This project was my final output from the Design/Engineering Masters program at Imperial College London and the Royal College of Art I attended between 2020 and 2022. The project was finished in June of 2022.
An excerpt from the paper:
Humans have always looked for ways to outsource cognitive load to objects and machines in the world, an outsourcing which has formed the basis for much of civilization. Recording taxes and laws on clay tablets allowed increasingly complex societies to keep track of more interlocking parts, while the abacus made it easier for early mathematicians to count and add. Modern humans use address books and spellcheck to outsource social obligations and typing, respectively. The underlying feature of these inventions is that they allow humans to outsource work that does not require innate genius or creativity, freeing up cognitive bandwidth to focus on more meaningful tasks.
The last 20 years have brought us “Tools for Thought”, a genre of software tools that have the goal of helping people share, organize, and manipulate their thoughts and ideas. However, a new paradigm in these kinds of tools is beginning to emerge, powered by recent and powerful advances in the field of AI. AI can now work not just with data, but with ideas as well, and some companies and people have begun to take advantage of this. These advances in AI mean we can build tools for thought that can be powered by pseudo-intelligent systems, allowing a potentially revolutionary rework of the relationship between humans and the computers they use every day.
This project aims to use AI as a tool to help manage and manipulate ideas and concepts, in order to help knowledge workers be more creative and work more efficiently. It is also, almost more importantly, an exploration into how humans can work effectively alongside AI in creative pursuits, and a journey to create a framework and corpus of research to inform the further development of these kinds of tools in an attempt to be a leader in this fast-moving and burgeoning field.
The project produced a fully working demo of a tool for thought called Alexandria, which facilities a new kind of collaborative relationship between human and machine by allowing for a creative dialog between the two agents, a departure from current ideas in the field. The project was successful, but identified innovative ways to facilitate collaboration communication that were not able to be executed in the time frame. These methods of communication include a multidimensional dialog system that facilitates divergent thinking by taking advantage of the unique benefits conferred by working with language-based AI systems. This should be the first idea to be studied in any further exploration of this topic.