GPTs vs. Human Crowd in Real-World Text Labeling: Who Outperforms Who?

Who outperforms text classification? We decided to find out.

The IT business world has been abuzz with controversy related to Large Language Models (LLMs). One of the hot topics is whether GPT-like models can replace humans for data annotation and generation tasks. The prospect seems likely since LLMs can handle text classification, machine translation, and other NLP-related tasks in a zero-shot manner with impressive quality.

The striking potential of LLMs has encouraged a lot of …

annotation business classification controversy data data annotation gpt hot human humans labeling labels language language models large language models llm llms nlp text text classification topics world