What is Chain-of-Thought prompting, and can it be used to generate text?
This article was co-authored with Annelie Tinworth, Lead UX Content Designer at Volvo Cars, localisation legend, and word nerd extraordinaire.
You know when you were at school studying Maths, and your teacher would always sternly remind you to “show your working out”? The more you write out all those steps, the easier it becomes to apply them to more complex problems. And that, essentially, is what Chain-of-Thought (CoT) prompting is about.
We’re going to get you up-to-speed on how this technique works, and how it could help give tools like ChatGPT steps to follow that you might take as a writer when writing to produce higher-quality content. Plus, we’ll touch on why it’s a potentially helpful one for any non-writing folk you work with who may still need to produce content.
In another piece, we’ve already looked at conversational chaining. Chain-of-Thought is a different kind of chain. (And yes, we also think terminology in this space is limited.)
Conversational chaining refers to iterating on one idea within one conversation within ChatGPT (or any conversational AI interface).
Chain-of-Thought is about splitting down a complex problem into its intermediate steps to illustrate how you might logically go about resolving…
