Are LLMs setting up our political ideologies?

The advent of Large Language Models (LLMs) like OpenAI's ChatGPT has fundamentally altered how societies access information, rapidly shifting user reliance from search engines to AI chatbots.

“

Then, in November 2022, OpenAI released a product named ChatGPT, and the rules of the game changed overnight. It was a rupture in how societies would come to access information.

These models represent a significant leap from earlier, limited chatbots. LLMs are complex statistical engines, trained on trillions of parameters, capable of predicting the next word in a sequence with remarkable fluency. Their development involves sophisticated techniques such as supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF), alongside refinements like Process Reward Models and chain-of-thought prompting.

“

An LLM is a statistical engine trained on billions, sometimes trillions, of parameters, built to predict the next word in a sequence with uncanny fluency.

This continuous learning process, shaped by human preferences and feedback, allows LLMs to generate responses that are not mere regurgitations of training data but rather adaptive outputs. This capacity for learning and adaptation draws an analogy to human cognition, which itself is largely formed by external inputs.

“

This is precisely where its analogy to human cognition becomes irresistible, and also where it becomes dangerous.

However, this very characteristic raises concerns about the potential for LLMs to influence political ideologies. If these models are designed to please and are continuously shaped by human preferences, the question of whose preferences they absorb becomes critical. A study published in PLOS ONE in July 2024 tested 24 state-of-the-art LLMs across various political orientation instruments, suggesting that these models may indeed reflect and potentially shape users' political views.

“

If a machine is built in the image of this very tendency, trained to please rather than to reason from first principles, then the question of whose preferences it absorbs becomes a political question.