Large Language Models (LLMs) such as (Chat)GPT or BERT will become (or have already become) an indispensable part of the writing process in professional settings. They can also be a support for writing tasks in the context of data-to-text automation.
They offer diverse applicability, making them invaluable throughout the copywriting process. These models prove to be versatile solutions, adaptable to a wide range of copywriting tasks.
- In the early planning stages, LLMs emerge as powerful tool for both inspiration and research. Backed by studies showcasing their capacity to boost creativity, these models streamline the ideation process. Engaging with LLM like ChatGPT during this phase fosters the generation of innovative ideas, refines concepts, and allows exploration of various writing styles.
- In the writing phase, LLMs can be instructed to generate content based on specific briefings. Similar to briefing a copywriter, you can specify the topic, style, length, and context, allowing the LLM to produce tailored text.
- In addition, as effective editors, LLM streamline the revision process by correcting errors, improving readability, and enabling a time-efficient workflow.
Because they have such great potential and many of their results sound good and are appealing in terms of language, they raise high expectations for their applicability.
The problem here is not so much minor technical weaknesses, but rather in their basic functionality. They can be disappointing when used professionally, especially with large amounts of text.
To help you set realistic expectations when working with LLMs, here are 5 things to know.
1. LLMs are best when they have no responsibility.
On the one hand, this means that you have to keep the steering wheel in your hand and not give the LLM any unsupervised tasks. At least not major ones.
Secondly, if they are used to produce texts that are critical in some way, for example where liability cannot be ruled out or in the medical field, the result needs to be thoroughly checked at every stage. Such narrow guidance makes LLMs less useful for such texts.
Its strength lies in producing texts that are intended to convey simple information or to be entertaining, and where high levels of accuracy and reliability are not particularly required.
2. LLMs work best with common themes, wording, languages or styles
The more familiar the topic or other text criteria are, the better the quality. The results vary considerably for texts with general or more specialized requirements. This is particularly noticeable for less common languages. However, it can even play a role for different product categories; for example, for well-known fashion products, they can produce significantly better and longer texts than for niche products.
3. LLMs are not experts
They have neither the factual knowledge (which they are not able to look up) nor the experience in the respective areas.
This is not as noticeable in the well-known, general areas as it is in the niche areas, but even there the level of knowledge is not at the level of an expert. So, you need to bring your own expertise to bear on the text, and you need to be careful about how the prompt is worded.
It remains your responsibility to check the facts and to verify all judgments, evaluations, and statements made.
4.Their strength is language
Their mastery lies in language: they can express complex issues in simple terms and switch between different styles. They neither produce run-on sentences (unless desired) nor string nouns together to form abstract statements. Also they make almost no grammatical or spelling mistakes and know an incredible number of variants for every phrase.
You do not need to check the LLMs for linguistic problems, except in a few isolated cases, and you can use them as critical editors.
5. You need to get familiar with the LLM you use
As with all tools, the more you practice, the better the results and the easier the work. Each of the LLMs has its own special features and differences that you need to be aware of. Writing with LLM sis also more different from conventional writing than you may initially think. You will write less of the actual text, but you will set out the features and requirements much more clearly, and you will spend more time skimming suggested texts and editing finished texts.
