Artificial intelligence (AI) has revolutionized the field of communication in recent years, and has gone from being an optional tool to a must-have resource if you don’t want to be left behind. But I’m sure you already know that.
Two years ago, we shared some basic recommendations to encourage communication professionals to experiment with AI and start using it in their tasks. Now it’s time to take it a step further and integrate it into routines and work processes consistently, effectively and ethically.
To help you, I share five learnings I wish I had known when I first started using generative AI. I have been using AI intensively for two years now. Every day I use some, if not more, generative AI tool to perform communication tasks: from copywriting and analysis to strategy planning for my clients and for #TheImperdibles school.
These are the five lessons that have taken me hours of experimentation and testing, and that I would have loved to have known at the beginning of the journey:
Give complete prompts: define the role, context and objectives.
In AI, the quality of the instructions is everything. A simple sentence is often not enough: the more detailed and clear the prompt,, the better the results. Apply the ‘garbage in, garbage out’ maxim: if you enter imprecise instructions, you will get mediocre answers.
Define the role you want the AI to adopt (communication expert, copywriter…); provide context (target audience, channels, brand tone) and explain the end goal of the task. If you also specify the format in which you expect to get the response and give an example, all the better.
Here’s an example: “Give me a post for LinkedIn on the impact of AI in communication” is too vague.
It will give you better results to ask: “Act as an expert in communication and social media. Write a post for LinkedIn about the impact of AI in communication. The post should be between 400 and 500 words. Include a CTA that encourages people to leave a comment. Here are 2 examples of posts I have published on this topic so you can see my writing style and approach, and apply them to this text”.
Train the tool: experiment with various options and choose the best one (for you).
Not all generative AI tools work the same. It is essential to test different options and evaluate which one best suits your needs and the way you work.
Spend time experimenting, do the same task with several tools and compare their responses. Repeat the process as many times as necessary until you find ‘your AI tool’, the one you understand best and the one that works best for you.
Training the tool does not mean configuring it or knowing how to code, but understanding how to interact with it to get what you are looking for and using it repeatedly, telling it what it does well and what it does not, so that it learns and gives you answers that are more and more aligned with you.
Continuing with the example of the LinkedIn post: the first time you ask it for this task, you will probably not be convinced by the result. If you tell the tool what you like about the proposed post and what you don’t, make adjustments and pass it your final version, it will learn. In my experience, from the 3rd time you repeat this process, the results will pleasantly surprise you.
Invest in a paid license of the tool of your choice
The free versions are an excellent starting point, but they have limitations in features, speed and privacy. If you find a tool that works well for your workflow, I recommend you purchase a paid license. I assure you it will be an excellent investment, with a good return.
This will not only give you access to more features with fewer limitations, but will also guarantee a higher level of privacy and security. And yes: it will also give you more refined and higher quality results.
By the way: keep in mind that, for most tools, you can purchase licenses on a monthly basis. So you can sign up when you need it and unsubscribe when you are not going to use it. Although I warn you that it is very likely that when you get used to it, you will never want to work without it again.
Use it only in tasks where it really adds value.
Although AI can be useful in many areas, it is not a magic solution for everything. Prioritize those tasks where its implementation will save you time and improve quality.
And how do I know which tasks to apply it to and which not to apply it to? Well, once again, you will have to try and experiment to get the answer. In case it inspires you: in the example of the LinkedIn post, I would ask ChatGPT to give me ideas for a post about the impact of AI in communication. I would choose one of the proposed ideas to further refine with ChatGPT. Then I would give him sample posts and directions for length, approach, etc. But I would never directly ask you to write the post without further details.
Automate the operational and apply it only to what really enhances your skills. Being selective is key for AI to really improve productivity and not end up being a time thief.
Do not upload confidential information or personal data.
Ethics and data security should be a priority for a communication professional. AI tools, especially in their free versions, can use the content we give them to further train their models.
Avoid uploading sensitive, confidential or personal data-related information. Caution is essential to protect your work and that of your clients.
In my case, I apply a rule in projects with clients: I only upload to the AI tool information that is published openly on their website. If I need to work with data that is not public, I either ask the client for permission, or I anonymize it before uploading it.
Define your guidelines for ethical and responsible use of AI to avoid scares and legal problems.
And I add a bonus to the list: don’t feel bad about using AI in your work.
Using AI does not make you less professional or detract from your work. On the contrary, it shows that you are taking advantage of the tools available to enhance your performance.
Remember that it is not the AI that does the work: it is you who guides, configures and reviews each step. Working hand-in-hand with these tools not only allows you to be more efficient, but also empowers your capabilities and gives you superpowers. The key is to integrate AI as an ally and not as a substitute.
By the way: I wrote this post with the help of ChatGPT. I told him I was preparing a post for a communication agency’s blog, told him the topic and dictated the introductory paragraph and some personal learnings. From there we chatted until we refined the 5 recommendations. I then gave him ideas for each tip so he could prepare a draft. I edited the draft and, before closing the conversation, I shared the final version with him so he could review my final tweaks.
I didn’t give him examples of posts because I already have him trained: months ago I spent several hours to have him analyze my writing style, the type of articles I write, the expressions I usually use… With this information I created a GPT that doesn’t replace me, but helps me to write my articles. The result is not bad at all, don’t you think? How did the post sound to you?
*Article written by Cristina Aced, digital communication consultant and teacher, founder of the school of #LasImperdibles de comunicación
