What stories does generative AI tell about people with disabilities?
Do you want to envision and shape GenAI to improve the representation of people with invisible disabilities and chronic conditions?
The field of GenAI moves so fast that our predictions about what the future with AI will look like change every other month. In addition, what we generated yesterday may not be possible to generate tomorrow or vice versa due to regulations around AI. Regardless of the rapid changes and the progress with GenAI tools, one thing seems to persist: misrepresentations of people with disabilities and chronic conditions. This is likely to stay the same unless we actively work on as it is generated among others due to a lack of data, algorithmic or data biases, and a lack of inclusive practices in testing and implementation.
In this half-day workshop...
We will start by presenting three personas of people with invisible conditions (disability or health condition). Each persona will be accompanied by a scenario of how they used GenAI to create a representation of themselves and what impact the generated results had on them. The personas and scenarios are based on a previous workshop conducted with people with disabilities. Then we will identify what the problems are and what actions could solve them. Finally, we would develop a dystopic and a utopic scenario of what would happen if the problems identified are not recognized or addressed, and what would happen if the problems were recognized, and actions are taken towards their resolution.
The aim of the workshop
Generate material for a manuscript that you, as the workshop participant, will have the opportunity to be a part of as an anonymous participant, contributor (in the acknowledgements), or co-author.
Workshop potential outputs
-
A list of actions AI-related professionals should take to improve the cases of the personas that will be presented in the workshop
-
Future scenarios based on the worst and the best cases of future GenAI
-
Material for the next workshop with people with disabilities and chronic conditions related to the project
For participating... (in person)
You will need to have some experience working with AI; it can be technical (e.g., LLMs, Datasets, Model training, etc.) or theoretical (e.g., AI and inclusion, AI ethics, AI policy, etc.), or in-between (e.g., prompt engineering, AI design justice, AI systems interface, etc.). It does not matter if you have this experience because of practice or research.
You will be given this informed consent form in the workshop.
Do you need more information?
Contact
vasiliki.mylonopoulou@gu.se
Workshop starts at 10.00 am
Room: Päätalo/Main Building A062, Kalevantie 4
NOTE: The schedule will be flexible to accommodate the participants and their needs
Time | Activity | Description |
|---|---|---|
10.00 - 10.50 | Introductions | We introduce ourselves and the experiences we had with GenAI. The facilitators introduce the project and the prework that was conducted for the workshop to take place. |
10.50 - 11.05 | Break | |
11.05 - 11.50 | Speculative Design | World Café: Every 15 minutes, we will change the task
Task 1: What is the problem, and how could it be solved?
Task 2: Create a dystopic scenario for the persona you have in front of you
Task 3: Create a utopic scenario for the persona you have in front of you
|
11.50 - 12.00 | Break | |
12.00 - 12.45 | Presentations and Discussion | Present the scenarios and reflect on them through discussion |
12.45 - 13.00 | Next steps and Debriefing | Facilitators summarize the results of the activities, present the next steps and open the floor for questions |
Q&A
What are the methods used in this workshop?
We will implement two methods in this workshop; however, you do not need to be familiar with any of them to participate. We will use a foresight / speculative design framework that you can find here under Figure 2 and below (bold letters), and we will apply the world cafe method, which you can find here. We will post the details of the workshop proposal we submitted as soon as it is available, where you can see the theoretical background in detail. Keep in mind that the proposed framework and methods are tools to help us be productive and constructive in the workshop; therefore, adjustments may happen on the fly. However, you do not need to care about any of these as a participant.
How will my personal data be handled?
We will not collect any personal data in the workshop. However, we may keep your contact information if you inform us that you want to have a follow-up or keep working with us. Download informed consent for this workshop.
Is there an information sheet with all the details I need to know about the research and my data?
Yes, your personal data will not be collected for research purposes; some demographic data may be collected, but not enough for people to identify you. This workshop is part of a bigger research project funded by the Vinnova innovation agency of Sweden. More information can be found in the registration form before you start filling in the related fields.
Can people with disabilities and chronic conditions participate in this workshop?
Of course, if they fulfil the participation criteria. Please contact Vasiliki as soon as possible for potential accommodations if you have not already done so in the registration form.
What is the role of people with disabilities or/and chronic conditions in this project?
Even though the workshop focuses on people with technical skills, the project's starting workshop was with people with disabilities and chronic conditions, and representatives for corresponding disability communities. The aim was to find out how ChatGPT represents them and how it can be prompted to represent them more fairly in the recruitment process. The reason have not make these groups of people have not worked together at the same workshop is due to practical reasons e.g. time and budget limitations.