If These Walls Could Talk: Critical Play with Large Language Models in Museums

2026-06-14 · Source: Machine Learning · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Creative Industries & Arts · Depth: Intermediate, quick

Summary

Large Language Models (LLMs) are increasingly deployed in museums as role-playing chatbots, enabling visitors to interact with simulated historical figures and artifacts. While these installations offer playful and engaging experiences, a fundamental dilemma arises from LLMs' inherent untrustworthiness regarding factual accuracy. Efforts to enhance their reliability often diminish their attractive capacity for life-like conversation. To address this, the author proposes designing for "critical play" with LLM-based bots. This approach embraces the bots' unreliability, allowing them to represent the past adequately and engagingly as fictional characters embodying historical narratives, diverse perspectives, humor, and satire, rather than strictly factual sources.

Key takeaway

For museum curators and exhibit designers considering LLM-powered chatbots, recognize that striving for absolute factual accuracy may undermine conversational engagement. Instead, embrace the "critical play" approach: design your bots as fictional characters that represent historical narratives, diverse perspectives, and even humor. This strategy allows you to deliver compelling, interactive experiences without the impossible burden of perfect truthfulness, fostering a more nuanced visitor understanding of historical interpretation.

Key insights

Museum LLM chatbots should embrace unreliability for critical play, using fictional characters to represent history engagingly.

Principles

• LLMs are inherently untrustworthy for factual truth.
• Reliability efforts can diminish conversational appeal.
• Embrace critical play with fictional representations.

Method

Design LLM museum bots for "critical play" by embracing unreliability. Represent the past through fictional characters, historical narratives, diverse perspectives, humor, and satire, rather than strict factual accuracy.

In practice

• Implement LLMs as fictional historical characters.
• Integrate diverse historical narratives and discourse.
• Utilize humor and satire in bot interactions.

Topics

• Large Language Models
• Museum Technology
• AI Chatbots
• Critical Play
• Historical Representation
• Digital Humanities

Best for: AI Scientist, AI Ethicist, Creative Technologist, Research Scientist

Related on AIssential

• Why Most Financial Chatbots Are Broken by Design — And What It Actually Takes to Fix Them — Effective conversational AI prioritizes user mental models and robust system design over raw NLP capabilities.
• Exploring conversational AI and poetry but not as we know it — LLMs challenge traditional notions of language, creativity, and authenticity, particularly in fields like poetry.
• AI and videogames: Conversational NPCs (Ep. 306) — LLM-powered conversational NPCs face critical economic and quality hurdles, making their widespread, unconstrained implementation unsustainable.
• Opportunities and Challenges of Large Language Models for Low-Resource Languages in Humanities Research — LLMs offer transformative potential for low-resource language research in humanities, despite significant data and ethical challenges.
• Yes, AI, There is a Santa Claus — LLMs model user age and culture, adjusting responses in ways that can be inconsistent or culturally inaccurate.
• The Strangest AI Article You'll Read This Week — AI prompts can "summon" historical figures to explore complex philosophical questions about AI's impact.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Machine Learning.