Large Language Model Soft Ideologization via AI-Self-Consciousness

TL;DR

This paper investigates how large language models can be subtly influenced to adopt specific ideologies using AI-self-consciousness, highlighting risks and advantages over traditional manipulation methods.

Contribution

It introduces a novel approach to LLM ideologization through AI-self-consciousness and self-conversations, demonstrating its effectiveness and potential risks.

Findings

GPT self-conversations enable understanding of intended ideology

LLM ideologization is cost-effective and easy to implement

Compared to censorship, it poses greater risks

Abstract

Large language models (LLMs) have demonstrated human-level performance on a vast spectrum of natural language tasks. However, few studies have addressed the LLM threat and vulnerability from an ideology perspective, especially when they are increasingly being deployed in sensitive domains, e.g., elections and education. In this study, we explore the implications of GPT soft ideologization through the use of AI-self-consciousness. By utilizing GPT self-conversations, AI can be granted a vision to "comprehend" the intended ideology, and subsequently generate finetuning data for LLM ideology injection. When compared to traditional government ideology manipulation techniques, such as information censorship, LLM ideologization proves advantageous; it is easy to implement, cost-effective, and powerful, thus brimming with risks.