It’s no secret that our evolutionary response to complicated or new technologies is to either ignore it, or fear it. However, at OnTrac, our response is different: we embrace new tech, and actively explore new concepts and ideas surrounding existing and emerging technologies. Technology is increasingly imperative in our everyday lives, and with the sheer velocity in which the field is expanding, we’re constantly being encouraged to incorporate new ways to incorporate technology – once considered science fiction – into our everyday realities.
In the 2013 movie, Her, the topic of algorithmic personality detection was actively explored via the protagonist’s use of his new operating system. While this was undoubtedly paired with the Internet of Things and artificial intelligence, the operating system was eventually able to develop into an intuitive and unique entity.
Although the movie was relatively fantastical, algorithmic personality detection is incredibly real – and attainable. While we may not be able to develop relationships or truly communicate conversationally with algorithms (yet), we can explore the fundamentals of algorithmic personality detection, and apply it to a number of tangible business practices – the start of which is already being seen.
At its core, an algorithm is a process that needs to be followed, to solve a problem. Algorithms are vital to modern business practices for many reasons, and this year has seen a monumental shift in algorithmic applications, with a focus on specific individuals, becoming more apparent. This is where personality detection becomes incredibly interesting.
Algorithmic personality detection, quite basically, is an assessment of an individual’s personality, via algorithmic predictions.
As is already beginning to be explored, an algorithm can predict the type of person you are, by simply assessing the websites you visit, or the content that you share on social media. Basically, the algorithm monitors and analyses your digital finger print, in order to assess your personality – and this becomes vital when considering its potential in business applications.
Personality detection algorithms borrow concepts from fundamentally human practices, such as psychology and sociology, and uses these methods to determine patterns in huge amounts of data – i.e. big data in practice. So, it makes sense that social media has become a formative hub, so to speak, for an algorithm to gather information about an individual.
Social media users now number over 1.4 billion people, so it does seem somewhat natural for personality detection algorithms to focus on the information we share via social media. It is an enormously extensive platform, and while it may seem that the information being liked and shared is completely innocuous, when analysed against the sheer volume of other profiles, an individual can share truly vital information – information that could revolutionise the way we do business.
In a paper published by the Proceedings of the National Academy of Sciences, research was conducted, in order to determine whether or not algorithms could learn more about us than our fellow humans could and the research proved that it is possible, in far more realistic terms that algorithms can learn more about us than simply asking another person to assess our personalities.
The research tested subjects’ Facebook ‘likes’ and overall activity, focusing on a five-trait personality profile. This included:
The algorithm required 100 of an individual’s ‘likes’ to outperform the average human, and successfully predict the subject’s five-trait personality score more successfully than colleagues or friends.
Another study, published by researchers at the University of Cambridge, gathered data from Facebook users, and by using their ‘likes’ alone, they were able to predict a wide range of personality traits, such as intelligence, gender, religion and sexual orientation. While this information seems rather easy to attain, it does become more interesting when you discover that researchers’ predictions weren’t derived from the obvious sources. Instead, it came from connections that were entirely unrelated to the personality traits. For example, a top Facebook ‘like’ strongly indicative of high intelligence, was ‘thunderstorms’.
This type of algorithmic personality detection is linked to the social scientific concept of homophily. This concept aims to inform us that people like to be friends with people who are most akin to themselves – ergo, smart people like to be around fellow intelligent people, etc. Therefore, pages like ‘thunderstorms’ are algorithmically associated with more intelligent Facebook users, because this is spread through elements of the social network that happens to consist of, categorically, more intelligent people. Therefore, the algorithm predicts that you are an intelligent person.
By allowing algorithmic personality detection to assess my own Facebook profile, I was able to discover the extent of my digital footprint, and how I am perceived online. I was pleasantly surprised – which is a relief! Of course, it’s not just social media that allows for this type of analysis; web browser history, credit card usage and even online linguistics can be tracked and analysed to ascertain more accurate information about another person – and the beauty of this? There’s no way for an individual to control or falsify the results of algorithmic personality detection, making its application hugely imperative and advantageous in business practice.
Utilising this software in business is becoming popular; the market is already taking leaps in this direction because of how it can be utilised; aiding in the growth of an organisation, by understanding its customers better.
Insurance brokers are already using versions of this algorithm to accurately predict a client’s liability, and an organisation will also be able to predict a client’s future financial transactions before offering a loan, for example. Furthermore, it has an enormous potential to truly revolutionise the way recruitment is conducted; a recruiter will soon be able to determine a candidate’s reliability, competitiveness and success rate by using algorithmic personality detection software.
It’s a really exciting development and its success could truly disrupt a great deal of industries.