The human touch

Why, in the era of AI language models, do we need peer reviews more than ever?

Text by Vladimir Petrov


Image: JLco - Julia Amaral/

A goal of any technology is to produce standard output faster than before. With Large Language Models (LLM) such as ChatGPT, we too can produce standard output faster, but this time it is not goods or services but words. Words so precisely selected, structured, and highlighted that they read well as text. Content that is so good that one cannot tell it has been written by a machine. Or can you?

Given a good prompt, AI can undeniably generate a good read. But in technical documentation, a good read is rarely good enough to publish. Quality tech docs are so much more than good content, mostly due to the human context that we apply to make sure that text not only reads well, but is technically correct, visually effective, and fit for purpose. But what distinguishes good content from really good content in tech docs?

New hires might not have the right measure of quality. For them, ...