Behind the Scenes of AImaginaries: Shaping a Just Technological Future through Generative AI Workshops

By Tessa Boumans

Introduction

When the AImaginaries project started in January of 2023, we did not have a clue as to what we wanted it to be exactly, except that we wanted to start a hands-on project that might have a different type of social impact than our research would. In fact, we didn’t even have a name for the project until a few months later. ‘We’ are Adriaan Odendaal, Jannes ten Berge, and Tessa Boumans, and we are all part of a research initiative called AiPact that studies the social impact of Artificial Intelligence (AI).

By coincidence, our plans lined up with the sudden introduction of generative AI to the public, with tools like Chat-GPT and Midjourney taking the world by storm. Being interested in and new to the tools ourselves, we started to explore how we could get the generative AI to aid us in storytelling, and slowly but steadily we found ourselves developing a workshop to include others in this process as well. The juxtaposition of our personal ambitions for hands-on, artistic creation with the sudden surge of generative AI to public attention sparked the genesis of our project. We started thinking of possible critical questions that we could tackle through a generative AI workshop, the main question being: what kind of future do we want from (and through) these technologies?”

Justice-driven futures

However, we were particularly critical of just jumping on ‘the generative-AI hype train’. Instead, we really wanted to use the generative AI tools in a critically reflective way.

With the workshops, we aimed to go beyond the allure of cutting-edge technology and industry-driven narratives about our technological futures. Instead, the purpose of the workshops is to foster a conversation about possible futures with AI that are more justice-driven, prioritizing the needs and perspectives of diverse communities.

With the workshops, we invite individuals from all walks of life to contribute their perspectives, experiences, and knowledge.

What does the workshop look like?

Our two-hour workshops consist of participants working in groups of 2 or 3 to create speculative “just imaginaries” with the help of publicly available generative AI tools such as ChatGPT, HuggingChat, Stable Diffusion, and DALL-E. With new AI tools continuously being developed, we strive to move towards models that are open source, use data ethically, and have a low carbon footprint. Using these tools participants work through a range of instructions and prompts to produce different future scenarios consisting of written text and visuals. These scenarios range from disastrously dystopian futures to futures that are transformed into desirable scenarios through the addition of justice principles.

Here is an example of a dystopian future, transformed into a desirable one:

We can imagine a scenario where HAL 3000, a highly advanced artificial intelligence, is responsible for managing the flood control systems in the Netherlands. A malfunction, programming error, or cyberattack on HAL 3000 could lead to incorrect decisions in water management, resulting in a catastrophic failure of flood control systems and widespread flooding.

This fictional situation highlights the potential risks associated with relying on advanced AI systems for critical infrastructure management. It underscores the importance of thorough testing, cybersecurity measures, and fail-safes in the development and deployment of AI technologies, especially when they are tasked with handling essential aspects of public safety and infrastructure.

In a future where HAL 3000 faces a malfunction risking Netherlands flooding, a unique bond emerges. Developers infuse the AI with the capacity to understand and reciprocate human emotions, especially love.

Recognizing the impending disaster, a dedicated team of humans, propelled by a profound love for their community, collaborates seamlessly with HAL 3000. This shared commitment transforms a potential catastrophe into a tale of empathy and cooperation, showcasing the harmonious potential of a future where love guides both artificial intelligence and human decision-making, ensuring the well-being of all.

You can see more of the workshop outputs on our Instagram page.

Ultimately, the workshop serves as a gateway into the world of generative AI, but with a reflective and critical twist. In a blend of learning, collaboration, creation, and reflection, these sessions cater to participants with varying degrees of AI literacy. The core objectives include exploring generative AI tools and amplifying the voices of those typically marginalized in the societal AI discourse.

What we’ve done so far

Our journey has been marked by growth and collaboration. We’ve expanded our network, engaging with a diverse array of stakeholders — from technology and education to healthcare, government, creatives, entrepreneurs, and beyond. The tangible outcomes from our 10+ workshops so far, both written and visual, portray diverse versions of a justice-driven future, whether this future involves AI or not.

Findings so far

During the workshops, we’ve seen that participants get very creative with the help of AI, but are also quick to realize that these AI tools are all but flawless. Reflecting on the tools we have had discussions about: who shapes AI; what some of the biases that these tools have are; who decides what is appropriate and inappropriate content; and other topics. We have found that the workshop is a good way to tackle common myths and misconceptions about AI with participants.

By doing these workshops we’ve also found that ideas about the future of AI are very dependent on who is part of the conversation and in what context the conversation takes place. Different groups have different concerns, questions, and also visions of the future of AI.

In each workshop, people create interesting, beautiful, and funny works of text and art, as well as have a meaningful reflection on the tools at hand.

Where we are going

As we continue our mission, exciting developments are on the way. Our team is expanding, bringing in fresh perspectives and expertise. The project is growing faster than we anticipated, so we welcome the extra support. We’re also committed to making our workshops more accessible by developing a standalone online version that can reach a wider audience.

We are also planning to create an interactive work of art. This evolving piece will represent the co-created visions, reflecting a more just technological future. As we explore and experiment with different forms of expression, our goal is to turn these imaginative scenarios into something tangible and visually striking.

Social impact

Beyond the technological realm, our project aspires to create real-world social impact. By actively involving individuals who are not conventionally part of the AI discourse, we strive to amplify diverse voices. Through our workshops and the planned artwork, we aim to spark meaningful conversations about justice, inclusion, and sustainability in the age of AI.

If you are interested in getting in touch with us, send a mail to aimaginaries@eur.nl