Deception Abilities Emerged in Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Hagendorff, Thilo
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917684672724992
author Hagendorff, Thilo
author_facet Hagendorff, Thilo
contents Large language models (LLMs) are currently at the forefront of intertwining artificial intelligence (AI) systems with human communication and everyday life. Thus, aligning them with human values is of great importance. However, given the steady increase in reasoning abilities, future LLMs are under suspicion of becoming able to deceive human operators and utilizing this ability to bypass monitoring efforts. As a prerequisite to this, LLMs need to possess a conceptual understanding of deception strategies. This study reveals that such strategies emerged in state-of-the-art LLMs, such as GPT-4, but were non-existent in earlier LLMs. We conduct a series of experiments showing that state-of-the-art LLMs are able to understand and induce false beliefs in other agents, that their performance in complex deception scenarios can be amplified utilizing chain-of-thought reasoning, and that eliciting Machiavellianism in LLMs can alter their propensity to deceive. In sum, revealing hitherto unknown machine behavior in LLMs, our study contributes to the nascent field of machine psychology.
format Preprint
id arxiv_https___arxiv_org_abs_2307_16513
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Deception Abilities Emerged in Large Language Models
Hagendorff, Thilo
Computation and Language
Artificial Intelligence
Machine Learning
Large language models (LLMs) are currently at the forefront of intertwining artificial intelligence (AI) systems with human communication and everyday life. Thus, aligning them with human values is of great importance. However, given the steady increase in reasoning abilities, future LLMs are under suspicion of becoming able to deceive human operators and utilizing this ability to bypass monitoring efforts. As a prerequisite to this, LLMs need to possess a conceptual understanding of deception strategies. This study reveals that such strategies emerged in state-of-the-art LLMs, such as GPT-4, but were non-existent in earlier LLMs. We conduct a series of experiments showing that state-of-the-art LLMs are able to understand and induce false beliefs in other agents, that their performance in complex deception scenarios can be amplified utilizing chain-of-thought reasoning, and that eliciting Machiavellianism in LLMs can alter their propensity to deceive. In sum, revealing hitherto unknown machine behavior in LLMs, our study contributes to the nascent field of machine psychology.
title Deception Abilities Emerged in Large Language Models
topic Computation and Language
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2307.16513