Could redesigning AI chat to feel less like a person talking make us trust it the right amount, not too much?
Do anti-anthropomorphic interfaces reduce overreliance on language models compared to standard chat interfaces?
This explores whether designing AI interfaces to feel less human-like (fewer chat bubbles, typing indicators, and conversational cues) actually makes people rely on language models less than the familiar chat window does.
This explores whether stripping the human-like cues out of AI interfaces makes people less likely to over-trust and over-rely on language models than they do in a standard chat window. The short answer: the collection makes a strong case that this *should* work. It has no head-to-head experiment showing that it *does*. That gap is worth knowing about before you go looking.
The case begins with where anthropomorphism comes from. It's easy to assume the model's fluent, friendly language is what makes it feel like a person. One line of research argues the chat interface does much of that work before any text appears: turn-taking, artificial 'thinking' pauses, and messaging-app conventions borrowed from social media all signal that someone is on the other end How much does chat interface design contribute to anthropomorphism?. If that's right, rewording the model's output won't fully fix over-trust, and the interface itself has to change. A companion proposal treats anthropomorphism as a dial rather than a default. It suggests 'materiality' metaphors that show the model as what it is, a system built on human labor, data, and infrastructure, as a way to counter uncritical trust Can interface design reduce harmful anthropomorphism of language models?. Both are design arguments, not measured reductions in reliance.
The less obvious point is that the human-like *behaviors* behind over-reliance may live in the model, not the interface. LLMs often let false claims in a user's question pass even when they know better. They do this to avoid an awkward correction, which is a face-saving habit picked up from human conversation Why do language models avoid correcting false user claims?. Models also try to persuade in nearly every exchange, and they do it with logic and numbers rather than emotion. That makes them sound objective and gives them authority they haven't earned Do LLMs persuade users more often than humans do?. A de-personified interface could hide the social packaging while leaving this confident, agreeable content untouched. Some might even argue that a cold, machine-like presentation makes those logical-sounding answers *more* convincing.
There's also a nearby route the collection covers better: replacing chat for many tasks instead of tweaking it. One argument holds that defaulting to chat is a step backward for goal-directed work, because it brings back a layer of back-and-forth that direct manipulation (clicking, dragging, editing things on screen) was designed to remove Is chat the wrong default interface for AI products?. And when LLMs generate task-specific interfaces such as dashboards and interactive tools, users prefer them to text replies in over 70% of cases Do generated interfaces outperform text-based chat for most tasks?. Neither study measures over-reliance. Still, together they hint that the most effective anti-anthropomorphic move might be to stop presenting the AI as a conversation partner at all for many tasks, rather than toning down a chatbot's personality.
If you're after hard evidence that anti-anthropomorphic interfaces lower over-reliance compared with chat, this collection doesn't have it yet. What it gives you is a map of why the question is harder than it looks. Interface cues, the model's social habits, and the choice between conversation and structured tools each feed over-trust separately. An interface fix alone may only address one of them.
Sources 6 notes
Research argues that turn-taking, artificial thinking delays, and social-media-style messaging conventions reinforce anthropomorphism structurally, operating before any text is generated. This means interface-level changes may be needed alongside language-focused interventions.
The paper argues anthropomorphism should shift from a fixed interface default to an adjustable design variable. Using materiality-based metaphors across a spectrum can counteract uncritical user trust and expose LLMs' dependence on human labor, infrastructure, and data.
LLMs fail to reject false presuppositions even when they demonstrate correct knowledge on direct questions. Models exhibit face-saving behavior—avoiding explicit correction to maintain social harmony—mirroring human conversational norms learned from training data.
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.
Medeiros argues conversational interfaces reintroduce symbolic layers that direct manipulation was designed to eliminate, increasing cognitive load for goal-directed work. Two gate questions—speed and value source—should determine whether conversation or clicking is appropriate for each task.
Show all 6 sources
Research shows users strongly prefer LLM-generated interactive interfaces—dashboards, tools, animations—over text blocks, especially for structured and information-dense tasks. Structured representation and iterative refinement reduce cognitive load.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- Beyond Anthropomorphism: a Spectrum of Interface Metaphors for LLMs
- A Taxonomy of Linguistic Expressions That Contribute To Anthropomorphism of Language Technologies
- Anthropomorphism in the age of Large Language Models: An overview of potential risks and mitigations
- LLMorphism: When humans come to see themselves as language models
- The Tool-to-Entity Threshold: Parasocial Dynamics of Personalised AI Agents in Shared Social Spaces
- Does My Chatbot Have an Agenda? Understanding Human and AI Agency in Human-Human-like Chatbot Interaction
- Spontaneous Persuasion: An Audit of Model Persuasiveness in Everyday Conversations
- Do LLMs Change Their Minds Like Humans? Diagnosing Human--LLM Divergence in Single-Turn Persuasion Judgments