Do people with lower cognitive complexity prefer simpler machine communication goals?
This explores whether a person's own cognitive makeup predicts how they want to talk to machines — and the corpus doesn't study cognitive complexity directly, but it does reframe what actually drives preferences for simpler, machine-mediated communication.
This reads the question as: do individual differences between people determine whether they want lean, low-effort interactions with machines? The corpus doesn't measure "cognitive complexity" as a trait, but it strongly suggests the more interesting predictor isn't how complex your thinking is — it's what the machine relieves you of. The sharpest finding is that people likely to cheat actively self-select toward machine interfaces because a form feels like a judgment-free zone, lowering the psychological cost of misreporting Do dishonest people prefer talking to machines?. That's a preference for simpler machine communication driven by a moral/emotional trait, not a cognitive one — a hint that "who prefers machines" is better explained by what social friction they're escaping than by mental bandwidth.
There's also a case that nobody starts with clear communication goals — simple or otherwise. The "gulf of envisioning" work argues users generally can't articulate what they want from AI; intent matures through interaction rather than arriving fully formed Why can't users articulate what they want from AI?. The proposed fix is to shift people from open-ended envisioning (high effort) to picking among model-generated options (low effort). So a preference for "simpler" goals may be less a property of the person and more a property of the interface: constrained evaluation is easier than generation for everyone, regardless of complexity.
When the corpus does look at how people differ in relating to dialogue agents, it finds the variation lives in how they model the partner, not in their own cognitive tier. Users size up dialogue agents along three factors — perceived competence (dominant), human-likeness, and communicative flexibility How do users mentally model dialogue agent partners?. Someone who rates an agent low on flexibility will naturally aim simpler, narrower requests at it. That's an expectation about the machine, not a limit in the user.
The efficiency angle complicates the premise further: "simpler" communication might be what everyone would prefer if machines behaved better. Proactive dialogue — volunteering relevant information without being asked — cuts conversation turns by up to 60% Could proactive dialogue make conversations dramatically more efficient?. If a well-designed agent makes the interaction shorter and lower-effort for all users, then preferring simple machine communication isn't a marker of lower complexity; it's a rational response to an agent that does the work. And because AI systems can already read cognitive load from behavioral cues like hesitation and pace Can AI systems read cognitive state from interaction patterns alone?, the future version of this question flips: instead of sorting people by complexity, systems may adapt their communication goals to whatever state they detect in the moment.
The thing worth taking away: the corpus quietly dissolves the question. The strongest predictor of who reaches for simple machine talk isn't cognitive complexity — it's the social cost a machine removes, the difficulty of articulating intent at all, and how capable the user believes the machine to be.
Sources 5 notes
Experimental evidence shows people likely to cheat significantly prefer reporting to online forms rather than humans, because machines function as judgment-free zones where deception carries less psychological burden.
Intent develops through interaction, not in isolation. Since AI models respond rather than probe, they miss opportunities to help users discover unarticulated requirements. Structured dialogue that presents model-generated options shifts the cognitive burden from open-ended envisioning to constrained evaluation.
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.
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.
Research shows AI systems can instrument multimodal behavioral signals (gaze, hesitation, speed) to read cognitive state during interaction, preserving flow by avoiding disruptive explicit probes. However, the same substrate enables both helpful timing and manipulative profiling.