SlowAI
Project Type
Strategic Intervention, Ethics Mapping and UX Design
Date
2024
Location
As Artificial Intelligence develops at an unruly rate and newer tools and technologies start arising from it we consider the impact it has on students in higher education, they have access to a broad range of technology but operate within the confines of higher education which keeps them accountable through a rigid set of rules of academic integrity. AI here challenges their educational pedagogy, Human values, transparency around AI among other ethical issues. Students, often under the pressures of time and maybe even wild curiosity, use these new tools without a comprehensive understanding of the effects it may have on their learning. While being open to newer technologies is important the Defects in education around AI cause unintended consequences. SlowAI is an academic citation and education tool that uses a top-down and lateral approach to addressing this issue. We realised that the answer to technologies we don’t fully understand isn’t stopping its use altogether.
So, SlowAI involves students, tutors and the higher education institutions in thinking more deeply about regulations and values around AI image generation without ignoring AI as an important current and future tool.
The Behavioural Solution
We propose an enhanced AI generation tool that encourages Mindful use of AI for image generation, by promoting users to be more thoughtful, responsible and creative in the journey of Image generation.
Rather than directly jumping on drafting a prompt, we propose an engagement that first starts with users reflecting on their purpose of using AI. Following this, the users are nudged to create a rough sketch of their expected outcome. This needs to be uploaded and is not for the system to derive details for generating images, but rather for the users to have visual context and clarity before they proceed to draft a prompt. The sketch thus acts as a reflective medium. Upon completion of this step, the users are invited to start defining their prompt, the prompt interface with its divisions adds another layer of enhanced experience, by reducing the cognitive load to writing sentences. The image generation process remains the same as for other prominent Generative AI Tools and the output is a set of 4 images, created based on the given output. The users are then asked to access the usability and ethical cheks of these images and offered to share and/or download them, with a subtle reminder of citing appropriately. If the user is unhappy with the outcome, they can re-generate the image, change their prompt or restart the entire journey. Additionally, there is a dedicated library, which houses all the previous. interactions with the tool, thereby acting as an Online Portfolio showing the AI Using practices of the user. This Library webpage can be password protected and shared with recruiters, acads, experts.