the act of design is often referred to as problem-solving. however, this view often restricts the scope of design to addressing only known-knowns or known-unknowns while excluding the potential to identify underlying systemic issues (hill, 2012).

building on this, it is worth revisiting herbert simons definition of design, transitioning from an existing situation to a preferred.

to further explore the mechanics of problem solving and design, we can refer to a common foundation of reasoning patterns in problem-solving, which consist of three components: "what" (elements), "how" (the pattern of relationships), and the "result" (observed phenomenon). these concepts are integral to various reasoning processes, including deduction, induction, and abduction (dorst, 2015).

due to the high uncertainty of abductive methods, designers must analyze the systems they want to improve. because of our bounded rationality, only parts of the system, so-called mental models, can be recognized. (adapted graphic from iwabuchi, 2021)

as an integral part of the design process, assumptions and biases can influence the course and outcome of the design. liedtka (2014b) identified nine biases that can harm decision-making within the design process.

design processes typically start with certain assumptions about the environment and problem space, which constrain our judgment and the extent to which we can explore new ideas, as they tie our thinking process to established beliefs. these assumptions limit our ability to imagine what is possible or preferable, instead perpetuating the status quo.

if we want to implement more profound societal changes, we need to question our thought structures in order to tackle problems like climate change or or social injustice. however, when large language models are used for thinking about future scenarios, they will project the past into the future through their probabilistic thinking.

my aim was to consider the designers' ideas or assumptions in collaboration with large language models. the intention is to identify gaps in the design argument that may arise due to the mental models, assumptions, and biases of the designers.

during the thesis, i developed sacrificial concepts to make the potential of ai to support the design process tangible. these concepts were designed to exaggerate different trajectories to provoke discussion about adverse consequences and the desirability of these tools.

'the map,' symbolizes a comprehensive guide to help designers navigate the often ambiguous landscape of a project.

the magnifying glass metaphorically 'glows' on unexamined beliefs and makes them explicit.

'the search dog,' embodies an innate sense of direction and can see what's important by highlighting the most important assumptions.

through iterative testing, i created a prototype to challenge ideas and make implicit assumptions visible. the prototype can be found and tested here:
prototype.anticipate.studio

an explanation of the rationale from a design perspective is available here:
anticipate.studio