Communicators of all stripes are torn by the promise and peril of Generative Artificial Intelligence. From marketing to PR, from advertising to internal communications, communicators are moved nearly to tears by AI’s time-saving capabilities. AI handles mundane tasks with ease, freeing up precious time for more creative endeavors.

At the same time, generative AI’s ability to write and research has many communicators wondering how long they’ll have jobs. In a perfect world, company senior leaders would recognize that AI will allow them to expand their operations without needing to hire more people. We don’t live in a perfect world, though, and there will be leaders aplenty who will conduct layoffs so their companies can do no more than what they’re already doing with fewer people on the payroll.

Both elation and fear are called for. However, communicators have not yet grasped where AI’s role in communications is headed.

Agents Everywhere

Sometime this year, according to OpenAI’s publicly announced product roadmap, the company behind ChatGPT and DALL-E will introduce generative AI agents for anybody to use. At its annual Build event this year, Microsoft announced Copilot Agents, which (the company says) anybody can build inside the Copilot Studio. Microsoft’s announcement came on the heels of Google sharing AI Agents, its own version of the task-performing assistant.

AI agents can execute multi-step tasks independently to produce the desired outcome by whoever submits the prompt. While OpenAI, Google, and Microsoft’s agents are not yet available to the public, you can catch a glimpse of how they’ll work with AgentGPT.  It’s not perfect by the farthest stretch of the imagination, but it offers a glimpse into a radically changed future. (It’s not to be confused with another AgentGPT, which produces human-like agents to interact with customers.)

Generative AI agents could well bring about the Communications Singularity.

The idea of a communications singularity germinated in my mind while reading the latest edition of Ethan Mollick’s Substack, “One Useful Thing.” In the May 26 edition, headlined “Four Singularities for Research,” the Wharton professor ruminated on how generative AI could change the nature of academic research. In the newsletter, he notes:

“I don’t mean The Singularity, the hypothetical moment that humans build a machine smarter than themselves and life changers forever, but a narrow version. A narrow singularity,” Mollick explains, “is a future point in human affairs where AI has so altered a field or industry that we cannot fully imagine what the world on the other side of that singularity looks like.”

Try as I might, I cannot imagine what the world of communications will look like on the other side of AI agents being employed in support of professional organizational communications.

An Agent-Created Campaign

An example sheds light on why this future, as I see it, is so fuzzy.

I prompted AgentGPT with the following instruction: You are a product marketer for a consumer packaged goods company. The company will launch a new Smart Plant Pot with built-in sensors and Wi-Fi connectivity that monitors soil moisture, sunlight, and temperature. It sends notifications to your phone when the plant needs watering, adjusting for different plant types. It features customizable LED grow lights, an integrated app with plant care tips, and voice assistant compatibility for voice commands. Produce the marketing campaign for this product, including a product name and slogan, a launch press release, a list of reporters and publications that would be interested in the product, and other typical elements of a launch communication.

The tasks it assigned, before reaching the maximum of five “loops” that come with a free account—included the following:

• Research the target audience and identify the key benefits and features of the Smart Plant Pot.
• Develop a product name and slogan that highlights the unique selling points of the Smart Plant Pot.
• Write a press release announcing the launch of the Smart Plant Pot, including details about its features, benefits, and availability.
• Compile a list of reporters and publications that cover topics related to gardening, smart home technology, and lifestyle and send the press release to them.
• Write a social media campaign for the Smart Plant Pot, including a series of posts highlighting its key features and benefits.
• Write a blog post about how the Smart Plant Pot can improve plant care and make gardening easier for everyone.
• Write a product description for the Smart Plant Pot to be used on the company’s website.

This list represents just the tasks. AgentGPT also completed them. The audience research task, for example, identified gardening enthusiasts and busy professionals, assigning key benefits and features of the Smart Plant Pot that would appeal to each audience and then suggesting appropriate campaign elements. It developed a product name and slogan. It wrote multiple press releases. It created a six-part social media campaign, something I had not included in my prompt. It was about to write the blog post when the five available loops ran out.

All the while, I sat back and watched.

If I had a paid account, it would have just kept going. The same instruction on one of the frontier models (ChatGPT or Gemini, for example, which offer image generation) could even produce logos, promotional graphics, social media ads, and voiceover copy for audio promotion. (I used Midjourney to create the graphic at right, which included a potential logo engraved on the product.) Even the schematics for in-store displays aren’t beyond generative AI’s capabilities. If you have subscriptions to AI audio and video generators, it could even create commercials and jingles, all based on thorough audience research.

Our Uncertain Future

Granted, the press releases AgentGPT wrote aren’t very good, nor are the social media posts. All of that is improving with the latest releases of various frontier models. As Mollick points out in his New York Times bestselling book, Co-Intelligence, the AI you’re using right now is the worst one you’ll ever use.

Communicators have long struggled with clients and employers who think their ideas for marketing campaigns or internal communication strategies are fine because, they reckon, everyone’s a communicator. With AI agents, some of those bosses will figure they no longer need communicators at all since they make the assignment and run with whatever the agent produces.

I don’t see that as the most likely scenario in most companies. While AI agents can perform many tasks as well as—if not better than—humans, they lack nuanced understanding, creativity, and strategic thinking. Collaboration between humans and AI seems more likely.

I envision a world where communicators use AI agents to conduct comprehensive research and make recommendations based on the results, generate initial drafts and preliminary graphics, kick-start thinking on messaging, and handle distribution. This would free up time for them to craft compelling narratives, build relationships, and create innovative strategies that resonate on a deeper level with their audiences. AI can assist in identifying trends and patterns in data, but it takes human intuition and experience to translate those insights into effective communication campaigns. It is incumbent on communicators to demonstrate this collaboration’s effectiveness and quantify its benefits over machine-only communication.

But still, this is speculation. I don’t know what communications will look like after agents suddenly appear in our toolbox.

I still believe the role of communicators will evolve rather than disappear. We will most certainly need to develop new skills to work effectively with AI. Training and upskilling will become essential, and organizations that invest in their communicators will see the most significant benefits, while those that dispense with communicators in favor of AI-generated campaigns will probably suffer.

Whatever the outcome, the Communications Singularity is almost here. While we wait to see exactly how that plays out, we must put aside fears of the rise of AI and see it instead as an opportunity to redefine our roles, focus on what we do best, and drive our industry forward in ways we never imagined possible.