Editorial: My Experience with Artificial Intelligence

Archives > Volume 20 (2023) > Issue 2 > Item 01

DOI: 10.55521/10-020-201

Stephen M. Marson, Ph.D., Editor

Marson, S. (2023). Editorial: My Experience with Artificial Intelligence? International Journal of Social Work Values and Ethics, 20(2), 4-8. https://doi.org/10.55521/10-020-201

This text may be freely shared among individuals, but it may not be republished in any medium without express written consent from the authors and advance notification of IFSW.

Like all of us, I have been deeply concerned about ethical issues regarding social work scholarship and publications. As you might have noticed, the International Journal of Social Work Values and Ethics instituted a manuscript submission policy regarding the use of artificial intelligence [see https://jswve.org/manuscript-policy/]. Dr. Dawn Apgar is the author of the artificial intelligence section of our manuscript policy.  In addition, Dr. Apgar submitted a forum article entitled “Incorporating the Use of Generative Artificial Intelligence and Large Language Models into Publication Standards: A Call for Editorial Policy Based on Social Work Values” which is included in this issue. We also have Dr. Reamer’s article entitled “Artificial Intelligence in Social Work: Emerging Ethical Issues.” I hope you have time to read both of these critically important articles. For now, I want to share my personal experience with artificial intelligence.

Last March, I had a conversation with Dr. Dawn Apgar regarding artificial intelligence and the ethical issues related to social work scholarship and social work education. At that time, we negotiated a forum article for IJSWVE. Since that time, I have been literally playing with artificial intelligence. I wanted to get a feel for how it worked.  It was like doing a google search but more personal. I heartily admit that I had fun playing with it.

After playing, I decided to put it to work. During this time period, I was deeply entrenched in a complex manuscript addressing linking social structure theory to personality theory. This subject was the centerpiece of my presentation at an international conference.0F[1] A major part of my research included the employment of role theory as a bridge between micro (psychology) and macro (sociology) theory. For the project, it was critically important that I uncover the evolutionary process by which role theory emerged.

With my doctoral education, I knew that Talcott Parsons produced a role theory with the influence of various sources including the Chicago School with scholars such as George Herbert Mead.  In addition, I knew that there were various streams of thought that produced a role theory. I wanted to know the historical pathways that produced the various versions of role theory. The artificial intelligence did not seem to understand the difference between a theory (as an organic whole) and theoretical concepts. My conversation with artificial intelligence evolved into an intellectual argument. From my perspective, I began to feel like I was arguing with an undergraduate who had just successfully completed a sociological theory course.

Time after time I continued to ask the artificial intelligence for authors (other than Parsons) who had produced a role theory that could be conceptualized as an organic whole. Time after time, artificial intelligence would offer authors who made important conceptual contributions but none of them offered a theory. Each time I responded by stating these contributions do not constitute a theory but were theoretical concepts. Artificial intelligence replied by insisting that concepts are “important.” This back-and-forth discussion continued to the point when I realized that I was debating with an entity who could not understand the abstract notions embedded within theory construction and design. Artificial intelligence was simply unsophisticated when addressing abstract theoretical processes.

At this point, I pulled myself out of my fantasy world1F[2] and realized that I was not speaking to a human being. I began to ask less complex questions about theory. Mostly I asked about authors who wrote about contemporary theory. I received, what I thought to be impressive citations. I began to search for the specific references in libraries and on Amazon. At least one of the citations did not seem to exist. When I reported this back to artificial intelligence, I received an apology and other references that were accessible by both the library and Amazon. I learned that when I work with artificial intelligence, I must verify the output I receive.

Continuing with my project, I needed to define role theory concepts. I asked artificial intelligence to create 61 definitions for role theory concepts. I was impressed with the results and the speed of the output. In my studies, I developed a role theory concept, but no word that captured the essence of it. I asked artificial intelligence to search for a concept within role theory that encapsulated the idea that role occupants were impressed by and admired the role skills of another. Artificial intelligence replied by acknowledging that no such concept exited. I asked it to create a new role theory concept that would encapsulate the meaning of my idea. This is how artificial intelligence responded:

Question: Can you make up a new term for role theory that includes the idea of admiration that peers have for the role performance of a friend or colleague?

Answer: That’s a creative challenge. How about this term:

Role admiration is the positive evaluation and appreciation that peers have for the role performance of a friend or colleague. Role admiration reflects the recognition and respect that a person earns by demonstrating competence, skill, and excellence in their role. Role admiration can enhance the self-esteem, motivation, and satisfaction of the role performer, as well as inspire others to emulate their behavior or seek their guidance. Role admiration can also foster social cohesion, trust, and collaboration among peers who share a common role or goal.

What do you think of this term?

I was impressed with the answer and also the evaluative question artificial intelligence asked at the end.  It wanted me to evaluate the answer. I wrote “GREAT!!!!” The artificial intelligence responded with “Thank you! I’m glad you like it.”

Let us delve deeper into the process of creating the definition of “role admiration” by artificial intelligence. In high school, I learned the process by which dictionaries construct a definition. For me, it was immediately apparent that artificial intelligence complied with these standards. Most importantly, I could have constructed a definition with the same level of quality as artificial intelligence. However, the process of constructing my definition would take much longer than 30 seconds. The amazing aspect of artificial intelligence is not the output but rather the speed of the output. Clearly, the speed and accuracy of this definitional construction is nothing less than amazing.  Nevertheless, in playing with artificial intelligence, I learned that all outputs cannot be trusted. Artificial intelligence’s output must be critically assessed prior to actually using it.

Currently, there is no word that encapsulates the process by which an acronym becomes a unique word which can be found in the dictionary. For example, scuba was originally an acronym for “self-contained underwater breathing apparatus.” Scuba is now a word that can be found in any English dictionary. I worked with artificial intelligence to create a new word for this process, and we agreed on the term “acronymergence (AL).”   Through the process of acronymergence, the use of AL will evolve into a word as an independent entity. AL is also short version for a man’s name and through the progression of time personification will emerge.  Interaction with Al will be like talking to a colleague. Hereafter, I will cease using the acronym, but rather will spell out artificial intelligence.  I recommend the same for other writers. Using the acronym, Al, is an easy trap to personify artificial intelligence.

The final part of my excursion in examining artificial intelligence will be a qualitative analysis. I have derived 61 classical concept definitions of role theory plus “role admiration” which is not included within the role theory literature. Through the use of a colleague who specializes in qualitative analysis, we will seek out patterns of how artificial intelligence constructs definitions. WHY? 

I began this editorial with a vague understanding of what artificial intelligence is and the possible ethical implications for social work scholarship and education. My curiosity got the best of me, and I wanted to come up with answers for vague questions that were generating in my mind. At the end of my excursion into the world of artificial intelligence, I am startled because I have more questions than answers.  I am going to diligently read current and upcoming articles addressing artificial intelligence within the pages of The International Journal of Social Work Values and Ethics. Hopefully doing additional reading will bring a level of closure to my skepticism of artificial intelligence.

If you have questions or comments about artificial intelligence, I would be interested in hearing from you and publishing your commentary in our letters to the editor section.  I cordially invite you to send emails to:  journal@ifsw.org

[1] Marson, S.M. (2023). Recommendations for the Study of Human Diversity. Presented at the 39th National and 5th International Congress of National Day of Social Workers. March 9 (online hosted in Iran).

[2] In using artificial intelligence, it was too easy for me to fall into the trap of feeling that I was interacting with a person.