How to Explain AI: A Short Guide for the Communications Professional.

As a Communications Professional, you should be aware of artificial intelligence (AI), how to explain it to your clients and colleagues, and how to use it to your advantage. This primer provides an overview of AI for Communications. As well as examples of how to use it in your workflow.

⚠️ Before we begin, there are two main challenges that come with using Artificial Intelligence in Communications.

The first is whether or not AI can effectively communicate the complexities and subtleties of messages in sensitive and specific contexts. And second, if AI does find success in communicating complex messages, it raises questions about who should be responsible for decisions regarding its use and whether it should be used without a human in the loop. The debate is a useful one. And each professional, or the organization in which he or she works must determine the best approach.

As the founder of HelloScribe, the ethics question is constantly at the forefront as we determine how to safely deploy these artificial intelligence writing systems to our customers. Having said that, we believe the advantages far outweigh any potential harms.

So, what is Artificial Intelligence?

Artificial Intelligence is when computer programs learn how to do things that usually require human intellect, like understanding language, making decisions, and learning from experience. More precisely, it is the art and science of creating intelligent agents with human-like capabilities.

To quote this article — “The various systems we call “AI” today are far from sentient — they have no concept of existence, no motivations that aren’t specifically supplied by humans, and no reliable way to describe, organize, or audit the knowledge they use to perform their tasks. AI models are built around datasets that humans create, meaning they can be as flawed or successful as we build them to be.”

Creating an AI system requires what’s called a “knowledge base” or a dataset. This knowledge base consists of data bits that are related to one another in such a way that they form meaningful patterns or insights. With this foundation, you can teach an AI system how to do things on its own by giving it various tasks to complete. As it gains “experience” in completing these tasks, it can carry them out with increasing levels of proficiency.

There are different types of AI, each with its own focus. Some areas of AI include Robotics, which is focused on creating intelligent machines that can complete mechanical tasks on their own; Computer Vision, which is focused on making computers able to understand the world by recognizing objects and movements; and Natural Language Processing (NLP), which is focused on understanding and processing language through voice or text.

As you can imagine, each branch of AI has various sub-disciplines. Within Natural Language Processing, for example, these include but are not limited to:

Text Mining — Used to find specific information, phrases, or even whole sentences within a text corpus using various techniques like keyword search or text classifier analysis.

Sentiment Analysis — Used to identify emotion in text like determining whether a piece of text is positive, negative, indifferent, etc.

Machine Translation — Used to analyze and translate text between different languages.

This primer will focus on one of the latest and most advanced Natural Language Processing systems, commonly referred to as Language Transformers.

What is a Language Transformer?

A Language Transformer is a kind of A.I. that can turn a written instruction into a related output. It does this using what is called a Large Language Model (LLM). Think of a Large Language Model as a modern digital library of words, phrases, sentences, and pictures. Since this library has so much data, the computer code that runs on top of it (The Transformer) can easily recognize patterns and relationships between its contents.

The Language Transformer can match your written instruction with those words, phrases, sentences, or pictures that it has learned. Importantly, it is not a database, so it’s not pulling information to give back to you. It’s a transformer. So in a quite literal sense, it’s transforming an instruction into a corresponding output using mathematical probability.

The Transformer is NOT a database that returns information to you. It is a program that uses mathematical probability to turn an instruction into a corresponding output.

Here is a simple example of a Language Transformer in use. Say you want to write an email to your team requesting a meeting on matters of importance. You want the tone of the email to sound urgent. You open your computer and start writing. Only this time, instead of writing the email in the 10 or 15 minutes it’ll take to find the right words, you are instructing an AI to do it for you. It will write your email in 5 seconds or less.

In the example above, our AI understands the COMMAND<< Write an email>>, the CONTEXT <<a meeting about matters of importance>>, and the TONE<<urgent>>.

It then gives you a series of outputs in grammatically perfect English. (it can do so in any other written language)

In this case, it returns five options from which to choose. But it can produce tens, hundreds, and even thousands of variations that match your instruction. It is virtually limitless.

Now think of this in relation to your work, and you immediately begin to see how it can be applied across a range of use cases. From writing emails and newsletters to coming up with creative ideas for internal campaigns, public outreach, marketing programs, and so on.

Cool, right?

Now that you understand a little bit more about A.I, and how it works in the realm of communications, let’s look at how to apply the technology in your daily practice.

Meet Sherry.

She is a Communications Manager for a large bank. She spends her days managing and coordinating communication efforts for her company. In her work, she uses a variety of channels to reach out to stakeholders, including emails, newsletters, and social media. She also uses the company intranet (Facebook for Work) to share company news with staff.

In her world, there’s no such thing as a 9–5 workday. There’s only 24/7. And she’s always on the lookout for ways to improve her work. Whether it’s finding a better way to communicate with high-net-worth customers, or working with her marketing team to create more effective materials, she’s always looking for ways to do her job better.

She’s working on a change management program. Her bank wants to introduce some new software for employees. This is a major exercise. And it’s Sherry’s job to create and manage the change communications for the bank’s 4,000+ employees.

She has to:

1. Communicate the changes in a clear and concise way to ensure that employees understand what is happening and why.

2. Create targeted campaigns designed to keep all employees up-to-date on the latest developments, including any applicable training resources or assistance available.

3. Post regular updates on her company’s blog and social media channel throughout the course of this major change rollout so there are no surprises when it comes time for them to start using the new software.

4. Maintain close communication with her team members throughout the process so they are fully informed and able to handle any questions or concerns as they arise

5. She also needs to be aware of how employees are reacting to the news — if they’re feeling overwhelmed or scared, she may need to provide additional support ahead of time.

Naturally, her job involves a lot of thinking and writing, and like many of her peers, she does her best work at night — away from the office and free from distractions. But sometimes she’s really tired, and fatigue sets in. This causes serious bouts of writer’s block, and getting from a blank page to a finished result can be unusually tough during these times.

She has been hearing a lot about artificial intelligence lately, and it intrigues her. After some deliberation, Sherry decides that perhaps AI can help with her writing process. She starts looking for an application that would fit the bill and eventually stumbles on one she likes.

• She brainstorms key messages for the program and is able to quickly decide which ones are important to the project, and the kinds of content she will need her team to develop.
• She drafts her emails, newsletter outlines, social media posts, and blurbs for articles. And she uses the tool to rewrite individual messages that are too turgid or unclear.
• She generates speech intros, and icebreakers that she can prompt her VP to use during her presentations. She even generates bullet points for the PowerPoint slides.
• She generates scenarios in which users might respond negatively to the change and runs simulation meetings with her team to create troubleshooting plans.
• She’s even able to help her internal marketing team come up with storylines and anecdotes to use in their videos about the new software.

It’s fascinating stuff. And as she goes along she discovers five important things:

1) You are in control.

Ultimately, AI is a tool that can help you complete tasks more efficiently. Like any tool, it needs to be used properly in order to be effective. You need to be clear about what tasks AI can help you with and set boundaries on its use so that it doesn’t become a distraction.

2) Let AI help you think, not just write.

If you’re still relying on Google search as your sole source of inspiration, you’re doing it wrong. A.I can help you to access and generate ideas that you might not be able to come up with on your own. This can be incredibly valuable when it comes to problem-solving and generating novel ideas. The large language models in use today are not just an arbitrary collection of words and pictures. They are a repository of potential solutions — the largest repository of potential solutions in existence. And you can instruct your A.I to give you a specific recommendation for your particular challenge.

3) Better context equals better results

When you provide AI with the right information, it can help you achieve the results you’re looking for. The key to this is being able to give clear, simple instructions. Just like you would give your intern who’s just learning about your business. If you’re using one of these tools, and expect to get a good result just by typing a few random, obscure words, you’ll likely get a big fat turd. Garbage in, garbage out, as the old saying goes.

4) It’s not a magic bullet

AI is not a magic bullet. It is not a one-size-fits-all solution to all of your communication needs. While AI can be incredibly helpful in a variety of settings, it is not a replacement for human intelligence. It should not be used in place of human creativity, critical thinking, or understanding of context. You should use AI as a tool to help you complete tasks more efficiently and uncover additional insights that you may not be able to come up with on your own. Don’t expect it to understand every nuance and subtlety. That’s your job.

5) Ignore the critics

Can you imagine the first person who used a calculator? Or the first person to write with a quill? How about the person who first printed a book? These were all innovations that met with criticism at first. Some people deemed them ridiculous and inconvenient. But eventually, everyone came to accept them and see the immense potential they had. Critics can be a lot like that. Their opinions don’t always matter.

Sherry’s experience with AI has been positive overall. She finds that it’s a helpful tool that can save her time when writing and it can help her generate ideas. She advises others to give it a try, but to be clear about what AI can and can’t do, and to set proper boundaries.

Closing thoughts

Artificial Intelligence presents a number of opportunities for communications professionals. And as it becomes more prevalent, there will be an increasing demand for skilled professionals who can understand and use it to improve communication outcomes.

We hope this short primer has given you a better understanding of the basics. As with any tool, misuse can lead to harm, so use them responsibly. As there are no guarantees of how successful AI will be in any given scenario, so have patience, practice often, and keep trying new things.

Ready to explore A.I. tools even more? Head over to HelloScribe and get started. You’ll find the tools that can help you get more done 10x faster.