Compling — An Introduction

Interdiscplinary

In my opinion, anything “computational” is understood as complex problem-solving. How can we go from step A to B? Can we find an easier version of this problem? Where do we see connections in what we’ve learned and what the problem asks?

Likewise, linguistics is the scientific study of language. By combining both fields together, what we end up with is rather intriguing — the study of language’s structure through computer science and statistical algorithms. In some definitions, computational linguistics is synonymous with natural language processing processing (NLP), which focuses on how computers understand and generate human language.

Display of alphabet letters in various fonts on small cards arranged in a grid behind a metal mesh

Relevant

We are entering, as Robert Downey Jr. put it, the age of AI. Thousands of AI tools have been created after ChatGPT, all of which use compling and NLP for its barest function — communication.

Yet these large language models (LLMs, another word for machines like ChatGPT) don’t actually understand human language. They can’t actually produce completely original text, or be in tune to the little nuances of language the way a human can naturally. Instead, developers train LLMs by feeding them mass amounts of data, from publications to literary analyses to images of dogs.

AI is taking over the world, and so compling follows.

A life-sized robot constructed from construction equipment and machinery, displayed indoors at an exhibition with red panels and 'SANY' branding.

Compling Resources

  • NLP vs. theoretical linguistics, learning the linguistic fundamentals for CL, python, semantics, all the cool stuff, where to get started, how to progress

  • Some comic short stories I’ve drawn, informative articles connected to the syllabus

  • Helpful resources outside of my blog that I’ve found useful from personal experience

  • Some practice problems and ways to get involved in the world of compling/CL!