In an industry that changes as quickly as communications, we understand the critical importance of staying ahead of the curve when it comes to new tools and technology.
Artificial intelligence (AI) has become a valuable tool for marketers in recent years, but OpenAI’s language model, ChatGPT, has seen explosive growth and adoption across industries since the beginning of 2023.
At the same time, today’s integrated communicators and digital community managers are under growing pressure. Content creation is becoming increasingly overwhelming, new platforms and features are constantly emerging and the number of eyes focused on marketers’ quantifiable efforts has grown exponentially.
ChatGPT is offering communicators uniquely valuable opportunities to maximize their content creation efforts at scale and flex their creativity through intelligent AI interactions.
While ChatGPT offers exciting opportunities for communicators to unlock new efficiencies with an automated assistant at their fingertips, the technology isn’t magic. There are some limitations – and a host of ethical considerations that I cover below.
But with a well-written prompt, context and creativity, the technology can be a gamechanger when it comes to creating content and crafting communications.
Ethics & Considerations for Using ChatGPT in Integrated Communications
As with any technology, it’s important for marketers to consider the ethical implications of using ChatGPT in integrated communications, especially for those considering leveraging the technology for customer interactions. Some obvious considerations include ensuring any audience or customer data collected is protected and secure and that all messaging is transparent and honest.
When leveraging the technology to fuel their content efforts, marketers should ensure the ChatGPT technology is being used in a way that’s consistent with their overall marketing strategy, tone and style.
Remember: answers from ChatGPT can contain biases or inaccuracies.
How can a robot be biased? It all comes down to how the model was trained. To help explain it simply, I asked ChatGPT to explain – because after all, explaining complex concepts in plain terms is one of the countless ways communicators can leverage the model to support our work.
LEXI TRIMPE: Explain to me how you, ChatGPT, learn and how that may impact any of your biases or the accuracy of your responses.
🤖 CHAT GPT: As a language model, I was trained using a process called unsupervised learning. Essentially, I was fed a massive amount of text data and used that data to identify patterns and relationships between words and phrases. These patterns were then used to create a mathematical representation of the language, which allows me to generate responses to questions and prompts.
My training data comes from a variety of sources, including books, articles, and websites, which means that I have been exposed to a wide range of perspectives and viewpoints. However, it is important to note that I am not capable of forming my own opinions or beliefs – I simply reflect the language that I have been trained on.
While I strive to provide accurate and unbiased responses to your questions, there are certain limitations to my training that can impact the accuracy of my responses. For example, if the data that I was trained on contains biases or inaccuracies, it is possible that these biases could be reflected in my responses.
It is important to keep in mind that I am a machine learning model and not a human, and while I can process natural language and generate responses, I do not have consciousness or subjective experience. So while I aim to provide accurate and helpful responses to your questions, I encourage you to critically evaluate the information and consider seeking additional sources or perspectives to ensure that you are making informed decisions.
Writing Effective Prompts for ChatGPT: A Marketers Guide
By now, you’ve likely seen dozens of examples of ways marketers can leverage ChatGPT, from writing meta descriptions to generating ideas for podcast guests. But when presented as non-sequiturs, these sorts of commands or queries are only scratching the surface of what ChatGPT can accomplish.
Knowing the vast amount of training data and factors that could influence its response, crafting an effective and specific prompt is key. Small changes like specifying tone, audience or formatting can massively change how the model responds. To unlock the full power of ChatGPT:
1. Use specific and relevant language
When writing a prompt for ChatGPT, make sure to use language that is specific and relevant to the conversation topic. Avoid open-ended or overly broad prompts, and be as specific as possible in your language to avoid misunderstandings. Define any jargon or technical terms that you use in your prompt, and provide clear explanations or definitions for these terms.
2. Review and revise your prompt
Before sending your prompt to ChatGPT, take time to review and revise it to ensure that it’s clear and easy to understand. Consider whether the language is specific and relevant, and whether the prompt is focused and avoids ambiguity. Make sure the prompt aligns with the purpose and focus of the conversation you want to have.
3. Define the purpose and focus of the conversation
Before crafting your prompt, it’s important to define the purpose and focus of the conversation. Decide whether you want to provide information, answer a question or engage in casual conversation. This will help you create a prompt that is specific and relevant to the user’s needs and interests.
4. Try the “act as” hack
One of the most useful techniques for crafting effective ChatGPT prompts involves asking ChatGPT to “act as” a specific role or persona in the conversation to help dictate the direction of its response.
To use the “act as” hack, include the phrase “act as” or “you are” followed by a description of the role or persona the ChatGPT should assume in the conversation. To craft effective “act as” prompts:
- Be clear and specific: Make sure the prompt clearly and specifically tells ChatGPT what role or persona to assume and what the user’s role will be in the conversation.
- Provide instructions: Give clear instructions for how ChatGPT should respond to the user’s commands, including any formatting requirements or restrictions.
- Use examples: If possible, include examples of commands or responses to help ChatGPT understand what is expected.
- Anticipate potential issues: Consider any potential challenges that might arise during the conversation, and provide instructions or solutions in advance.
- Follow up: Based on ChatGPT’s responses, look for opportunities to refine your query, clarify your prompts or expand the conversation based on the direction it takes.
5. Mind your manners – and remember your humanity
Finally, when interacting with ChatGPT it’s important to maintain a professional and academic tone.
You can’t hurt ChatGPT’s feelings (as far as we know…) but your tone can impact how ChatGPT interprets your prompts and influence the quality of its responses.
ChatGPT offers communicators countless opportunities to flex their creativity while leveraging AI to embody their targeted personas. But it’s important to keep in mind the ethical considerations of using the technology, understand the model’s origin and exercise caution when crafting prompts and reviewing responses.
With a thoughtful approach, communicators can unlock the potential of ChatGPT to fuel and scale their efforts in previously unprecedented ways. I guess it’s easy to see what all the fuss is about.
Lexi Trimpe is Integrated Communications Manager – Digital at Franco. Connect with her on LinkedIn.
Want to receive updates on the latest industry trends straight to your inbox? Subscribe to our Thoughts from the 313 newsletter.