Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Language Models as Critical Thinking Tools: A Case Study of Philosophers

Andre Ye, Jared Moore, Rose Novick, Amy X. Zhang

TL;DR

This paper investigates whether language models can function as tools for critical thinking, using philosophy as a case study. It reports on 21 professional philosophers' views and experiences with LMs, including live GPT-4 interactions, and qualitatively analyzes the responses. The authors identify two core deficiencies of current LMs—low selfhood and low initiative—that undermine substantive critical thinking, and they propose the Selfhood-Initiative model to formalize this gap. They then outline three roles for LMs (Interlocutor, Monitor, Respondent) and discuss implications for metaphilosophy and interface design, arguing for living-script-inspired, human-centered collaboration between philosophers and AI.

Abstract

Current work in language models (LMs) helps us speed up or even skip thinking by accelerating and automating cognitive work. But can LMs help us with critical thinking -- thinking in deeper, more reflective ways which challenge assumptions, clarify ideas, and engineer new concepts? We treat philosophy as a case study in critical thinking, and interview 21 professional philosophers about how they engage in critical thinking and on their experiences with LMs. We find that philosophers do not find LMs to be useful because they lack a sense of selfhood (memory, beliefs, consistency) and initiative (curiosity, proactivity). We propose the selfhood-initiative model for critical thinking tools to characterize this gap. Using the model, we formulate three roles LMs could play as critical thinking tools: the Interlocutor, the Monitor, and the Respondent. We hope that our work inspires LM researchers to further develop LMs as critical thinking tools and philosophers and other 'critical thinkers' to imagine intellectually substantive uses of LMs.

Language Models as Critical Thinking Tools: A Case Study of Philosophers

TL;DR

This paper investigates whether language models can function as tools for critical thinking, using philosophy as a case study. It reports on 21 professional philosophers' views and experiences with LMs, including live GPT-4 interactions, and qualitatively analyzes the responses. The authors identify two core deficiencies of current LMs—low selfhood and low initiative—that undermine substantive critical thinking, and they propose the Selfhood-Initiative model to formalize this gap. They then outline three roles for LMs (Interlocutor, Monitor, Respondent) and discuss implications for metaphilosophy and interface design, arguing for living-script-inspired, human-centered collaboration between philosophers and AI.

Abstract

Current work in language models (LMs) helps us speed up or even skip thinking by accelerating and automating cognitive work. But can LMs help us with critical thinking -- thinking in deeper, more reflective ways which challenge assumptions, clarify ideas, and engineer new concepts? We treat philosophy as a case study in critical thinking, and interview 21 professional philosophers about how they engage in critical thinking and on their experiences with LMs. We find that philosophers do not find LMs to be useful because they lack a sense of selfhood (memory, beliefs, consistency) and initiative (curiosity, proactivity). We propose the selfhood-initiative model for critical thinking tools to characterize this gap. Using the model, we formulate three roles LMs could play as critical thinking tools: the Interlocutor, the Monitor, and the Respondent. We hope that our work inspires LM researchers to further develop LMs as critical thinking tools and philosophers and other 'critical thinkers' to imagine intellectually substantive uses of LMs.
Paper Structure (20 sections, 1 figure)

This paper contains 20 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Background and Related Work
  3. Language Models as Thinking Tools
  4. Language Models as Critical Thinking Tools
  5. Philosophy as Critical Thinking, Critical Thinking as Philosophy
  6. Methods
  7. Language Models Are Not Good Critical Thinking Tools (So Far)
  8. How do philosophers philosophize?
  9. Language Models are neutral, detached, and nonjudgmental
  10. Language Models are servile, passive, and incurious
  11. Designing Language Models for Critical Thinking
  12. The Selfhood-Initiative Model
  13. Three Roles for Language Models as Critical Thinking Tools
  14. Discussion
  15. Challenges for Language Modeling
  16. ...and 5 more sections

Figures (1)

  • Figure 1: Left -- The selfhood-initiative model for critical thinking tools. Right -- Illustrative sample interactions between humans and LMs playing different roles. Other alternatives are possible. The excerpt from "The Monitor" is taken from 68204752-8786-318f-a155-4726d074c535.