The Three Ages of Data Science: When to Use Traditional Machine Learning, Deep Learning, or an LLM (Explained with One Example)

The evolution of data science is framed across three ages: traditional machine learning, deep learning, and large language models (LLMs). A practical use case of tweet sentiment analysis is used to compare these different approaches hands-on. The author argues that older methods remain valuable and the best model choice depends on balancing accuracy, cost, and scale. The article then begins a Python implementation for an "easy" classification scenario using traditional ML techniques like TF-IDF and linear models. This sets the stage for tackling more complex versions of the problem with more advanced models.