Ethical applicationen of AI

How to make ethical algorithms

Thesis project description

A while back an algorithm was implemented in jobcentres that predicted who were at risk for becoming long-term unemployed. When implemented it predicted that people above the age of 60 and those whose parents came from other countries were more likely to become unemployed. The algorithm did not solve a problem with unemployment, but instead just predicted the biases that employers have when they are looking to hire new employees.

Instead an algorithm that focused on competences might have been more ethical to implement.

I'm very interested in working with ethical applications of algorithms that have direct influence on people wether it is in the public or private sector is not important.

Hopefully the project will evaluate a problem or problems, and come up with guidelines or solutions for a way to ethically implement algorithms

Managerial value of the research

Information and guidelines for making better and more ethical algorithms and applications of AI.

Research topic motivation

People sometimes assume that technology is unbiased and algorithms just give you an idea of what the "truth" is. However this is far from true. Every time an algorithm is being created, decisions are being made about the purpose of it, and what data to apply, these decisions are made by people, and people are inherently biased. My hope is to make companies more aware of biases, so they can make better solutions/algorithms

Why collaborate with us

This is something that I'm extremely passionate about. Throughout my studies I have had a big focus on ethics, data, and diversity, and I would love to apply it in a business setting. 

I have specialized myself within the field of big data, and I have found the topics ethical applications, biases, and representation of people especially interesting.