Artificial and Architectural Intelligences in Design
Hypothesis:
If human emotions can be mapped with specificity towards a human-centric design goal, it could become a blueprint for a new era of responsive architecture
Inspired by Greg Lynn's critique of static architecture, the approach diverges towards exploring human emotions as parameters for responsive design. Hypothesizing that mapping human emotions can guide design processes, a proposed framework redefines the relationship between four key players: the architect, the user, the built environment, and AI. Central to this framework is the creation of a knowledge library, leveraging digital documentation of projects for algorithmic analysis.
Framework Overview:
The framework outlines the potential for every project ever designed and documented digitally, to be part of this expandable knowledge library. This library serves to be the foundational database that algorithms could leverage on, to pull data points and references from. The diagram shows the visual representation of the relationship between different contributors. The human user and the built environment interact with each other and the emotional data that comes out of the interaction becomes input data for the knowledge library. The AI algorithms conceived by designers can then make sense of the data and organise them into information that is usable by human designers. The human designers can then refine the algorithm to sieve out more specific and relevant information and eventually make decisions for the AI to proceed with design output and subsequent iterations.
Application Example:
Old Folks Home: Illustrating the framework through a hypothetical old folks home design, emotional data collected through surveillance and biosensors inform design decisions to enhance user experience. Algorithms analyze patterns, facilitating a curated interior environment conducive to sustained positive emotions.
Graph on Emotions:
One of many ways of organising emotional data is through a matrix. Determining threshold and normalising the dataset
Here is a chart showing one of the many ways of charting and organising emotional data by the nature of the emotions, whether they are on the calm side or exciting side, closer to the unpleasant or pleasant side in comparison to other monitored emotions. A threshold can be determined for each emotion and then the data can be normalised for comparison and ideal state setting.
Graph on Elderly in Old Folks Home:
Circles represent the dominating emotion. The size of the circles represent the quantity of similar responses gathered
This is a visualisation of the hypothetical old folks home data being monitored. Where each circle represent the dominating emotion experienced by all tracked users are experiencing at the particular timeframe. The larger the circle means the higher the number of similar responses gathered. At points where the emotion circles are far away from the acceptable threshold for a relatively long period of time, the responsive design interventions could come to change the situation.
Application Example 2 - Prison Cell:
In contrast to the old folks home, a hypothetical AI-managed prison cell demonstrates a different design goal. Rather than just maintaining a balanced emotion status, the AI also makes decisions to implement environmental changes when the emotional threshold is surpassed, acting as a responsive system of control.
Graph on a Prisoner in Prison Cell:
Large changes in the emotional status of the prisoner will mean the implementation of intrusive control measures
In the hypothetical scenario of the the prison cell, the AI looks for large changes in emotional status of the prisoner which is indicative of the need for intrusive control measure. In comparison to the example of the old folks home, the prison cell data has very different threshold settings that need to be carefully considered by the designer and the algorithm.
But when the system goes rogue...
The video below is a depiction of the potential drawbacks of this framework that have resulted in an "unthoughtful" application of the system. The context of the project is about a future prison fully managed and controlled by an AI algorithm, designed to contain highly dangerous prisoners that need to be kept away from the human population.