CHI Paper MOC
# Overview
- Basically like my thesis (See Atlas/Projects/Thesis) but with a different methodology
- Thesis: researches creates AI art
- CHI Paper: participants create AI art
- Goal: Full paper submission to CHI 25? and/or WIP paper submission to CHI 24
# Inspiration
- One Week in the Future: Previs Design Futuring for HCI Research
- Human Enough: A Space for Reconstructions of AI visions in Speculative Climate Futures
- Designing for Narrative Influence:: Speculative Storytelling for Social Good in Times of Public Health and Climate Crises
- Poetics of Future Work: Blending Speculative Design with Artistic Methodology
- Articulating (Uncertain) AI Futures of Artistic Practice: A Speculative Design and Manifesto Sprint Approach
# Ideas
- Workshop where participants create AI art based on the following prompt: “Imagine a day in the life in 2040 for the Filipino”?
- base idea…you’re essentially exploring the use of AI tools to envision the future…how does the AI influence their perceptions?
- Must educate participants beforehand (e.g. how to write good prompts)
- maybe involve not just have them do AI art, but AI writing as well
# Questions
- Logistics of running a workshop or a sprint, especially in a remote setting
- tackling AI in an ethical way?
# Notes: 7/17/23
- better if participants write the prompts kasi
- kailangan mo igeneralize…don’t evaluate usability of chatgpt….evaluating usecases of AI tools for this context (envisioning the futures)
- Pros and Cons
- sprint: prepared paper methodology, defined na
- workshop: faster
- diary study:
- ^ kailangan ng justification, prepared
- ^ write research protocol, expected outputs, present possibilities to Jordan
- define participants and their background…should they have experience?
- get at least 5 participants…masyadong marami na ang 10
- Ethical Limitations
- participant recruitment: consent, recruitment
- no problem with AI
- ^ you can make this part of the workshop, their perception on AI…dapat related siya sa main topic
- based on the results of the workshop…this is what we think…and this is how it relates to greater literature
- identify positioning of paper and how it contributes to the greater literature
- HCI contributions: empirical, artifact
- expected outputs:
- prompts won’t be very valuable
- “how do they envision the future”?
- now that we have AI genertation tools, what future do they see based on the tools?
- another consideration: technical set-up
- set up discord server for the participants, connected to Richard’s computer, Midjourney
- just use one discord account, and send multiple prompts (for video)
- issue: Stable Diffusion is pretty resource intensive
- Dall-E: not good quality
- set up discord server for the participants, connected to Richard’s computer, Midjourney
- “what if we use generative AI tools to enhance the D&D storytelling experience?”
- take exact same 1 week into the future workshop….
- don’t draw…generate with AI tools
- if you follow structure, you can ask them to generate multiple images
- do workshop parts in FigJam open session or Miro
- Timelines
- If you want to aim for CHI 24: late breaking work (like preprints or posters) January 25, 2024
- pwede siyang remote
- considered semi-archival, which means you can reuse your paper in the next CHI (if you’re going to make it a full paper)
- you need to have not just research done, but also have the paper written
- paper length: 8 pages long, excluding references, double column format
- You should be done with the workshop by October
- you can do this within the next 2 months, just figure out logistics
- Recommendation: Do it ASAP! by middle of first semester
- email Maam Nikki so that you can get clearance pa
- CHI paper is small part of thesis (the research part), you have to ask for permission on this
- the rest of the process is the thesis na your creative work
- If you want to aim for CHI 24: late breaking work (like preprints or posters) January 25, 2024
- make research participants background more specific
- “They self-identify…”
- how many years of experience?
- for students, they should probably come from specific majors
- pre-filter survey
# Action Items
Figure out research protocol, explore other concepts (e.g. sprint), then send feedback to Jordan
- ask about sensemaking
- talk 1-2 sentences about the intersection’s results
- you can start interviewing experts
Once it’s clear, execute, then observe
- You need at least 2-3 hours for the workshop
- think about compensation
- or explicitly state that it will be voluntary
Write your research questions
- this will be part of your contributions
Write email to Ma’am Nikki talking about collaboration and asking for meeting / Zoom call
- Ask Jordan + Richard to check before sending
Write about desired research protocol, put in presentation and ask for feedback from Jordan
Write out your research questions
# Research Protocol
# Project Overview
Title of Project: Keywords: Generative AI, Design Futuring, Speculative Design,
# Background and Current Approaches
# Rationale and Objectives
- Rationale
- Provides the justification or reasoning behind the research project. Explains why the study is necessary and why the research question or problem is worth investigating.
- Discuss the background context of the research, including any gaps in current knowledge or literature, potential practical implications, and the significance of the research problem.
- Objectives
- Outline the specific goals or outcomes the researcher aims to achieve through the research project. They define what the study intends to accomplish.
- Clear, concise statements that describe the expected results or milestones of the research. They may include specific research questions, hypotheses, or tasks.
# Research Questions
- In what ways does AI support the generation of ideas for an utopian future for youth in developing countries?
- What are utopian and dystopian futures typically imagined by youth in developing countries, and what role does technology play in these?
- How might design fictions as a method of futuring elicit new ideas that speak to concepts of a utopian reality for youth in developing countries?
Notes:
- maybe replace AI with AI art generators? unless ChatGPT will also be used
# Hypothesis Statements
- AI Art generators make the process of design futuring easier and more efficient
- Imagined futures are deemed “utopian” or “dystopian” depending on the accessibility and impact of technology in these scenarios
- Design methods as a method of futuring can foster critical thinking and imagination
# Conditions and Variables
# Possible Conditions
- No AI tool v.s. with AI tool?
- best tested out in a group setting, like a workshop. creating design fictions w/ ai v.s. w/o
- w/ design fiction v.s. w/o design fiction
- pre-study and post-study, measure level of critical thinking and imagination? but how
# Possible Variables
- RQ 1
- Independent variable: Use of AI-driven idea generation tools
- Dependent variable: the amount of creative support a user feels during the design futuring process
- RQ 2
- Independent variable: Imagined/elicited futures from participants
- Dependent variable: Role of technology in the scenarios? Or perceptions regarding role of technology?
- RQ 3
- Independent variable: Use of design fictions as a futuring method
- Dependent variable: Critical thinking and imagination?
- Others
- Participant demographics: Collect demographic information about your study participants, such as age, gender, educational background, and experience with technology.
- Cultural context: Consider the cultural context of participants and how it may influence their responses and perceptions.
# Technical Script
to see how were gonna do the studies/workshops.
# Targeted Contributions
- Empirical
- The utility of AI art generators in the design futuring process
- Findings on how much design fictions can foster critical thinking and imagination?
- Elicited futures from youth in developing countries
- Artifact
- Workbook (a novel system)
- Website gallery of participants’ design fiction outputs (if this counts as novel designs)
- Methodological
- reevaluating design sprint method of Ivanov et. al (One Week into the Future)
- introducing a “reflective” component based on method from Harrington & Dilahunt (Eliciting Tech Futures from…)
# Evaluations
# Study 1: Pilot Study Protocol
# Study Design
- Only one participant will be recruited for this pilot study.
- The goal of this pilot study
- Evaluating the impact of introducing a “reflective” component into Ivanov et. al’s design sprint method
- Evaluating the results of using AI art generators for design futuring
- The participant will undergo a 5 day sprint wherein they will do the following:
- Get introduced to design futuring + Stable Diffusion
- Envision utopian and dystopian futures through co-designed? speculative design fictions
- Produce concept sketches, storyboards, and comics that examine the implications of these futures
# Task
# Measure (Analysis Criteria)
- Qualitative instruments:
- 1:1 Interview
- Content + thematic analysis
- Quantitative instruments:
- Creativity Support Index
- Mixed-Initiative Creativity Support Index: MICSI rates co-creative systems along experiential axes relevant to AI co-creation. We administer MICSI and a short qualitative interview to users who engaged with the Reframer variants on two distinct creative tasks.
- Can be collected through VideoAsk platform
- Creativity Support Index
# Results
# Study 2: Testing with Focus Group
- Assessing effects and potential benefits will be done in a further and larger user study.
# Study 3 (Complementary): Workshop Setting
# To-Do
Design workbook
Create educational content for VideoAsk
To read (for background and current approaches):
- Human-AI Co-creation: Evaluating the Impact of Large-Scale Text-to-Image Generative Models on the Creative Process
- A Poetics of Potential Form: The Role of Architectural Creative Writing in Augmented Visual Imagination (AVI)
- Trash to Treasure: Using text-to-image models to inform the design of physical artefacts
# Resources
- https://www.figma.com/community/file/1074773789352934150/Comics-in-Figma
- https://www.figma.com/community/file/1228326072903106794/AI-Comic-A-Day
- To request:
- https://dl.acm.org/doi/abs/10.1145/3591196.3596616 (background and current approaches)
- https://dl.acm.org/doi/abs/10.1145/3581641.3584095 (instrument)
- https://dl.acm.org/doi/10.1145/3596925 (background and current approaches)
- https://dl.acm.org/doi/abs/10.1145/3591196.3596819