Since generative AI hit the mainstream in late 2022, most newsrooms have been trying to find out how to use OpenGPT and similar tools to (cheaply) create content, improve journalistic work, or simplify or eliminate time-consuming tasks.

Some of the trials went quite badly, especially when ChatGPT was used unsupervised or used as a knowledge tool and not solely as a language tool.

As noted in this article published in MIT Technology Review, large language models (LLM) are “notorious bullshitters,” and their main task is to sound plausible, not to be factual correct. After all, everything the machine spits out is based on probability and what the next “best” word is to sound good — not on any understanding (let alone a deep understanding) of the input given.

The use of ChatGPT as a language tool for journalists and their work in a newsroom can be extremely valuable in many places. However, a human check is always required.

7 concrete applications of ChatGPT for newsrooms

If you have experimented with ChatGPT or similar tools and gained your first experiences, here are seven concrete applications where you can begin working earnestly with these tools. Start honing your prompting skills for your daily newsroom work, and create some “master prompts,” whether you have access to GPT-4 or any other language model, such as Bard or Claude.

1. Copy editing and grammar checking

The most obvious applications are simple copy editing and grammar checking of articles. The results are absolutely usable and of high quality, even with quite simple prompts. If you add instructions, such as following a certain writing style, you can refine and improve the response.

2. Creating and optimising headlines

ChatGPT is already widely used for creating and/or optimising headlines and article teasers. Asking for multiple examples surfaces usable results, sometimes superior to human-generated versions.

Additions to the prompt, such as creating cliff hangers without using questions or describing who this article is intended for, improve the quality of the response. Giving examples of what a good article headline looks like also makes a big difference.

3. Summarising articles

Creating article summaries is another popular application that often works. Even if ChatGPT and LLMs in general can’t count characters or words, terms in the prompt like “short” or “very short” work quite well.

4. Newsletter creation

Using ChatGPT to create newsletters containing teasers or summaries of articles is another simple but useful application. Either the different source articles are part of the prompt (which can be cumbersome if there are lot of long articles) or the article summaries are created beforehand (in the case of creating teasers) and then used in the prompt.

Even instructing ChatGPT to create text transitions between the topics works satisfactorily in some cases. Again, as with all other examples, trial and error improve the results.

5. Combining articles

Another time saver is to combine different articles about a certain topic or connected topics into a new article. The result can offer a concise overview and summary of a series of articles for the reader. Here the border between using a LLM just as a language model or to create new journalistic content gets blurry, so every newsroom needs to decide where the red line is.

6. Target group specific rewriting

This includes taking a current affairs story and rewriting it using simpler language or in a writing style suitable for children or people who are not fluent in the native language of a country.

7. Social media posts

This takes the summary and target group-specific writing to another application by writing posts for Facebook, X, or other social media platforms. As noted, LLM can’t count the exact number of characters, so some experiments are needed. If you try to give a character count and undershoot, you might get good result.

5 fundamental tips for good prompts

In any of these applications, the quality of the prompt determines the quality of the response. Sophisticated “prompt engineering” or “prompt design” is not trivial, and there are job ads offering US$150,000 or more for skilled prompt engineers.

Sometimes the best prompts are longer than the response. A lot of thinking goes into designing prompts. Not every journalist will or has to become a bona fide prompt engineering specialist.

Here are five starter tips for improving the quality of the responses quickly.

1. Be clear and specific

The system does what you tell it to do, not what you want it to do. Ambiguity can lead to unexpected or undesired responses.

2. System prompts

Perhaps the most important part of the prompts is the system prompt. It describes the role a model should assume for the task ahead and can influence the result dramatically.

For instance, starting a prompt with “you are a journalist” followed by asking it to “explain what the sun is” will have a very different result than “you are a philosopher” with the same task.

3. Reduce hallucinations

State in the prompt that the model should only use the data provided in the prompt as a source for the task. It doesn't eliminate the risk of hallucinations completely but reduces it.

4. Provide examples

Providing ChatGPT examples of what you want the result to look like is also very powerful. For instance, if you ask for headline suggestion for an article, provide examples of other headlines that are written in the style you want to receive in the result.

5. Refine prompts

ChatGPT is a conversational tool, so refine the prompts as you go. It remembers conversations and learns from past prompts and responses. Keep asking, and experiment with different phrasings and structures.

These applications and tips are only the first steps into the rabbit hole of using generative AI in a journalistic context. As the systems continuously evolve, so does the science and art of prompt design.

The key factors for every journalist are constant experimenting and learning by doing. Most importantly, ChatGPT and similar tools can help journalists work by automating and streamlining tasks, but it cannot replace human intuition and creativity.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .