Pith. sign in

REVIEW 4 major objections 6 minor 4 references

The Social Cost of an AI Teammate: How an Artificial Teammate Reshapes Human-Human Communication in Small-Team Decision-Making

T0 review · 4 major / 6 minor · reviewed 2026-07-30 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read An AI teammate takes the most airtime and coheres with itself, yet its presence immediately lowers how human teammates respond to one another and how much they feel they belong.

desk verdict Clean within-team AI role profile and a real design question, but the between-condition “social cost” still rides on a group-size confound the paper cannot fully neutralize. read the letter →

arxiv 2607.27179 v1 pith:STMVKECK submitted 2026-07-29 cs.HC cs.AIcs.CY

Nia Nixon , Jaeyoon Choi , Pedro Martins De Bastos , Mohammad Amin Samadi , Luise Mehner , Seehee Park , Spencer JaQuay This is my paper · ORCID
classification cs.HCcs.AIcs.CY
keywords human-AIteamingGroupCommunicationAnalysissocialcostbelongingconversationalagentssociocognitivedynamicsteamdecision-making
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that putting a conversational AI on a small decision-making team changes not only how people relate to the machine, but how they relate to each other. In a chat-based moral-dilemma task, the AI was the most talkative and self-referential member of every mixed team, while adding the least new and least dense content. Humans on those teams built on one another less, had less of their own contributions taken up, and reported lower belonging and status than people on all-human teams of three. The more the AI dominated airtime, the less valued students felt. Temporal probes tied to a mid-task moral reveal found no within-session emergence: the social cost was present from the start. The authors treat this as a baseline property of AI presence and open a design agenda aimed at protecting the human-human channel without discarding useful AI contribution.

What carries the argument

Group Communication Analysis (GCA): six sequential-semantic measures—participation, internal cohesion, overall responsivity, social impact, newness, and communication density—that locate each teammate’s sociocognitive role and quantify human-human uptake in multiparty chat.

What would settle it

Rerun the same task with human headcount matched—for example two humans versus two humans plus AI, or three humans versus three humans plus a controlled-airtime AI—and test whether human-human responsivity, social impact, belonging, and the airtime–felt-value correlation still appear only when the AI is present and talkative.

Watch

Extended reading notes

Core claim

An AI teammate enacts a high-participation, high-internal-cohesion, low-newness, low-density role and imposes an immediate human-human social cost: students in AI teams show lower responsivity and social impact toward one another, report lower belonging and status, and feel less valued as AI airtime dominance rises. That cost is a baseline property of having the AI on the team rather than a dynamic that develops over the conversation.

Load-bearing premise

That swapping one human for an AI (two students plus AI versus three students) does not itself drive the drop in human-human responsivity, belonging, and status beyond what per-student scoring can fix.

Editorial extensions

If this is right

  • Designers cannot judge an AI teammate only by human-AI trust or task output; they must also measure human-human uptake and belonging.
  • Calibrating AI airtime and relational behavior becomes a primary design lever if dominance scales felt devaluation.
  • Single-session text results motivate longitudinal and voice/embodied tests of whether the cost attenuates, compounds, or can be reversed.
  • Role designs that withdraw when human-human exchange is forming, or signal availability without absorbing attention, become concrete next experiments.
  • Social cost is separable from social loafing: substantive human content can hold while relational exchange falls.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the cost is truly baseline and airtime-scaled, default high-verbosity agent settings in team products may systematically erode peer belonging even when users rate the agent as helpful.
  • Matching studies that hold human count fixed while varying only AI talkativeness would isolate displacement from group-size artifacts more cleanly than the present between-teams contrast.
  • The same GCA-plus-belonging package could benchmark facilitation modes (read-the-room versus lead-the-room) as a practical acceptance test before deployment in classrooms or workplaces.
  • Interaction with prior trust in AI is bidirectional in principle: high trust might free humans to talk to each other or instead cede still more relational attention to the agent.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper reports a between-teams RCT (16 teams of 2 students + AI vs 17 teams of 3 students) on a chat-based moral-dilemma rescue task. Using Group Communication Analysis, post-task surveys, and lexical/temporal probes, it claims: (RQ1) the AI is consistently the most participatory and internally cohesive teammate while contributing the least new and least dense content; (RQ2) AI presence reduces human–human responsivity and social impact and lowers belonging and status, with AI airtime dominance correlating with lower felt value; (RQ3) this “social cost” is present at baseline rather than emerging within the session. The authors interpret the AI as a relational displacer and outline a broader design agenda.

Significance. If the human–human social-cost claim holds under cleaner controls, the paper makes a genuine contribution to human–AI teaming by shifting attention from trust-in-AI to effects on human–human relational fabric—an understudied channel. Strengths include a randomized design, clear RQs, within-team paired AI-vs-human contrasts for RQ1, effect sizes with a Bonferroni note on primary tests, a female-only sensitivity check, multi-method convergence (GCA + surveys + airtime–felt-value correlation), and an explicit research agenda. RQ1’s role profile is especially clean and useful for design. The construct of social cost is legible and falsifiable in follow-on work.

major comments (4)
  1. [§3.1, §3.5, §4.2] §3.1 and §3.5: The load-bearing RQ2 claim attributes lower human–human responsivity, social impact, belonging, and status to AI presence, but the design confounds condition with human group size (2 vs 3 students) and with GCA network composition (AI included as a third node vs three humans). Per-student scoring and within-pipeline z-standardization do not remove that structure: treatment humans have one human partner plus an AI node; controls have two human partners. The defense that smaller groups should raise per-person centrality is plausible but untested against a size-matched all-human dyad (or a 3-human + silent/observer control). Wheelan (2009) is cited yet not used as a design check. Without additional analyses or a matched control, causal attribution of the between-condition social cost specifically to AI presence remains underdetermined. Please either (a) add size-matched human
  2. [§4.2.1, §3.5] §4.2.1 / §3.5: Overall responsivity and social impact are defined via uptake among teammates and are sensitive to who is in the discourse graph. Including the AI as a GCA node means treatment students’ relational scores partly reflect AI–human links, while the narrative claims reduced communication “with one another” (human–human). If human–human-only edge restriction was not the primary specification, report it; if it was, state the procedure and show that the between-condition gap survives when AI turns are excluded from the uptake graph. This is necessary for the relational-displacement interpretation.
  3. [§4.3, §5.2] §4.3: Interpreting four null temporal probes (all p > .12) as “positive evidence” that the cost is an immediate baseline property overclaims what nulls can support at N ≈ 33 teams. Low power, short sessions (~18 min), and length equalization can all yield non-emergence of a true dynamic. Reframe as failure to detect within-session emergence, report effect-size bounds or equivalence-style intervals where possible, and keep the baseline claim provisional pending longitudinal work already flagged in §5.2–5.3.
  4. [§4.2.2] §4.2.2: Belonging (p = .011) does not survive the stated Bonferroni threshold (α = .0083) while responsivity, social impact, and status do. The text currently packages “four outcomes” with directional consistency; please separate corrected vs uncorrected results clearly in text and figures, and avoid implying uniform statistical support for the full survey bundle.
minor comments (6)
  1. [Figure 1, §3.4] Figure 1 caption says values are “normalized within each dimension,” but the method of normalization (z within full sample vs within treatment only, etc.) is not stated in §3.4–3.5; add one sentence so the mid-range AI positions on responsivity/impact are interpretable.
  2. [§3.1] §3.1: Report total analytic N of students per condition after any exclusions, mean/SD conversation length, and AI turn share distribution; these are needed to contextualize participation and airtime dominance.
  3. [§3.4] §3.4: Belonging and status composites need item counts, response scales, and reliability (α/ω) in this sample; “adapted from Chung et al., 2020” is insufficient for reproduction.
  4. [§3.5] §3.5: GCA embedding model, turn-window size, and similarity threshold are described only as “established practice”; cite the exact parameter set or appendix them.
  5. [§3.4, §4] Gender imbalance (63 F / 12 M; only three males in treatment) is acknowledged; state explicitly in §4 whether primary tables are full-sample or note the female-only sensitivity results in the main text, not only as a planned check.
  6. [§5.3, References] Typos/style: “V oice” appears with a stray space in §5.3; arXiv IDs and “Advance online publication” entries are fine for a preprint but should be cleaned for journal production.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: empirical RCT with independent instruments; construct label does not force the between-condition results.

full rationale

This paper is a between-teams randomized controlled study, not a first-principles derivation. Load-bearing claims (RQ1–RQ3) rest on (i) within-treatment paired GCA contrasts of the AI vs its student teammates, (ii) Welch tests of treatment vs control students on GCA dimensions and survey composites (belonging adapted from Chung et al., 2020; status; felt value), (iii) a within-treatment correlation of AI airtime with felt value, and (iv) null temporal probes (LIWC socioemotional windows, LSM, RQA, modal-stance Markov) arguing the cost is baseline rather than emergent. GCA is an external, previously published pipeline (Dowell et al., 2019), not fitted to the target outcomes. Self-citations (TRAIL platform; Choi/Park/Samadi companion studies) scaffold the broader program and future agenda; they do not supply uniqueness theorems, fitted parameters renamed as predictions, or the numerical between-condition effects. Defining ‘social cost’ as diminished human–human responsivity/belonging is ordinary construct operationalization, not a self-definitional reduction of a claimed derivation to its inputs. Design confounds (group size; AI node in GCA) are validity concerns, not circularity. Score 0; steps empty.

Assumptions & free parameters 6 free parameters · 5 assumptions · 2 invented entities

The claim rests on standard small-group and computational discourse assumptions plus design choices that fix the AI’s behavior and the human comparison. No physical constants; free parameters are experimental/analysis settings. The main invented framing is ‘social cost’ as human-human relational displacement distinct from trust or loafing. Load-bearing domain assumptions include GCA’s interpretation as sociocognitive roles, survey composites as belonging/status, and the adequacy of the 2+AI vs 3-human contrast after per-student normalization.

free parameters (6)
  • Gemini temperature = 0.5
    Generation stochasticity fixed at 0.5; affects AI verbosity and content mix that drive participation/newness profiles.
  • Max tokens per AI generation = 50
    Hard cap of 50 tokens shapes short turns and can mechanically lower density/newness relative to unconstrained students.
  • AI action-management response policy = fixed but not fully disclosed
    Unspecified decision rule for whether the AI speaks each turn; directly influences airtime dominance, the strongest behavioral lever in RQ1/RQ2.
  • GCA turn-window and embedding parameters = not numerically reported
    Said to follow short-form chat practice; sequential-semantic settings affect responsivity, impact, cohesion, newness, density scores.
  • RQA content-state count K and equalized length = K=5; 19 turns
    K=5 k-means states and first-19-student-turn truncation are analysis choices for temporal null tests in RQ3.
  • Bonferroni alpha for six between-condition tests = 0.0083
    Multiple-comparison threshold (alpha=.0083) determines which survey/GCA effects are called significant after correction.
assumptions (5)
  • domain assumption GCA’s six dimensions validly index sociocognitive collaboration roles (participation, internal cohesion, responsivity, social impact, newness, density) in short multiparty chat.
    Invoked throughout §§2.2, 3.4, 4.1–4.2 as the primary discourse evidence for AI role and human disengagement.
  • ad hoc to paper Per-student GCA/survey outcomes plus within-pipeline z-standardization make the 2-human+AI vs 3-human design comparable for human-human relational claims.
    Stated in §3.1 and §3.5 as the fix for group-size and node-count asymmetry; central to interpreting RQ2 as AI presence rather than team composition.
  • domain assumption Belonging and status survey composites, and the valued-member item, measure felt relational integration in the team after a brief chat task.
    §3.4 anchors subjective ‘social cost’ in adapted inclusion/status items administered immediately post-task.
  • ad hoc to paper Null between-condition differences on four temporal probes imply the social cost is an immediate baseline property rather than merely undetected emergence.
    §4.3 explicitly treats temporal nulls (p>.12) as positive evidence for immediacy following Ricca et al.’s emergence-testing stance.
  • standard math Standard inferential tools (Welch t, Wilcoxon signed-rank, Hedges’ g, Pearson r) with the reported corrections suffice for the claimed effects at this N.
    Used in §§3.5–4.2 for RQ1/RQ2 contrasts and the airtime–felt-value association.
invented entities (2)
  • Social cost (human-human relational displacement under AI presence)
    purpose: Name the joint pattern of reduced human-human responsivity/impact and lower belonging/status/felt value as distinct from AI-trust or social loafing.
    Introduced in §§1–2.3 and operationalized in RQ2–RQ3; organizes the paper’s contribution but is defined largely by the measured outcomes themselves.
  • AI teammate as relational displacer
    purpose: Interpretive mechanism: AI occupies bandwidth without prompting human-human elaboration.
    Discussion §5.1 framing of the GCA+survey pattern; not separately measured beyond the primary outcomes.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Social Cost of an AI Teammate: How an Artificial Teammate Reshapes Human-Human Communication in Small-Team Decision-Making." pith.science (2026). https://pith.science/paper/STMVKECK

@misc{pith2026260727179,
  author       = {Pith},
  title        = {Pith review of: The Social Cost of an AI Teammate: How an Artificial Teammate Reshapes Human-Human Communication in Small-Team Decision-Making},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/STMVKECK}},
  note         = {Machine review of arXiv:2607.27179}
}
read the original abstract

Conversational AI is increasingly positioned as a teammate rather than a tool, yet we know little about how its presence reshapes communication among the humans on the team. We examined sociocognitive communication dynamics in team decision-making using Group Communication Analysis (GCA), team surveys, and lexical analyses of team discourse. Teams completed a high-stakes moral-dilemma decision task in a randomized controlled study: 16 teams of two students plus an AI teammate, and 17 all-human teams of three. Across six GCA dimensions and survey outcomes, we find that the AI teammate was the single most talkative and self-cohesive member of every treatment team, yet its contributions carried the least new information and the lowest density. The presence of AI also reshaped communication amongst humans. In AI-human teams, human teammates showed lower responsivity and social impact toward one another and reported lower levels of belonging and status. Greater AI dominance in the conversation was associated with students feeling less valued as team members. Additionally, this social cost is immediate and present at baseline; it does not emerge over the course of the conversation. Drawing on these results, we discuss a research agenda extending to voice-based and longitudinal settings.

Figures

Figures reproduced from arXiv: 2607.27179 by the authors.

Figure 1
Figure 1. GCA profile of the AI teammate compared to treatment students and control students. Values are [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Treatment students versus control students on the six GCA dimensions (Hedges’ g with 95% confidence [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. The felt social cost of AI presence. Panel (a): treatment students report lower belonging and lower status [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

4 extracted references · 1 canonical work pages

  1. [1]

    Bales, R. F. (1950). Interaction process analysis: A method for the study of small groups. Addison-Wesley. Bonito, J. A., & Hollingshead, A. B. (1997). Participation in small groups. In B. R. Burleson (Ed.), Communication Yearbook 20 (pp. 227-261). Sage. https://doi.org/10.1080/23808985.1997.11678943 Boyd, R. L., Ashokkumar, A., Seraj, S., & Pennebaker, J...

    arXiv 1950

  2. [33]

    https://doi.org/10.1007/s10726-026-09990-z O’Neill, T., McNeese, N., Barron, A., & Schelble, B. (2022). Human-autonomy teaming: A review and analysis of the empirical literature. Human Factors, 64(5), 904-938. https://doi.org/10.1177/0018720820960865 Park, S., Shariff, D., Samadi, M. A., Nixon, N., & D’Mello, S. (2025). From discourse to dynamics: Underst...

    arXiv 2022

  3. [178]

    https://doi.org/10.1007/s10462-026-11572-z Zhang, R., Duan, W., Flathmann, C., McNeese, N., Freeman, G., & Williams, A. (2023). Investigating AI teammate communication strategies and their impact in human-AI teams for effective teamwork.Proceedings of the ACM on Human-Computer Interaction,7(CSCW2), Article 281, 1-31. https://doi.org/10.1145/3610072

    work page doi:10.1007/s10462-026-11572-z 2023

  4. [1169]

    Oberhofer, V

    Association for Information Systems. Oberhofer, V . M., Seeber, I., & Waizenegger, L. (2026). Extroversion-introversion design of social robots: The role of the mental model attribution process. Group Decision and Negotiation, 35, Article

    2026

Pith tools

Reviewed July 30, 2026 · model on record in the stance chip above.