Affective Computing - when technology meets emotion

Written at Jul 28, 2023 9:31:08 AM by Ksenia Gorska

Last week my female distriBind colleagues and I attended the Women in Data Science conference organised in London.

During the course of the day we had a chance to listen to 4 amazing speakers, who brought to our attention a few very important and interesting aspects of data science and the data-driven decision making process. While all of the lectures were really engaging, the one given on by Akshi Kumar affective computing really resonated with me and gave me lots of food for thought.
blogs (1)

Imagine that you come back home after a tough day. You are tired and agitated. Who’d be able to better read your mood and adjust to the current situation? A computer or a puppy…?

A puppy is capable of recognising cues in a human’s body language and assessing whether its positive or negative. It can notice changes in the tone of our voice and react accordingly. It will either come to greet us with joy or prefer to stay away and leave us alone for the time being. That’s an example of its own emotional intelligence that it developed by observation, therefore here puppy’s emotional quotient is higher than the computer’s.

Akshi took us on a tour of her world and made us aware of the current social and ethical issues in affective computing; that is when the technology meets emotion. Nowadays, more and more systems and technological devices have the ability to read, decipher, process and simulate human affects (BBVA Open Mind 2016). It does however pose a question of whether providing emotional capabilities to artificial intelligence is something we should be striving for?

In the world where AI would be able to detect one’s emotional state and try to manipulate one’s behaviour to achieve certain responses; we risk stripping ourselves out of the mystery associated with one’s uniqueness and risk falling victim to categorising human beings based on their personal features. At present affective computing raises issues around user’s privacy, autonomy and informed consent (MIT Edu 2021). One might start to wonder whether there are things that simply shouldn’t be automated?

Personally I think that while the virtue of emotional intelligence shouldn’t be all left out to software, back-end data processing for the delegated authority definitely can. distriBind won’t try detect how you are feeling today, but it will handle the data processing across your written, premium and claims bordereaux and give you some of your valuable time back. Maybe that will make you feel better, even if we can’t compete with the puppy…

Let us know what you would like to do with all the additional time on your hands!

Share article

Ksenia Gorska

Senior Business Data Architect