Your partner is jealous of ChatGPT
How people sometimes misperceive the use of AI.
The Intelligent Friend - The newsletter about the AI-humans relationships, based only on scientific papers.
Before we begin, I first wanted to thank all the new readers of this newsletter! I hope you will enjoy this issue and future ones. For everyone: if you haven't already done so, here is a survey where you can tell me who you are, what you do and what you would like to read in future issues. It lasts less than a minute and is super nice to know a little more about you!
Intro
It may have occurred to you, in relation to your partner or to a close friend, that at one time that person was not giving the right weight to you and your relationship. Perhaps it was superficial. Or forgot something important. Well, this kind of negativity can also happen because of the Artificial Intelligence. And today's authors tell us why.
The paper in a nutshell
Title: Artificial intelligence and perceived effort in relationship maintenance: Effects on relationship satisfaction and uncertainty. Authors: Liu, Kang, and Wei. Year: 2023. Journal: Journal of Social and Personal Relationships.
Main result: AI-augmented relationship maintenance behaviors could lead to perceptions of unbalanced partner effort and compromise the desired relational equity and relationship satisfaction.
Starting from the investment model and equity theory, the study tries to understand what the effect of using ChatGPT (and AI) for a relationship with another person is on the perception of that person.
Relationships and efforts
Imagine it's your birthday. Your partner or best friend finally gives you their present. It is that person's gift, so it is especially important. You open it and... you like it! You have chosen something original but fitting, similar to what you expected but with a touch of surprise.
You are happy, you thank him sincerely and you are glad for having such a person by your side every day. But then he or she reveals to you that he / she was so good in choosing the gift thanks also to ChatGPT's help: starting with his / her idea, he then elaborated and evaluated a series of options, and then chose the final one.
I guess you feel angry. But how come? Yet, the gift is the same. The person is the same. The choice, too, was basically the same. But the use of AI makes you nervous. It makes you feel that the person took less effort to find the gift than you imagined. Well, the study tests this specific feeling.
Despite the fact that we are all good friends and partners, and we want to be objective and want to be less analytical as possible in personal relationships, we weigh what the other person does. To keep a relationship going, we maintain it. According to Ogolsky & Monk (2018), relationship maintenance is the mental and behavioral efforts to maintain a relationship as stable and fulfilling1.
In this context, of course, we can use AI to help us manage our relationship. For example, we could ask how to write something sensitive or problematic, advice for a birthday present, or create something personal. I know, it will sound strange to some. But it is not so different from other behaviors we already do. The reminders on your iPhone, the internet searches, the cosigning to a friend.
I would consider them as practices similar in substance to the AI approach, with the main difference being the human person (in the last case) or another type of tool (in the first two).
When we talk about behavior to maintain a relationship, we usually refer to efforts that have three main characteristics2:
ensure the continuation of a relationship;
maintain the stability of the relationship;
keep the relationship satisfying for both parties involved.
But, of course, when we are in a relationship we make a number of assessments because our real goal is to understand how the other person (friend or partner) behaves. This, in fact, leads us to be more or less satisfied, and to be uncertain about the other person's equal willingness to continue the story.
In analyzing relationships, scholars focus on two outcomes:
OK, it is clear what is to be taken into account. But how do we actually make these assessments? What is it that we actually put under analysis? What actually makes us dissatisfied?
Investment model and equity theory
As you may have guessed, I like explaining scientific concepts by using examples from the most common everyday life. So, imagine this situation.
It's your birthday (again, a day of celebration!). All your friends are there, except George, a surgeon at the local hospital with a long shift to support. As part of the party, you don't think much about his absence, but in some cases you feel it. While you are talking to one of your friends about how interesting
's new book you are reading about AI6 is - I actually bought it and am waiting for it! - there is a knock at the door. It's George! He tells you that he left his shift an hour early so that he could go out there. You are really surprised, because, you think, 'wow, who knows what it must have cost him to be like that after that exhausting shift'.
The investment model, the first important framework considered in the paper, illustrates a dynamic that in a sense we often guess: when we look at an interpersonal relationship, we evaluate it in terms of the investments a person makes7.
These investments are costs of various kinds (money, time, etc.) that the person bears in order to maintain the relationship. The more this investment increases, the more theoretically the commitment to the relationship should increase, as should the satisfaction of the other person8.
However, not all costs are equal: those that are not 'recoverable', such as money or time, are the ones most valued by partners9. On this, scholars cite a study that I found really interesting Camerer (1988) observed that while efficient gifts like cash or precisely what the partner requested (high utility gifts) are seen as impersonal and less valuable, less efficient gifts that involve personal thought and effort in selection are more valued and common in relationships (exactly the initial example of this issue)10.
Personal investments are a delicate dynamic. Having a partner who is warm-hearted to us must be really exciting. Or a friend who cares a lot and shows it. Maybe he calls us every day to ask how we are or what we did today. But if we don't call him for a day, for the second day, and for quite a while, he might be upset. He may think that, on the other hand, you are not investing as much as he is investing. We've seen that this investment signals a person's commitment, right? Well, your repeated failure to act could be a sign of unequal behavior between the parties.
Equity theory, the second important perspective used by the study, stresses the idea that people pay attention not only to the costs suffered by the other person, but also to what is the equity between the costs borne by themselves and those of their own friend or partner.
The role of AI
Now, what does technology, and especially AI, have to do with this? First, we use technology to maintain our relationship more than we think. Social media, messages, video calls. They are all concrete examples. There are even several cases in which partners get angry because the other side isn't sharing the story they posted on Instagram from their boyfriend. And although technology, when it has a low weight in the relationship, can have positive results11, personal effort matters. Indeed, some scholars have studied the relationship between likes or messages and the escalation of a relationship. Guess what: it's the seconds that predict them the most, not the likes12.
Whether or not it may surprise you, there are many ways that AI can be used in a relationship13. However, AI makes our lives easier. In other words, it decreases our effort needed to maintain the relationship. But the other person is no less smart than us. When she realizes this, she could perceive this act very negatively, as it would signal a lower commitment on your part and could lead the person to be more uncertain about the relationship. This is exactly what scholars hypothesize.
To test this, the authors conducted an online experiment with a 3 (agency: self-without-augmentation vs. AI-augmented vs. human-augmented) × 3 (relational task: support-giving vs. advice-giving vs. birthday) design celebration) design, focusing on friendship contexts. Participants interacted with a hypothetical friend named Taylor, who responded to them via text message tailored to the relational task—either offering support, advice, or birthday greetings.
To vary the “agency”, in the AI-augmented scenario, participants were informed that Taylor utilized an AI content generation platform called "do-re-mi" to formulate responses. In the human-augmented condition, it was explained that Taylor engaged an online writing community also named "do-re-mi," drawing on its members to help draft the reply. In the control condition, known as self-without-augmentation, participants were told that Taylor crafted their response independently.
Interestingly, the researchers' findings show that actually using AI both decreased the satisfaction and increased the uncertainty of the other party in the relationship. Furthermore, the reason is precisely represented by the perceived effort: since with AI we think that there is less effort on the part of the other person, we believe that he is less committed and we begin to think that he is not investing the right amount in us.
However, these results do not show a significance of the difference between the "AI-augmented" and "Human augmented" conditions.
This means that there is almost the same negative result whether the help comes from the AI or from another person. Really intriguing!
Before moving on to the take-aways and research questions, I wanted to hear your opinion on a topic:
commented one of my notes (P.S. thank you a lot!) that the best way to offer a paid option is doing it at the beginning. I started this newsletter recently, but I think it is very valid advice, especially to accompany the growth path.In this regard, I was thinking of creating a weekly appointment for premium subscribers in which I choose 4/5 interesting papers on AI and summarize them as a Briefing (4-5 lines each). It could be interesting for those who don't have the time to read papers but they would be interesting to be constantly updated (the papers would be on various topics related to AI).
Please tell me your honest opinion, both on the format and on the possibility of paid support. I would be really happy about it!
Take-aways
Make the effort without AI, and you will be rewarded: The study reveals that when AI is used to assist in relationship maintenance, it leads to a perception of reduced effort by one’s partner. This perception negatively impacts relationship satisfaction, as the effort put into maintaining the relationship is a crucial factor in how valued and satisfied individuals feel.
Uncertainty matters: the utilization of AI in relationship tasks also increases uncertainty within relationships. Participants felt more unsure about the stability and future of their relationships when AI was involved in tasks typically requiring personal effort and emotional investment.
It's not who that matters, but what: comparing AI-assisted and human-assisted scenarios, the study found that there were no significant differences in perceived effort between these two. However, scenarios without any assistance were viewed as more appropriate and led to higher relational satisfaction, suggesting that personal effort is highly valued in relationship maintenance.
Further research directions
Investigate how AI augmentation affects interactions between actual friends or partners across different types of relationships, examining its long-term impact on both relational and individual well-being;
Discover the specific mechanisms and outcomes of AI in the maintenance of relationships, exploring how it modifies relational dynamics and affects perceived closeness and support;
Study reciprocal actions and perceptions of effort and commitment, particularly how AI users might attribute improved relational outcomes to their own efforts, potentially skewing self-perception of commitment and affecting the sustainability of relationship satisfaction.
Thank you for reading this issue of The Intelligent Friend and/or for subscribing. The relationships between humans and AI are a crucial topic and I am glad to be able to talk about it having you as a reader.
Has a friend of yours sent you this newsletter or are you not subscribed yet? You can subscribe here.
Surprise someone who deserves a gift or who you think would be interested in this newsletter. Share this post to your friend or colleague.
P.S. If you haven't already done so, in this questionnaire you can tell me a little about yourself and the wonderful things you do!
Ogolsky B. G., Monk J. K. (2018). Maintaining relationships. In Vangelisti A. L., Perlman D. (Eds), The Cambridge handbook of personal relationships (pp. 523–537). Cambridge University Press.
Dindia K., Canary D. J. (1993). Definitions and theoretical perspectives on maintaining relationships. Journal of Social and Personal Relationships, 10(2), 163–173.
Arriaga X. B. (2001). The ups and downs of dating: Fluctuations in satisfaction in newly formed romantic relationships. Journal of Personality and Social Psychology, 80(5), 754–765.
Fuller-Iglesias H. R. (2015). Social ties and psychological well-being in late life: The mediating role of relationship satisfaction. Aging & Mental Health, 19(12), 1103–1112.
Theiss J. A. (2020). Uncertainty perspectives on relationship maintenance. In Ogolsky B. G., Monk J. K. (Eds), Relationship maintenance: Theory, process, and context (pp. 69–85). Cambridge University Press.
https://www.amazon.com/Co-Intelligence-Living-Working-Ethan-Mollick/dp/059371671X
Rusbult C. E., Martz J. M., Agnew C. R. (1998). The investment model scale: Measuring commitment level, satisfaction level, quality of alternatives, and investment size. Personal Relationships, 5(4), 357–387.
Kubacka K. E., Finkenauer C., Rusbult C. E., Keijsers L. (2011). Maintaining close relationships: Gratitude as a motivator and a detector of maintenance behavior. Personality and Social Psychology Bulletin, 37(10), 1362–1375.
Camerer C. (1988). Gifts as economic signals and social symbols. American Journal of Sociology, 94(Suppl), S180–S214.
Camerer C. (1988). Gifts as economic signals and social symbols. American Journal of Sociology, 94(Suppl), S180–S214.
Taylor S. H., Zhao P., Bazarova N. N. (2022). Social media and close relationships: A puzzle of connection and disconnection. Current Opinion in Psychology, 45, 101292.
Sosik V. S., Bazarova N. N. (2014). Relational maintenance on social network sites: How Facebook communication predicts relational escalation. Computers in Human Behavior, 35(12), 124–131.
Hancock J. T., Naaman M., Levy K. (2020). AI-mediated communication: Definition, research agenda, and ethical considerations. Journal of Computer-Mediated Communication, 25(1), 89–100.
Cover credits: BBC (with the addition of the ChatGPT logo)