How might you make use of Generative AI?

Generative AI tools allow you to create new content (text or images) from short or complex prompts. This can be helpful as you are drafting new content, as long as you check the output carefully. Here are some examples: 

• Idea Generation: Generative AI can help writers with brainstorming initial ideas and generating creative concepts. By inputting prompts or keywords, the AI model can generate a range of potential angles, headlines, or content ideas that can serve as a starting point for the copywriter's work. 

• Content Expansion: writers can use generative AI to expand their content. For instance, if a copywriter has written a few paragraphs or sentences, they can input that text into the AI model to generate additional related content. This can help in expanding the scope of an article, blog post, or marketing copy. 

• Editing and Refinement: Generative AI can help writers during the editing and refinement process. By inputting draft copy, the AI model can suggest alternative sentence structures, vocabulary choices, or improvements to enhance the clarity and impact of the writing. This can help writers refine their work more efficiently. 

• A/B Testing: writers can use generative AI to generate multiple versions of advert copy, email subject lines, or social media posts. By running A/B tests with different generated variations, writers can quickly identify which versions perform better in terms of engagement, click-through rates, or conversions. This makes it easier to test different approaches in a data-driven way. 

• Language and Tone Adaptation: Generative AI can assist writers in adapting their content to specific language styles, target audiences, or brand guidelines. By refining the AI response with successive prompts, copywriters can generate text that aligns with the desired tone, formality, or cultural nuances. This helps ensure consistent and appropriate messaging across different contexts. 

It’s worth emphasising that you’ll still need to apply your expertise and experience to ensure that you supply the right prompts and context at the beginning of the process. And you’ll need to apply your critical judgement to check the result, especially if you are developing informational content to help people make decisions. 

Coming soon – Retrieval-based AI tools 

A current weakness of generative AI is that it is, quite literally, making things up. Based on a very refined language model and a fair bit of computing power, generative AI essentially guesses what would be the most plausible response and outputs the result. Because language patterns are predictable, this next best guess approach is surpisingly effective. However, generative AI tools are not usually able to make real-world fact checks at the moment. Some companies such as Google and Microsoft are linking generative AI tools to recent web content in an effort to improve their accuracy. 

The next step will be retrieval-based AI tools. These will work in a similar way to generative AI, but they will match and check results against a pre-defined data set. So, for example, you could instruct a retrieval-based AI tool to check against a library of published guidance when producing answers to queries. This should radically improve the reliability of these tools for people seeking factual information.