WE SPECIALISE IN FINDING FANTASTIC OPPORTUNITIES
FOR DIGITAL AND DATA SPECIALISTS WITH THE MOST INNOVATIVE BUSINESS ACROSS EUROPE AND THE USA.
The role of ethics in modern data science
Artificial intelligence is becoming a common presence in our day-to-day lives.
It influences our choices as consumers, it determines what is put before us in social media, and it is even having an impact on legal systems.
But who judges whether or not these data-driven decisions are actually...fair?
Algorithms are making important decisions
Professors in some higher education institutions are beginning to bring years of research into bias and discrimination inherent in machine learning to their teaching.
Ethics and policy have a role in the future of data science, as there are issues that arise from the fact that algorithms are now making important decisions that impact people's lives dramatically.
There's excitement, but also concern
The question is whether the fact that these decisions that are driven by data necessarily means that they can't be discriminatory.
There is definite excitement in the prospect of pushing rigorous and empirical data analysis to the heart of high-stakes decision making. But research has shown that data can function as a mechanism by which certain biases are relearned by the algorithms that use it.
And as algorithm use continues to rise, so do many concerns regarding privacy.
Machine learning can be harnessed to make inferences about deeply personal information based on evidence drawn from unlikely sources.
An algorithm may be able to determine sensitive health conditions in a person based on seemingly benign social media activity.
Yes, you read that right.
The world needs to be prepared for the possibilities that these computational techniques give rise to. And, even more worryingly, there is the possibility that unscrupulous individuals or corporations might use algorithmic techniques to collect sensitive data from people in a deliberate breach of ethical practice.
There is danger of bias being encoded
There is a growing number of people involved in the data science world who are pushing for greater fairness, transparency and accountability in algorithmic systems.
Research into the dangers of unintentionally encoding some kind of bias into automated decision-making devices is on the rise and is becoming increasingly important as such devices become more commonplace in everyday existence.
Ethics in machine learning is a rich area to explore, and the research is sure to help ensure data science continues to benefit humanity in the way we all want it to.
Interested in learning more?
We recently held an AI in Data Science Meetup in Berlin, where Luba Weissmann discussed the potential of behavioural data in credit scoring. Very thought-provoking.