Assessing the nature of large language models: A caution against anthropocentrism

TL;DR

This study evaluates large language models like GPT-3.5 using cognitive and personality tests, revealing they lack sentience but exhibit human-like variability and signs of poor mental health, challenging anthropocentric assumptions.

Contribution

The paper introduces a novel battery of tests to assess LLMs' capabilities, stability, and human comparability, highlighting their limitations and psychological-like traits.

Findings

LLMs unlikely to have developed sentience

GPT-3.5 shows high variability in responses

LLMs exhibit poor mental health indicators

Abstract

Generative AI models garnered a large amount of public attention and speculation with the release of OpenAIs chatbot, ChatGPT. At least two opinion camps exist: one excited about possibilities these models offer for fundamental changes to human tasks, and another highly concerned about power these models seem to have. To address these concerns, we assessed several LLMs, primarily GPT 3.5, using standard, normed, and validated cognitive and personality measures. For this seedling project, we developed a battery of tests that allowed us to estimate the boundaries of some of these models capabilities, how stable those capabilities are over a short period of time, and how they compare to humans. Our results indicate that LLMs are unlikely to have developed sentience, although its ability to respond to personality inventories is interesting. GPT3.5 did display large variability in both cognitive and personality measures over repeated observations, which is not expected if it had a human-like personality. Variability notwithstanding, LLMs display what in a human would be considered poor mental health, including low self-esteem, marked dissociation from reality, and in some cases narcissism and psychopathy, despite upbeat and helpful responses.