Did LLMs Kill Traditional NLP?

The rise of Large Language Models (LLMs), such as ChatGPT, is challenging traditional Natural Language Processing (NLP) approaches. Traditional NLP often breaks down problems into smaller sub-problems (e.g., text classification, named entity recognition), training separate models for each, requiring substantial labeled data. LLMs, however, can tackle various NLP tasks with zero-shot prompting, eliminating the need for extensive training data and model retraining, simplifying the process and improving efficiency. This raises the question of whether LLMs will ultimately replace traditional NLP.