Two faces of AI use in the workplace: implicit versus explicit

In her bachelor’s thesis “Discrepancy between explicit and implicit attitudes towards the use of artificial intelligence in the work context”, Cheyenne Schmidts, a psychology student at Macromedia University in Stuttgart, wants to find out whether there is a significant difference in this respect and which personality traits may influence it.

thebroker talks to Cheyenne Schmidts.

A study by Cognizant from last year states that generative AI will generate annual growth of up to 1 trillion US dollars by 2032, but at the same time could destroy 90 percent of existing jobs. What do you think? Link here.

On the one hand, of course, I’m also aware that people from a wide range of industries are worried about losing their jobs or what dangerous products, such as images, could be created with the help of AI. On the other hand, looking back, our society has already faced a technological change like this umpteen times and each time this change has been associated with fear and horror. That’s why I’m trying to embrace what’s coming and have a little faith in the whole thing. I think it is much more important, as with any other progress, to provide the right information and learn how to deal with it sensibly.

The title of your bachelor’s thesis is “Discrepancy between explicit and implicit attitudes towards the use of artificial intelligence in the work context and which personality traits may influence this”. Explain what is meant by this.

In the literature, a person’s attitude is divided into explicit, the conscious attitude, and implicit, the unconscious attitude. This is because our actual (unconscious) attitude towards an object, which is not directly accessible to us, is often overshadowed by influences such as social norms, as a result of which we may (consciously) express a different attitude to the one that is actually deeply rooted within us. There is a discrepancy between implicit and explicit attitudes. In addition, our attitude significantly influences our behavior. In cases where the implicit and explicit attitudes are not congruent, it appears as if our behavior towards an object does not match our attitude.

An example of this: My demonstrated behavior (“I use little to no AI assistants”) is based on my actual, implicit attitude (“I am afraid of AI assistants taking over my job”). However, I know that the requirements of my job (“You have to be open to change and be able to adapt flexibly”) contradict this attitude, which is why I express a different attitude, the explicit one (“I am open to the use of AI assistants”).

In addition, I assume that people who have a higher degree of adaptability in their personality have a significantly lower to no discrepancy between their implicit and explicit attitude towards the use of artificial intelligence in the work context and are therefore more likely to use AI assistants in this area.

How did you come up with this topic?

I study brand and advertising psychology, so I am particularly interested in which psychological backgrounds can be linked to marketing and used in a meaningful way. And it is now no secret that it is not only a person’s conscious processes, such as actively searching for information about a product, that contribute to a purchase decision/behavior, but also unconscious processes such as emotions or attention. As the decision-making of relevant decision-makers is seen as relatively rational in a professional context, I was interested in whether this is really the case, as these decision-makers are also nothing more than people made of flesh and blood and with a brain that follow certain principles. My topic was then modified a little and adapted to the latest state of research

Why do you relate these facts to all insurance intermediaries?

The latest developments in the insurance industry (GDV) show that AI is increasingly finding its way into this field of work and that the area of tension between insurance intermediaries is therefore particularly topical and particularly intense. I therefore expect a very interesting and meaningful result compared to less affected sectors.

Your investigation is based on a study you conducted yourself. Does this only apply to Germany?

For financial reasons, I was only able to develop the study for German-speaking countries. This naturally includes insurance brokers from Austria and Switzerland. Here is the link to take part in the study.

What does it ask?

First of all, a self-disclosure questionnaire asks about explicit attitudes, i.e. opinions and thoughts, on AI. This is followed by almost 30 questions that measure individual adaptability, also via a self-report questionnaire. Finally, a reaction time measurement takes place, which can be used to draw conclusions about a person’s implicit attitude.

In the reaction time measurement, the participants have to assign words shown to a given category as quickly as possible. What do you learn from this?

The Single Category Implicit Association Test (SC-IAT) follows the principle that people whose implicit attitude towards AI is positive can assign the words in this category to positive words more quickly than negative words. The words are assigned in such a short time that conscious decisions are not possible and therefore other factors (e.g. social norms) that overshadow the implicit attitude cannot play a role, as is usually the case.

How many insurance brokers have taken part in the study so far and when does the survey end?

As of today (May 21st, 2024), 57 people have accessed the link, but only 27 have taken part by the end. In order to acquire enough test subjects for my study, the study will run until June 7, 2024.

As a budding psychologist, what will you gain from the results of the study?

First and foremost, it’s about creating and conducting empirical research, which has to be completed as part of my Bachelor’s thesis. I am not making any economic or financial gain from it. Out of personal interest and ambition, I would be very happy about a significant result that, in the best case, confirms my assumptions on the topic already mentioned. Furthermore, significant results would also confirm that “rationally thinking” people in a decision-making role are not as rational as people think. This in turn could stimulate further research, particularly in the field of marketing, in order to apply classic models of purchasing decisions, as we know them from private consumers, to decisions in a professional context.

Which AI tools do you use yourself?

I use the AI assistants that very few people know they are: Siri from Apple, other smartphone functions, SmartHome and ChatBots. I’m still a little reluctant to use the classic AI assistants, such as ChatGPT.

What are you using AI for in your bachelor’s thesis?

I haven’t deliberately used artificial intelligence for my bachelor’s thesis yet. But I probably just wasn’t aware of it.

How many of your fellow students do you think are using AI for their Bachelor’s theses?

Actually, we recently had a brief discussion about this in a small group and it turned out that the vast majority use AI for smaller tasks, such as translations or literature research.

Cheyenne (Kassidy) Schmidts, 21 years old, a bachelor’s student of psychology at Macromedia University in Stuttgart, is currently researching the topic “Discrepancy between explicit and implicit attitudes towards the use of artificial intelligence in the work context” as part of her bachelor’s thesis.

A year ago, she and her fellow students conducted research on the topic of “Emotion and attention – The amount of visual attention of the six basic emotions”. This study showed that, in contrast to previous studies, emotional valence makes no difference to the generation of attention.

In addition to the exciting content of her studies, Cheyenne is also interested in sporting activities in her free time. Jogging, swimming and weight training are her primary pastimes. When she has a little more time on her hands, Cheyenne enjoys traveling to distant countries or simply relaxing and skiing in the mountains of South Tyrol.

Read also: ATTENTION: AI use from a legal perspective