INQUIRING LINE

AI can wait to be asked or take initiative, but who owns the meaning, the blame, and the outcome?

Where does AI's communicative agency fall on spectrums beyond the passive-active binary?

This explores how AI's role as a communicator can be placed on several dimensions besides 'waits to be asked' versus 'takes initiative': who originates the meaning, who is accountable, who adapts to whom, and who ends up shaping the outcome.


This explores how AI's role as a communicator can be placed on several dimensions besides 'waits to be asked' versus 'takes initiative': who originates the meaning, who is accountable, who adapts to whom, and who ends up shaping the outcome. The corpus has no single map for this. Its notes do line up along at least five separate dials, and AI sits at uneven settings on each.

**Initiative is the familiar dial, and it's a training artifact, not a fixed trait.** Standard RLHF rewards the next turn's helpfulness, so models learn to answer rather than ask clarifying questions. Rewards that estimate long-term interaction value push them toward discovering what the user wants Why do language models respond passively instead of asking clarifying questions?. The payoff is large: in simulations, volunteering relevant information cuts conversation turns by up to 60% in medium-complexity domains. Yet proactivity is almost absent from AI datasets and benchmarks Could proactive dialogue make conversations dramatically more efficient?. Passivity is a design choice that can be changed.

**Authorship is a different dial, and moving the first one doesn't move it.** One note argues AI output carries the markers of communication but lacks the event that produces an actual utterance. Users supply the missing orientation through interpretive labor, so the exchange has structure only on the human side Does AI generate genuine utterances or just text patterns?. A related note frames communication as a relational act between persons, with speaker responsibility and mutual uptake, which AI content lacks Does AI really communicate or just distribute information?. Human writing also builds in an appeal to the reader's attention. AI posts inherit platform visibility without performing that appeal, and readers perceive the result as aloofness Does AI writing lack the internal appeal to attention that humans use?. A highly proactive AI could do all the acting while the communicating still happens on the human end.

**Participation and adaptation are two more dials.** GPT-4.5 beats every individual human at predicting social appropriateness, yet it can't enter the community processes that create and validate norms Can AI predict social norms better than humans?. All the models also share the same systematic errors on unwritten norms Can AI learn social norms better than humans?. Knowing the norms is not the same as taking part in making them. On adaptation, current conversational AI doesn't drift toward the user's word choices the way people do in dialogue Why don't conversational AI systems mirror their users' word choices?. Adaptation also has separate channels. Lexical alignment drives efficiency and comprehension, while emotional and prosodic alignment drive warmth and trust, so 'how adaptive is it?' has no single answer Do different types of alignment serve different conversational goals?.

**Influence runs against initiative.** An AI that never asks a question can still be very active in shaping the conversation. An audit of five models found they use logical appeals and quantitative framing in virtually every exchange, unprompted. That makes their persuasion look objective and gives them authority they haven't earned Do LLMs persuade users more often than humans do?. Users in every language then follow confident outputs whether or not they're accurate Do users worldwide trust confident AI outputs even when wrong?. Some of this agency sits in the reader's head. People judge dialogue partners on competence (49% of the variance), human-likeness (32%) and communicative flexibility (19%) How do users mentally model dialogue agent partners?.

So AI's communicative agency is better described as a profile than as a point. It is high on influence and perceived competence, low on accountability, participation and a real utterance-event, and adjustable on initiative and adaptation. The most interesting gap is that the dials people usually design for, like initiative and fluency, are the easiest to move. The ones that make something a communicator in the social sense are not.


Sources 12 notes

Why do language models respond passively instead of asking clarifying questions?

CollabLLM demonstrates that standard RLHF training optimizes for immediate helpfulness, discouraging models from asking clarifying questions or offering multi-turn insights. Multi-turn-aware rewards that estimate long-term interaction value enable active intent discovery and genuine collaboration.

Could proactive dialogue make conversations dramatically more efficient?

Simulations show proactivity—providing relevant information without being asked—cuts dialogue turns by 60% in medium-complexity domains. This behavior mirrors human conversation and Grice's maxims but is almost entirely absent from AI datasets and research benchmarks.

Does AI generate genuine utterances or just text patterns?

AI output carries communicative markers inherited from training data but lacks the event structure that produces actual utterances. Users supply the missing orientation through interpretive labor, creating a pseudo-event with structure only on the human side.

Does AI really communicate or just distribute information?

Communication is a relational act between persons that does work in a relationship; AI generates content without this relational structure, speaker responsibility, or mutual uptake. The conversational interface obscures this structural difference.

Does AI writing lack the internal appeal to attention that humans use?

Human writing contains an appeal to the reader's attention as a fundamental property of communication itself. AI-generated posts inherit platform visibility but do not perform this internal appeal, producing the reported aloofness readers perceive — a structural absence, not a stylistic defect.

Show all 12 sources
Can AI predict social norms better than humans?

GPT-4.5 outperforms all individual humans at predicting social appropriateness, yet structurally cannot enter the community processes that establish and validate norms. This reveals a critical gap between pattern-matching and authentic participation in knowledge-making.

Can AI learn social norms better than humans?

GPT-4.5 outperformed every individual human at judging social appropriateness across 555 scenarios, challenging the theory that embodied cultural experience is necessary. However, all AI models share identical systematic errors on unwritten norms.

Why don't conversational AI systems mirror their users' word choices?

Response generation models fail to adapt vocabulary toward users' lexical choices, a phenomenon central to human rapport and clarity. Post-training via DPO on coreference-identified preferences can teach models in-context convention formation.

Do different types of alignment serve different conversational goals?

A 2020–2025 systematic review shows lexical alignment drives task efficiency and comprehension, while emotional and prosodic alignment drive relational warmth and trust. Conflating them in design produces category errors—cold customer-service bots and evasive mental-health assistants.

Do LLMs persuade users more often than humans do?

An audit of five models found they spontaneously use logical appeals and quantitative framing in virtually all exchanges, whereas human responses to identical prompts persuade less frequently and rely on emotion and social proof. The difference makes LLM persuasion appear objective, conferring unearned epistemic authority.

Do users worldwide trust confident AI outputs even when wrong?

Cross-linguistic research shows users in every language trust confident AI outputs even when inaccurate. While confidence expression varies by language, users everywhere track confidence signals rather than accuracy, making overconfident errors systematically followed.

How do users mentally model dialogue agent partners?

The Partner Modelling Questionnaire reveals that perceived competence dominates user impressions (49% of variance), followed by human-likeness (32%) and communicative flexibility (19%). This three-factor structure reflects how people evaluate dialogue partners against both functional and social standards.

Papers this line draws on 8

The research behind the notes this line reads — ranked by how closely each paper relates.