The Intelligent Friend - The newsletter about the psychological, social, and relational aspects of AI, based only on scientific papers.
Hello IF readers! This is the tenth issue of Nucleus, where you find actionable insights from research papers, links to articles, and useful resources related to AI's social, psychological, and relational aspects. I'm happy to announce that this is one of the first issues I've written with a 'guest' writer in my newsletter.
In fact,
collaborated on this issue, recently launching its newsletter regarding gamification in science, AI, and beyond: . This will be an experimental issue because it will be thematic: the papers will mostly concern the topic of gamification and AI. If you like this, don't hesitate to let me know by writing to me in chat or reposting on Note.I found his approach to content captivating and I suggest you, if you are interested in the topic, to seriously take a look at what he does. Enjoy this issue!
Psychologically advanced characters
By
As the world runs in circles trying to figure out what AI, and in particular, generative AI, means for them, researchers in the gaming industry are quietly finding ways to supercharge game creators. In Nucleus #9, Riccardo shared a study about personality traits in large language models: this study applies those generated personality traits to procedurally generate NPCs in turn-based RPGs.
They created a framework called PANGeA (Procedural Artificial Narrative using Generative AI), which will “generate narratives in turn-based role-playing games”. It works with a high-level set of instructions and parameters set by the game designer, and constraints like period and world setting. PANGeA then generates NPCs, using those parameters as well as the Big Five Personality Traits to be more human-like. These NPCs allow players to have a more natural interaction with them as opposed to the traditional, scripted dialogue options.
The generated characters also include a validation feature to provide game-appropriate responses in case they are asked questions outside the scope of the game. PANGeA integrates with Unity, one of the biggest game development engines, so integration into the development process should be fairly seamless.
There are so many cool applications for this kind of NPC generation and I’m excited to see if developers can bring them to life. It’s the kind of application that makes ME want to get into game dev!
Title: PANGeA: Procedural Artificial Narrative using Generative AI for Turn-Based, Role-Playing Video Games. Author(s): Buongiorno et al. Year: 2024. Journal: / (working paper). Link.
How do you behave when you play?
By
I’ve always been hesitant playing multiplayer online battle arena (MOBA) games because of their highly competitive nature. Competition, in turn, sometimes brings out the worst in people, and is sometimes made even worse in the virtual world. The act of deliberately irritating and harassing other players is known as griefing, and the perpetrators, griefers.
Many game developers have tried addressing griefing but the volume and complexity of data to be analyzed has thus far been too great. Enter GrieferLens, an AI-powered visual analytics tool that annotates and analyzes player behaviour during a match for experts to review. GrieferLens analyzes behaviours like non-participation or AFK to flag any suspicious behaviours. These results are then quickly reviewed by an expert to make a decision regarding griefing behaviours.
Experts using the tool noted that the time it took them to complete a review was reduced from 25 minutes to just 5 minutes. The visually intuitive nature of the tool also helped identify outliers easily. While the tool was only tested for League of Legends, a tool like this could be applied broadly to any multiplayer online game, looking at game-specific activities and metrics that will ultimately make these virtual environments just a little bit safer.
Title: Towards an Exploratory Visual Analytics System for Griefer Identification in MOBA Games. Author(s): Chen et al. Year: 2024. Journal: / (working paper). Link.
The issue(s) of the week
As you may have read (or maybe not!) I recently obtained my Master's Degree in Marketing (I'm super happy!) so I haven't had time to read much. By now you know that I strongly recommend only what I read personally, so this week I recommend only one issue (promising to recommend many more next week) by
, who has built a framework regarding AI tools in education and which in addition to being interesting (like everything Nick writes) is based on research and educational practice. I found it an exciting and inspirational effort, I recommend reading it.Personalized AI Gamification
This paper is an editorial that explores how gamification can be tailored to individual user needs to enhance engagement and outcomes across various contexts. The paper comprises three main studies, each focusing on different applications of gamification. I'll tell you about them briefly one by one because they have different actionable insights:
The first study explores the use of conversational agents (CAs) in educational settings. The findings highlight that while game-inspired PCAs have significant potential, the field requires more prescriptive design knowledge, diverse game elements, advanced AI algorithms, and intelligent adaptation to fully realize their benefits.
The second study examines the impact of gamification on fitness app users. Drawing on flow theory, the analysis reveals that gameful experiences and personalization preferences lead to higher levels of flow, where users are fully immersed in the activity (very interesting). This immersion correlates positively with user satisfaction, which in turn enhances loyalty and word-of-mouth intentions.
Keep reading with a 7-day free trial
Subscribe to The Intelligent Friend to keep reading this post and get 7 days of free access to the full post archives.