subjective nature of creativity measurement is a challenge.

Think critically

How might the cultural background or prior experience of a designer influence their interpretation of AI-generated sketches and their subsequent creative response?

05

Design Principles

"AI-generated design prompts should be calibrated for similarity to strategically influence the user's creative process."

Understanding how AI partners influence human creativity is crucial for developing effective co-design tools. This insight can guide the design of AI systems that strategically offer suggestions to foster specific types of creative exploration, leading to more innovative design outcomes.

06

What This Means for Your Design

An AI drawing tool can help you be more creative by showing you different kinds of drawings. If the AI drawing is very similar to yours, it helps you mix your ideas. If it's a bit different, it helps you explore new things. If it's very different, it helps you completely change your idea.

How to use in your project

  • 1.Discuss how AI-generated stimuli in your design process influenced your ideation phases, referencing the relationship between similarity and creative modes.
07

Add to My Project

08

Quick Cite

(2019). Relating Cognitive Models of Design Creativity to the Similarity of Sketches Generated by an AI Partner. Academic Publication. https://doi.org/10.1145/3325480.3325488 Retrieved from https://designdex.org/study/dc544a35-94a1-4a89-88b5-603ad9017a31/ai-sketch-similarity-influences-creative-design-shifts

Paragraph starter

The use of AI in the design process can significantly influence creative output. Research indicates that the degree of similarity between a user's initial design and an AI-generated suggestion can steer the user towards specific types of creativity, such as combinatorial, exploratory, or transformational, depending on whether the AI's contribution is highly similar, moderately similar, or distinctly different.

09

Source

Academic Publication

Relating Cognitive Models of Design Creativity to the Similarity of Sketches Generated by an AI Partner

journal · 2019

View source

Questions about this research

What does the research say about ai sketch similarity influences creative design shifts?
Designers of AI-assisted creative tools should consider tuning the similarity of AI-generated suggestions to intentionally steer users towards combinatorial, exploratory, or transformational creative thinking. Evidence: Academic Publication (2019).
Why does "AI sketch similarity influences creative design shifts" matter for design?
Understanding how AI partners influence human creativity is crucial for developing effective co-design tools. This insight can guide the design of AI systems that strategically offer suggestions to foster specific types of creative exploration, leading to more innovative design outcomes.
How can designers apply this research?
Designers of AI-assisted creative tools should consider tuning the similarity of AI-generated suggestions to intentionally steer users towards combinatorial, exploratory, or transformational creative thinking.
What were the main findings?
High similarity between user and AI sketches is associated with combinatorial creativity.. Medium similarity is linked to exploratory creativity.. Low similarity is related to transformational creativity.
What research method was used?
Empirical study involving a co-creative sketching tool and computational models of conceptual shift..
How strong is the evidence?
Evidence strength is rated Moderate effect, based on a 2019 journal from Academic Publication.
What should I do differently in my next project?
When developing AI tools for brainstorming or ideation, implement algorithms that can adjust the novelty or relatedness of AI-generated suggestions based on the desired creative outcome.
What are the limitations?
The study's findings may be specific to the particular AI model and sketch database used. The definition and measurement of 'creativity types' can be subjective.
Is there evidence that creative affects design outcomes?
When an AI's sketch is very similar to a user's, it tends to inspire combining existing ideas. Moderately similar AI sketches encourage exploring new directions, while very dissimilar AI sketches prompt more radical transformations of the initial concept. Understanding how AI partners influence human creativity is cruc Source: Academic Publication (2019).
Where does this design research apply?
Human-computer interaction in design sketching, computational creativity. It sits within modelling research on designdex.org.

Related research topics

creative design research · evidence on creative · does creative improve design outcomes · design studies for designers · creative and design findings · modelling research evidence