Moral Lab

Project description

Algorithms are already making many daily decisions. These decisions are automatically made using the data we feed them. What the algorithms decide may not always reflect what their creators or users have in mind. Could machines unintentionally be programmed to carry out ethnic profiling? Could health insurance be more expensive for somebody from Limburg because their medical statistics are influenced by mining activity?
Those dilemmas can be formulated into a series of questions. These questions can then be used for the basis of a research into the target group’s moral profile.
Our design challenge is to find the best way to collect that moral data through an interactive software solution.

Context

There are moral issues all around us. For instance: should people who are already relatively rich be subsidized for driving environmentally friendly? It will help the environment in an effective way, but it will increase inequality at the same time.

Add technology to these already complicated issues and you can also add several other problems to the mix. Tech companies make it easy to follow people from over the whole world, but at the cost of privacy and insecurity from their users. Governments can use more information than ever to serve their people but can also use this information to track or fine them. But if the fine is deserved, isn’t that a good thing then actually?

But do the people who decide over these choices have the moral authority to do so? At the Moral Lab we explore who should have this moral authority, and what can be decided through moral programming. That’s where this application comes in. Through the Moral Lab, data can be collected on a target group’s moral orientation. Future teams will then be able to build on the data we collect.

Results

Here are two major products that we have developed over the course of the semester.

API:
Modular (for the sake of future teams). Handles most of the logic for our web application.
It functions like a base api (you can extend your project api on this base to further specify the functionalities, that makes it so each future team can work from this).

WEBSITE:

MORAL SWIPER UI:
One of the previous Moral Lab teams’ interpretation of the project involved asking long and complex questions in a setting where heavy thinking on each answer was permitted.
Our response to our stakeholders’ request for a tool to collect moral data with was the “Moral Swiper”. By asking many small “Agree”/”Disagree” questions, we build a profile of a target group’s stance on the moral dilemma as a whole. As we do this, we also encourage users to answer as quickly as possible to get their “gut” feeling on each question.
We expect it will be interesting for researchers to compare these “gut” feelings towards a dilemma to more thought-out answers towards the same dilemma.

DILEMMA BUILDER DASHBOARD:
The structure we devised for our backend is reflected most clearly in the Dilemma Builder UI. In this piece of the project, it is possible to build up a dilemma with consideration to the paths between questions. This allows a researcher to build their own dilemma from scratch, without needing technical know-how. This is done via a drag and drop system.

About the project group

Vincent von Schilling:
Hallo, I’m Vincent,
I’m a 3rd semester student at Fontys University of Applied Sciences in Eindhoven studying in the ICT & Software Engineering route.
For the past semester, 4 days a week, Eric and I have been working using SCRUM development tactics to make the Moral Lab a reality for Interaction Design Fontys.

Eric van den Berge:
Hey, I’m Eric,
I’m a student at Fontys, Eindhoven and doing the Software route. That means that I’m working on the architecture & programming of the project.
Together with Vincent, we started work on the project in this semester that lasted for half a year and we spent 4 days a week working on this project to make sure we finished the project on a level we are proud of.
If you want to read more about the project, then keep reading to get more information on the context, the description and the results.