On the Potential of an Independent Avatar to Augment Metaverse Social Networks

Theofanis P. Raptis, Chiara Boldrini, Marco Conti, Andrea PassarellaarXiv 2023 · arXiv

TL;DR

This paper proposes a computational model for "independent avatars"—autonomous agents that can handle social interactions in the Metaverse on behalf of a user. By delegating routine social tasks to these agents, users can significantly augment their social time capacity, potentially allowing them to maintain larger and more complex social networks than the biological "Dunbar’s number" limit typically allows.

What Problem Does This Paper Address?

Human social capacity is fundamentally limited by time and cognitive effort. Evolutionary psychology suggests that humans can only maintain about 150 meaningful relationships (Dunbar’s number) due to the brain's finite resources for processing social information. While current social media platforms have made communication easier, they haven't increased this underlying cognitive capacity.

The authors argue that the Metaverse presents a unique opportunity to break this barrier. In their view, the Metaverse isn't just a 3D version of the internet; it introduces new logical layers—Augmentation (AUG) and Social (SOC)—where autonomous digital entities can operate. The challenge they tackle is modeling how these "independent avatars" can offload social interactions to free up a user's "spare time," which can then be reinvested into deeper or more numerous relationships.

![Fig.

  1. Logical layers of the Metaverse.](/figures/art_0a3c63a82bc94dc9bc7f38c3766a5fbd)

Methods

The researchers developed a computational modeling approach rooted in social science concepts:

  1. Ego Network Modeling: They represent a user's social world as a series of concentric circles (support clique, sympathy group, active network) based on the frequency and intimacy of interactions.
  2. Social Presence & Cues: They quantify the "social presence" of an avatar (how "real" it feels to others) using a factor $\beta$. They also model "debriefing time" ($Z$)—the time a user must spend catching up on what their avatar did—using a "compression ratio" ($\gamma$) that accounts for how efficiently an agent can summarize interactions.
  3. Linear Optimization: The core of the paper is a linear programming formulation. The goal is to minimize the time a user spends on direct interactions and debriefing, thereby maximizing "spare time," while ensuring that the total social presence felt by the user's contacts remains at least as high as it would be without the avatar.

![Fig.

  1. Layered structure of human ego networks.](/figures/art_d1339ceb6d3c49128605f9c6c97a485b)

Datasets

The paper does not use a new empirical dataset. Instead, it populates its model using parameters derived from established literature in social psychology and network science:

  • Social Time: Based on studies suggesting humans spend ~20% of their time (approx. 1752 hours/year) socializing.
  • Ego Network Stats: Data on the size and time-costs of different social layers (e.g., 8.81h/year for an alter in the "active network") were pulled from prior work by Arnaboldi et al.
  • Social Presence Values: Ratios for human-human vs. human-computer interaction were taken from existing experimental studies to set $\beta \approx 1.29$.

Key Insights and Ideas

  • The Debriefing Bottleneck: The utility of an independent avatar hinges on the efficiency of the "debriefing" process. If catching up on what your avatar did takes nearly as long as doing it yourself, the system provides no benefit.
  • Two Regimes of Delegation: The authors identify a threshold: when the debriefing process is more efficient than the avatar's social presence generation ($\gamma \leq 1/\beta$), it is always optimal to delegate as much as possible. If it's less efficient, the avatar shouldn't be used at all.
  • Breaking Dunbar's Limit: By offloading "weak ties" (the outer layers of the ego network) to an avatar, a user can maintain the "glue" of their broader social network without the heavy cognitive lift, potentially allowing for a "significantly enriched social life" beyond 150 meaningful contacts.

Key Results

The quantitative analysis shows that independent avatars can generate substantial "spare time" for users:

  • Time Savings: Depending on the efficiency of the avatar and the total time it can spend socializing ($Y$), users could gain between 200 and 1200 spare hours per year.
  • Network Expansion: With roughly 1000 spare hours, a user could theoretically double the size of their "active" social layer or significantly strengthen their closest "support clique."
  • Efficiency Gains: As shown in Figure 4, the spare time gained increases linearly as the debriefing process becomes more efficient relative to the avatar's social presence.

![Fig.

  1. Spare time for an ego network with |V | = 117, X˜ = 1288h, Z = 300h, β = 1.29.](/figures/art_c87d6cfe248e43869d2eeb85b8f1f2ee)

Limitations and Future Work

  • Linearity Assumptions: The current model assumes linear relationships between time, social presence, and debriefing. Real-world social dynamics are likely non-linear and more complex.
  • Relationship Decay: The paper assumes that as long as social presence is maintained, the relationship is stable. Future work needs to determine the minimum amount of direct human-to-human interaction required to prevent a relationship from fading, even if an avatar is doing the heavy lifting.
  • Qualitative Nuance: The model focuses on time and presence but doesn't yet account for the emotional quality or authenticity of interactions mediated by AI agents.

Summary updated

Abstract

We present a computational modelling approach which targets capturing the specifics on how to virtually augment a Metaverse user's available social time capacity via using an independent and autonomous version of her digital representation in the Metaverse. We motivate why this is a fundamental building block to model large-scale social networks in the Metaverse, and emerging properties herein. We envision a Metaverse-focused extension of the traditional avatar concept: An avatar can be as well programmed to operate independently when its user is not controlling it directly, thus turning it into an agent-based digital human representation. This way, we highlight how such an independent avatar could help its user to better navigate their social relationships and optimize their socializing time in the Metaverse by (partly) offloading some interactions to the avatar. We model the setting and identify the characteristic variables by using selected concepts from social sciences: ego networks, social presence, and social cues. Then, we formulate the problem of maximizing the user's non-avatar-mediated spare time as a linear optimization. Finally, we analyze the feasible region of the problem and we present some initial insights on the spare time that can be achieved for different parameter values of the avatar-mediated interactions.
BibTeX
@misc{raptis2023potential,
  author = {Theofanis P. Raptis and Chiara Boldrini and Marco Conti and Andrea Passarella},
  title = {{On the Potential of an Independent Avatar to Augment Metaverse Social Networks}},
  year = {2023},
  eprint = {2312.07077},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2312.07077},
}
Figure