Navigating the New Frontier of AI In Public Relations

Probably a decade or so ago, back when we did Fireside Chats with marketing and comms experts on Spin Sucks, I had Chris Penn as a guest (I just looked; it was seven years ago!). We talked about artificial intelligence.

He said something that day that has stuck with me all of this time. I’m paraphrasing, but he essentially said that someday, our jobs will be not unlike an orchestra conductor. We will have a team of robots that do the work we need done, and we will conduct how they all come together.

At the time, it seemed unlikely that that world would ever happen, but here we are. With all of the generative AI tools at our fingertips and semantic search changing the way we think about content creation, we are now conducting robots to do our bidding (but again, they’re still not doing our dishes or laundry!). 

Not a single day goes by that I’m not using ChatGPT, Perplexity, Gemini, and Opus to help with our work—and we become more dependent on each of those tools as they continue to improve. It’s like having a team of really smart interns working with us 24/7.

Penn was right: we are conducting an orchestra of robots.

You’re Not Immune to AI

Greg Brooks recently snipped from a Hacker News comment related to OpenAI’s new release and posted it on Reddit (which says a lot that I saw because Reddit scares me, but I go there to see what smart advice he’s doling out). 

The comment was, “We’re moving toward a world where every job will be modeled, and you’ll either be an AI owner, a model architect, an agent/hardware engineer, a technician, or just..training data.”

Let me repeat: you’ll be an AI owner, a model architect, an agent/hardware engineer, a technician, or training data.

Just as Chris predicted seven years ago, you are conducting an orchestra of robots in each of those scenarios.

I know many of you don’t want to use AI for whatever reason: it’s scary, will take your job, or can’t replace the human touch of what you do. A client told me the other day that he told his team, “Not using AI is like being an accountant who wants to do all of their calculations with paper and pen instead of Excel.” 

A-freaking-men!

You’re not immune to AI, and it’s not scary. Is it going to replace you? Of course not! Is your job going to change? For sure, it is! 

Some people think we’ll be replaced by robots and all sit around on our couches, watching Netflix. That’s not going to happen. But the way you do your job will absolutely be replaced. 

So why not be ready versus putting your fingers in your ears and singing, “La la la la la” as loudly as you can?

Let’s explore each scenario Greg snipped to get your juices flowing.

The AI Owner

The AI owner. Imagine, if you will, you have an expertise or a certain skill set. Crisis communications or strategic planning, for instance. Or you have a framework only your organization or agency uses—the PESO Model©.

This expertise will eventually require custom-trained models that you can use to work with customers or clients. 

A smart domain-focused PR firm, like one focused on manufacturing, law, or beauty products, could train those models and build a business by renting them to clients. 

It would make a lot of sense for my team and me to create one for the PESO Model, right (it’s like you’re looking into our strategic plan!). You then could achieve your PESO Model Certification® because it’s imperative for you to know enough to be able to conduct your orchestra of robots and then use our trained AI to help you execute it, tweak it, measure it, and improve.

But you don’t have to have a framework that an entire industry uses to be an AI owner. Let’s say your expertise is in crisis communications, and you work for an agency. Rather than have clients call you when something happens, you can train an AI to predict what might happen—and when—for clients you already have, prospects you’d like to work with, and an industry overall. 

To do that, there are four steps:

1. Gather historical data on past PR crises, including news articles, social media posts, and company announcements. 
2. Train AIl to recognize patterns and early warning signs of potential PR crises.
3. Deploy the model as a real-time monitoring tool that alerts clients and prospects to potential crises before they escalate.
4. Charge a subscription fee for access to the crisis prediction tool, with higher fees for advanced features like automated response suggestions and even higher fees to work with you to get them through the crisis.

The Model Architect

Perhaps owning the AI isn’t the right fit for you right now—either for your business, the organization you work for, or yourself personally. That’s OK! There are other ways AI will help you become more efficient as you evolve your career versus being replaced.

Another is the model architect. This person, team, organization, or agency focuses on improving existing AI models to support their processes. 

We’re already at the point where becoming a model architect will benefit you—and some of you may already be doing it without even realizing it. Your daily life has many practical applications: sentiment analysis, media monitoring, crisis prediction and management, content generation (ideas or first draft only!), and influence identification.

You’re probably using AI for most of those things already, maybe without even realizing it, especially if you use software for some of it. This is great because you’re already on the path to becoming an AI model architect.

To get yourself over the finish line and start offering it either as part of your skillset as you move up the ladder or to clients if you work for an agency, there are four steps for you to take:

1. Start learning how it works by reading, watching, and listening to as much AI content in PR as possible. You’re already doing that just by listening to this podcast. Keep that up because we’ll continue to discuss where AI is useful for you at that moment in time. And find others to learn from.
2. Complete small projects that allow you to understand how it works, how to prompt it to get what you need, and how to deliver what is being asked of you. For instance, a common question PR pros ask in the Spin Sucks Community is, “How do I provide an automated media and social media monitoring report so I don’t have to do it every day manually?” While some software will create the daily report and email it to your list of executives, Truescope, for instance, you still have to tweak it to get it where you need it. Instead of doing that manually, use AI to get you what you need before the email is sent out each day.
3. Once you have that figured out, move to something else. Keep testing, iterating, and trying new things. For instance, we have a client that hosts a weekly podcast. It used to take us FOREVER to create content for each episode, but with AI, it now takes a fraction of the time. We upload the transcript into the AI of choice (some of us use Castmagic, some use ChatGPT, and others use Perplexity) and prompt it to provide headline ideas, social media post ideas, four blog post ideas, timestamps, and quotes. From there, it’s pretty easy to draft the pieces we need to promote each episode without listening to them several times to create the assets we need to build the content.
4. Once you feel comfortable with how you’ve set the AI up to do the tasks you don’t want to do, you can launch it more broadly. For agencies, this might look like a subscription model where you charge clients for access to the AI to do what you might have done manually in the past. For in-house folks, this might be incorporated into your standard operating procedures and taught to colleagues to improve the team’s efficiency. 

The Agent or Hardware Engineer

The agent or hardware engineer is where many of us are right now. We’re experimenting and figuring out how to make AI work in our daily lives and for our organizations and/or clients. 

This is a start, but it’s not a way to build a career or business with a moat around it (aka get paid more or charge more). 

Suppose you want to use AI to increase your efficiency and effectiveness. In that case, building an army (or orchestra if we’re going with the conductor analogy) of engineers is where you should begin to focus your time. 

In this case, as an AI agent, you develop AI that automates and optimizes the various tasks you do every day. Think admin tasks, such as media list development or editorial calendars, media monitoring, sentiment analysis, and results reporting. All of the things you probably don’t enjoy doing very much, but are critical to the success of your job. 

Here is how you might do that:

• Build an AI agent that scans news articles and social media posts, summarizing key points and identifying potential PR issues. The reports can be daily, weekly, or monthly (depending on the topic) and can be emailed to you first and then to a larger group of stakeholders.
• A crisis management bot might monitor for signs of a PR crisis, send alerts, and provide recommended actions. This will enable you to quickly alert a client or senior leadership team that something is amiss and you should react quickly. 
• Build a chatbot that handles client inquiries on social media and in email, providing instant responses and escalating issues when necessary. This is especially nice when you have international clients or customers who are working while you sleep.

When you do this kind of work, you create an orchestra of robots to do your bidding for you so you can focus on the more strategic—and more valuable—parts of your job.

The Technician

The AI technician bridges the gap between developers, engineers, and comms pros. This person is learning how this works but doesn’t have enough experience to conduct an entire symphony alone. They likely have zero to seven years of experience and are learning the trade and how to use AI in their jobs. Eventually, this person wants to be an engineer, architect, or agent and is learning as much as possible along the way. 

An example of how a technician might work in today’s comms world is that you have the software set up to monitor media, social media, and trends, and, for the most part, the AI does its job. However, some oversight still needs to be done to ensure the AI is correctly trained on sentiment and share of voice.

The technician might log in every day and scroll through all of the mentions to confirm the sentiment and other metrics are reported correctly. If they are not, the AI technician corrects them, which trains the AI to tweak what it delivers the next time.

Eventually, the software does its job without the technician. That person can move on to tasks such as crisis management, trends reporting, search and AI optimization, and/or content creation. 

While not a glorious job, it’s necessary as someone moves up through the PR ranks. Once you know how to conduct one robot, you can eventually conduct an entire symphony.

The Data Trainer

The data trainer ensures the AI systems deliver accurate, unbiased, and ethical outputs. Just the other day, we had a conversation in the Spin Sucks Community about the dangers of AI, and some wondered if it’s too late to put the genie back into the bottle.

I do think it’s too late—and I think the pros far outweigh the cons, but there are many considerations PR pros should take into account. We talked earlier this year about how we can responsibly navigate weaponized information, and the same goes for AI. 

Unfortunately, PR pros will figure out that they can affect AI’s output by flooding it with biased training data. The same will be true using AI, just like we have whisper and smear campaigns. 

It’s highly unethical, but human nature and its sheer amplification make it tempting to think someone (or someones) won’t do it.

If you aim to be someone who trains the data, there are some pros and some cons to consider. 

The pros include:

• You get to decide what the AI brings to the forefront and what it ignores. You directly influence how these models assist in PR activities and can build and train them to provide what you need with the stroke of a key.
• Gaining AI and data science expertise can open up new career opportunities and positions within the PR industry and beyond.
• Ensuring AI models are trained on unbiased and representative data contributes to ethical AI development and use.

The cons include: 

• The role requires a deep understanding of AI technologies and the specific needs and nuances of the PR industry, which can be challenging to master, particularly early in one’s career.
• Collecting, cleaning, and annotating data is time-consuming and requires meticulous attention to detail.
• Even with the best intentions, there is always a risk of unintentional bias in the training data, which can affect the AI model’s performance and fairness.

Your Own Orchestra of Robots

As with anything, time will tell where this all goes. It’s an exciting time to work in our industry, full of innovation and technological advancements that we have yet to experience.

If you open your mind to the possibilities versus sticking your head in the sand because of the dangers, you’ll soon be on your way to conducting an AI orchestra of robots of your own.