Updated: Nov 27, 2020
Looking into the AI mirror
At some point in our lives, we’ve spent a, later realised, laughable amount of time taking tests to humour or find out insights about ourselves we may not have been aware of. Believe me, I’ve been there. From ridiculous ones that seem to be written by an eccentric in a basement, to extensive ones used in work settings, we have all wondered: who is asking us these questions and am I really a Phoebe?
Human psychology and the path to learning about our complex minds fundamentally peaks everyone’s interest: who am I? And this soul-searching process has undergone an exhaustive course of research and development, with answers still dubious. Lobotomies, which is one of the few psychiatric treatments to win a Nobel Prize, were once widely administered to treat mental illness and is now obsolete. Who’s to say all our current knowledge and treatments won’t be discredited in the next century, or even decade?
We’ve come a long way from drilling holes into our brains, extreme cold or hot baths, and electro-shocks that we now see in Netflix psychological thrillers. And as we climb Maslow’s Hierarchy, we seek information in the pursuit of self-actualisation. Personality tests are a vehicle into learning more about our identities, strengths, weaknesses, relationships, careers, and purpose in life. That’s why a fair amount of time, research and money goes into developing these tools to further society’s understanding of themselves as a whole.
With such a subjective topic, the research and creation of these tests start with constructs – specific characteristics we can measure. These need to be well-defined (where definitions are often debated in literature) and ideally need to originate from existing theory, or by developing new theory. Once you have the theory, you have the constructs, then you’ll need the questions. And then you’ll have to test whether the test really tests. It’s actually a complex process: trial-and-error, pilot studies, sampling, statistical analysis and continual development overtime.
It’s fascinating because sometimes we don’t fully appreciate the R&D that goes into the everyday things we come across.
A popular one you might want to try is ‘16 Personalities’ – a development of Carl Gustav Jung’s introvert and extrovert personalities and the Myers-Briggs test. The results can provide insight into personality traits, show famous people who have the same personality type, talk about strengths and weaknesses, romantic relationships, friendships, parenthood, career path and workplace habits.
But have you heard of an AI driven personality test? Zelfium, like 16 Personalities, asks 76 questions to give you your personality type ‘code’ by using big-data analysis through machine learning. Where it differs is that instead of being able to deliberate your answers, you’re on a timer. AI then factors in your answer scores, response time and profile information to produce a more accurate result: comprising of additional categories to other personality tests.
Sometimes it is frustrating when you are required to check a box for a question because your answer isn’t that simple. With R&D driving AI’s application, the ability to input and analyse detailed answers, including behavioural factors such as eye movement, will allow personality testing to go even deeper. The technology is yet to reach it’s full potential in this sector, but we’re excited about the future implications.
Have you ever taken a personality test? Share your thoughts below.