One Year After ChatGPT, Whither Comms? (Intellicomms)
The authority on mixternal communications
ChatGPT was released to the world one year ago on Nov. 30, 2022. No one has ever accused comms pros of being quick to embrace new technology, which is why early on I argued that we ignore artificial intelligence (AI) at our professional peril.
Three months after that, as I became more familiar with ChatGPT and other AI tools, it became clear to me that a line had been drawn in the sand:
One day not far in the future we will refer to life before and life after ChatGPT. I believe the release of the generative artificial intelligence platform in November 2022 was a watershed moment in human creativity and innovation, akin to the introduction of the affordable personal computer, and may be as disruptive as the printing press.
Artificial Intelligence is here to stay.
There’s too much money (and too much at risk in the realm of national security) for the genie to be put back into the bottle.
So where are we today with AI and mixternal communications? It’s clear that AI, particularly ChatGPT, has influenced our industry in many ways . . . and there’s still a long way to go.
Enhanced Efficiency and Productivity
AI can streamline common mixternal communications work. Everyday tasks like drafting emails and press releases, creating content (especially articles and visuals), and summarizing research and reports are waaaaaay more efficient when you use AI.
No longer do we need to stare at a blank page wondering how to get started and then spend hours on the first draft. The right prompts can get you 80% of the way there in seconds.
Affect on Job Roles and Skills
To my knowledge, no comms pros lost their jobs to AI this year.
But…but… In the next three years fewer comms jobs will be created, thanks to AI.
While AI can automate some tedious tasks (e.g., recapping hour-long meetings), it hasn't replaced the need for skilled communicators (yet?). Rather, AI’s presence can shift the focus toward more strategic, creative, analytical, and innovative roles.
As the now trite soundbite goes: AI won’t take your job, but someone using AI will.
Quality Over Quantity
ChatGPT and similar AI tools are handy for generating initial drafts and ginning up content ideas. But we still need humans to manage the quality, tone, context, and nuance, as well as to check for bias.
As I said on Simpplr’s Cohesion podcast, AI has its creative shortcomings.
Generative AI is not creative because it creates content based on material that already exists. It doesn't have much ability to come up with something original, and that's where the human using AI can stand out against this technology.
Generative AI doesn't handle rare events very well because, again, it's basing its answers on things that have already existed or happened. If you confront an activity or an event or a circumstance that you need assistance with that just hasn't happened before, you're not gonna get much help there.
Also, AI can’t love you.
Ethical Considerations
The use of AI has raised questions about ethics, particularly concerning data privacy and the potential misuse of AI for spreading misinformation.
But what about ethics in using AI to create mixternal communications? One of the more influential reports on this topic was published in 2020 by CIPR—three years before ChatGPT’s release. Needless to say, a living framework for how to use these tools ethically in our industry is needed.
My thoughts on the ethics of using AI in mixternal comms are evolving, but here are a few ideas ($) on whether and when to be transparent about how AI is used in your comms work.
Learning and Development
As AI tools become more integral to mixternal communications, there's a growing need for continuous learning and development. That’s why I created Intellicomms ($), a groundbreaking series that explores the intersection of AI and mixternal communications.
Intellicomms offers practical examples of how to use AI to enhance productivity, efficiency, innovation, and creativity in your comms role.
We’re in this together.
The Future of AI and Comms
We are experiencing a Cambrian explosion of AI tools. I maintain a database of AI tools across 16 categories (e.g., content generation, research assistance, operations) that comms pros can use to become more efficient, productive, and creative in their roles.
My database contains 287 tools and grows each week. Today alone I added six more to the gold mine.
Don’t expect any bold predictions from me on the future of AI and comms. I’m happy enough to proclaim that we should incorporate AI technology as much as possible into our everyday work and make it a part of long-term career plans.
AI will be as vital as the keyboard to do our work.
There is some variety in our small industry. Within our niches—exec comms, media relations, change management, editorial, etc.—there are numerous and specific AI tools that can enhance your work, so you can have a profound influence at your organization and within our profession.
I encourage you to explore. Those who genuinely embrace the opportunity may become richly rewarded for the effort ($).
I’m excited to see what we can do.
Back to the Future
In July I appeared on the Cohesion podcast with Simpplr’s Carolyn Clark to discuss the early days of AI and its influence on internal comms.
The tension in the conversation was between the imperative of using AI to improve our comms game and the fact that it was—and remains—all too new and too soon to know exactly where best to apply the technology. (Though I think things are coming into focus.)
I’m sharing excerpts from that conversation as a reminder that life seems to move simultaneously fast and slow.
On one hand, AI has grabbed headlines all year long, reflecting our collective fascination and anxiety over the emerging technology.
On the other hand, as far as comms goes, not much has changed. While many comms-rades have warmed to a couple of tools, the majority of our peers have yet to fold AI into their daily workflows.
The following excerpts have been lightly edited from a rough transcript. I hope the conversation provides an easy introduction to the role AI can play in upping your comms game.
Host: What is ChatGPT?
Shaun (a.k.a. Mister Editorial) The boring answer is that it's an artificial intelligence language model that's been trained on a vast array of text. It's been designed to understand and generate human-like responses based on the inputs it receives.
The more interesting answer is that you can think of it as a virtual assistant that can engage in conversations with you. It can understand your questions, statements, and prompts as long as you're using natural language, and then provide coherent responses and generate text that's contextually appropriate and sort of simulates human conversation.
In this section, Carolyn Clark describes how “you get what you put in.”
What role can ChatGPT play in internal communications?
It’s too soon to fully understand the role it's gonna play. A lot of people are asking for best practices for using AI or ChatGPT, and I'm saying best practices? We're still in the experimenting phase. We need to design these practices as we go along.
Right now I see it as a valuable virtual assistant, junior, associate slash intern. It has some very basic capabilities and can give anyone a headstart if you're running out of time or you've never been confronted with that question or problem before and you're facing a blank page and just don't know where to start.
ChatGPT can knock off hours and hours of your work week just by putting together a simple prompt.
For example, let's say you are new in your internal comms career and you're two years in, and the CEO wants to send a message to employees about the death of a high-profile employee.
How often do you write notes from the CEO expressing grief about the passing of a colleague? Well, maybe you asked ChatGPT for that first draft, just to get a sense of what that would even sound like and where to go rather than just trying to start from scratch. It’s a jump starter.
Carolyn explains how she uses ChatGPT as an “extra set of eyes” and to play a persona.
How do you use ChatGPT?
The first thing that people should think about is the name. It literally says chat in the name because it’s a chatbot. You can just ask it questions, it'll give you an answer, and then you can ask it to follow up or refine or revise it. It's like you're texting a virtual assistant so you can have that kind of conversation.
The next iteration of that is to assign ChatGPT a virtual assistant persona. You may not know the questions to ask, but you can just tell it: “Give me a headstart on something.”
For example, this week I am doing some live reporting for an internal-only conference, and our company uses Viva Engage as as our internal chat platform, a platform that is new to me. So I can ask ChatGPT, “how would you report live from an internal conference using Viva Engage?”
It gave me an answer, at least something to start with. It was more technical about how to set it up in the backend and use certain features and functionalities. But being unfamiliar with the platform, it gave me a way to start thinking about how to use Viva Engage for this project.
What should you not put in ChatGPT?
I absolutely do not put any information about my company into any of these AI platforms. With that example about how do I provide live updates from an internal conference using a certain chat platform, there's nothing about my company in that kind of prompt, nothing about my CEO, nothing about the products we make, nothing about our employees, nothing about where the conference is happening.
The reason you want to keep personal and proprietary information out of these AI platforms is that they own the data, the information is stored on their servers. It is vulnerable to cyber-attacks and the last thing that you want is to compromise your company or your personal information should a leak occur.
What are some ways you two both see ChatGPT AI has already begun to improve and transform internal comms or HR within companies?
I can't speak too much about HR. It's a space I don't have much experience in, but I know for more than 10 years, HR has been using AI to filter and read resumes as a first round of recruiting, for example.
Again, I think we're too soon to understand how internal comms is being transformed. That's for a couple of reasons.
One, comms pros are notoriously slow to pick up on new technology. I mean, remember when eight years ago everybody was claiming We need and app! The Apple app store launched in 2008!
Second, internal comms is one of the last functions to get the budget resources that are needed to invest in these new technologies.
If we see any transformation happening, it’s happening at the individual level.
It's happening on the fringes and it's happening with those who are ambitious because as I said in the introduction to my initial piece on Intellicomms, AI will not replace you, but somebody using AI will.
And so I think those who are experimenting with the technology right now on an individual level within their teams, maybe their team doesn't even know that they're using it.
For general purposes, the safest place and the most fun is probably with image generation, because when you ask AI to help create an image, it's a benign act.
You’re not asking ChatGPT to write a memo in a certain CEO's voice about a sensitive topic like layoffs or anything like that.
Many communications teams don't have a multimedia artist within the function, but you need graphics to entice readers to read your material. So why not experiment with something like DALL-E or Midjourney to create arresting graphics for your intranet story in a matter of seconds without having to tap a consulting service for thousands of dollars for a piece of artwork.
Image generation is the gateway drug to generative AI.
I can count on one hand the number of comms pros who I think are seriously talking about AI and corporate communications, and I'm talking with one of them in this podcast right now.
Carolyn explains how she uses ChatGPT to create FAQs and to check bias in her messages.
What makes you nervous about AI?
I've got so many spreadsheets, so many reports, so many thoughts about AI and I would not call myself an expert, even though I think I know more than most corporate comms professionals about this topic.
I look at what the actual experts out there are saying and doing, and they're kind of freaking out a little bit, and we're recording this podcast on a day in which hundreds of AI professionals signed a one-sentence statement about mitigating the risk of extinction from AI.
There are a few things happening here.
One is it's, you're gonna read about that statement because it's alarmist and media loves alarming stories. The second is, I think it is a bunch of cover-your-ass activity happening right now.
All these engineers, psychologists, and philosophers have created this technology and unleashed it, and now they're kind of apologizing for it and for trying to distance themselves from it, from the negative consequences of the monsters that they released.
I think there is mostly sincerity here because I think that the individuals who signed that statement believe themselves to be the trusted individuals who can help usher AI technology forward in a way that doesn't harm us. I think they're sincerely worried about nefarious actors leveraging AI technology to do bad things and tear society apart.
But I also think at the same time, they're like, we were against nuclear war caused by robots. We said so in a statement back in 2023. So don't blame us. But. I mean, they built it!
I don't think any of the concern is unfounded. Anybody who has read Ray Kurzweil's The Singularity Is Near should have a sense that the end might be near too.
If bad actors get a hold of artificial intelligence, I think the most immediate effects that we will see on the negative side are around disinformation, and that's fake photos, fake videos, and fake audio.
Already, platforms like Spotify are removing thousands and thousands and thousands of AI-generated music.
Russia is using AI-generated technology for propaganda in their war in Ukraine. They're using it on the Ukraine side, they're using it on the the Russian side.
I guarantee you that in the 2024 U.S. presidential election, we will contend with fake video, fake audio, and fake photography. Maybe even from an outside actor trying to manipulate the US election.
Jonathan Haidt, who is a social psychologist among other things, has a lot to say about how AI will ramp up that negativity and rage and misinformation and social media, which will create a vicious cycle for how we perceive our fellow citizens and so on. So I think disinformation is the most immediate and real fear for me.
I think AI has the potential to tear society apart.
Our American society is already fragile and I think the volume on disinformation will ratchet up.
Ironically, many people are talking about how AI will be needed to spot AI-generated content and mark it as valid or manipulated.
On the other hand, I think anybody who's seriously interested in artificial intelligence and its effect on society should approach the individuals who are in the media talking about this with a critical eye. Geoffrey Hinton is considered the godfather of AI. He retired from Google to speak out about the dangers of AI.
But guess what? Hinton built that AI! Weren't you aware of these dangers all along while you were building it and collecting your fat Google paycheck? Now that you're doing all right, it's okay for you to retire and speak out about this.?
And then Elon Musk speaks out about ChatGPT and slowing down the advancements in artificial intelligence. Well, Musk is a co-founder of OpenAI, which started ChatGPT, and Elon Musk is starting his own competing AI platform. So I think he's less about the dangers of AI and more about falling behind in the marketplace, and he wants to catch up.
So when, when people are speaking out against AI, ask, who are they? Where are they coming from? And weren't you aware of all this all along? Why didn't you tap the brakes before you released the monster? So just think critically about the voices we're hearing out there.
Carolyn talks about anxiety around determining whether sources are reliable.
Will AI take our comms jobs?
I don't think AI will replace you, but somebody using AI will. I also think that the number of jobs in our industry will be fewer because one person can do more using these technologies.
Or, I don't know, maybe it's the same amount of jobs and people's time is spent more on thoughtful strategic work rather than rote, junior-level material.
Five years ago there was a study of 200 tech applications and it found that 38% of PR work could be done using AI. Now we have ChatGPT and an explosion of AI tools. I haven't seen a study on this lately, but my guess is 60 to 70% of PR work now can be done using AI.
So what does that mean for the PR department?
Does that mean you still have five people working on PR?
Are they being more productive and efficient because they're using AI tools to write press releases, media advisories, and interview background notes?
Or does that mean that department goes from five to one or two and the money is reallocated somewhere else?
The point is, the two PR people left standing at their company are using AI. That's just the future of corporate communications. AI will be as vital as the keyboard to do our work.
Carolyn suggests AI can free up time for strategic work
How will AI affect the employee experience/employee engagement?
Employees don't care where the information comes from, as long as they get it. Many may or may not be able to spot generative content.
General Motors did a study a while ago on what they deemed social AI and how documents and information were surfaced for employees on the desktop.
What they found was that AI was able to reveal documents, people connections, and projects that they didn't know existed, but were popping up in their feeds because of keywords, history, and browsing.
AI made them more productive and more efficient.
AI increased their happiness. It decreased attrition among that workforce and increased the quality of the output.
So I think AI will have a bigger influence on how employees connect, share, and find information. Generative AI they could probably care less about.
Carolyn talks about how AI can help respond quicker to below-the-surface employee sentiment and provide better context for messages.
What can ChatGPT not do?
I don't know about the future, but right now, as we've mentioned before, it doesn't understand context, doesn't understand the context of your organization, and so the output from something like ChatGPT may be inappropriate.
Most importantly, it's not creative. Generative AI creates content based on material that already exists. It doesn't have much ability to come up with something original, and that's where the human using AI or not can stand out against this technology.
Generative AI doesn't handle rare events very well because again, it's basing its answers on things that have already existed or happened.
So if you confront an activity or an event or a circumstance that you need assistance with that just hasn't happened before, you're not gonna get much help there.
AI can't love you.
Carolyn talks about how AI lacks the human touch.
What are some resources on security and ethics around AI and ChatGPT?
Ethics and security are very specialized areas around AI. I wouldn't expect any comms person to become an expert in this area.
We do a lot of communications with our employees. Are we experts in organizational psychology? We use technology platforms. Are we necessarily experts in cybersecurity or do we understand supply chain ethics around the computers that we're assigning to new employees?
I wouldn't ask our comms-rades to freak out about not being an expert in the security space around this, but I do think they should have some general awareness and knowledge of the topic. If you're listening to this podcast, you're already on the right track.
Sam Altman, who is one of the founders of OpenAI, is doing a very good job of trying to explain ethical concerns around his company’s technology. He's on a lot of podcasts. You can find him in the media, seek him out. I think he is a thoughtful individual and he was recently testifying in front of Congress about this very topic.
IBM has a good primer on AI and ethics, so you can just google IBM Primer AI Ethics and you'll find it.
If you want to get into the weeds around cybersecurity ethics and technical concerns, you can go to Anthropic’s website. Anthropic is another AI technology company that tries to be ethically-minded with its output. They have a lot of white papers that are generally readable by a lay audience.
There's a lot of froth out there right now, and I think you don't need to stay on top of the day-to-day movements. Check in once a week or once every three weeks. There's so much to talk about in the news. It's a hot topic right now.
It'll, it'll simmer down or settle down, so you've got time to catch up.
Carolyn provides tips for staying on top of AI’s evolving landscape.
Thank you for being a part of Mister Editorial.
Read past editions of Mister Editorial here. If you’re enjoying what you’re reading, please consider recommending it to others. They can sign up here.
Have feedback? Email me at editorshaun@gmail.com. Connect with me on LinkedIn and X.
Disclaimer: Besides running Mister Editorial, I am the editor-in-chief of Digital Publications at Lam Research. The views in this newsletter are my own.