Communications Has Never Been More Important in the Age of AI

By SBS Founder John O'Brien

TL;DR: The comms industry’s path forward is to create a new digital trail for AI systems that considers how information is prioritized to better compete within a new search and information delivery paradigm.

ChatGPT has lit the world on fire. I use it nearly every day now and in the near future I think it will only get deeper into the daily workflow. I love it. Recently, something struck me that I can’t stop thinking about. Amid all the talk of ChatGPT one day being able to do the job of a comms professional (which I am not worried about) is the reality that comms is now more important than ever. Yes, I’m a biased source, but there’s a reality here that most have not considered.?

I have argued for years that the communications / public relations industry as a whole is undervalued. Attempts to measure the impact of announcements or campaigns produce nice graphs and may have some good insights, but they largely compartmentalize and limit the actual impact of the work. We see the value of media stories, blogs, Tweet storms, press releases, and more as limited to days, weeks or months. There is almost no value placed on that same information living on the Internet for years or decades. Nobody, communications professionals included, thinks about or understands how the comms work they do today will have value years from now. But it does have value. I can prove it.?

ChatGPT and other Generative AI models produce results that are synthesized from an ocean of data. If you don’t know anything about Large Language Models (LLMs) or Neural Networks, the gist is that data is “weighted” differently depending on where it came from, who wrote or published it, which terms are in the text, and many other factors. As data gets weighted it also gets filtered through to provide the most helpful answer to any query. What is the most important aspect to a helpful answer? Truth.?

Truth, along with relevance, are the most important factors in creating a system of value for consumers and businesses. As such, when massive AI models like ChatGPT are architected and trained, they prioritize sources of truth. Twitter, for example, would be low on truth rating for obvious reasons. If I were to post a blog about my favorite ski mountains on Medium it could be part of an LLM’s training, but not weighted high on the truth scale because--who am I? I have no history of writing about the best ski mountains in the world. I am not a journalist for Ski Magazine, right??

So now you see what I’m getting at.?

Search in the age of Generative AI will be highly reliant on two things:

1) What a person, business, group, political leader or political party says about themselves

2) What credible sources like academic research, journalists, and third parties interviewed by journalists say about all of the above

This breakdown also touches on something near and dear to my heart as a communications professional: the separation of go-to-market messaging and storytelling. Because search is changing from delivering blue links to delivering synthesized results with context, if you ask ChatGPT about a specific company or that company's product or service, it will very likely pull directly from that company’s own description. Have you asked ChatGPT what your company is? Do you like the answer? If not, how can it be changed? This is an important exercise for marketers to start today. Along with adapting go-to-market messaging, this might also lead to the return of the press release – clear, fact-based artifacts that normally pass through legal would be high on the truth scale for AI systems.???

The biggest thing though, in my opinion, is storytelling and journalism. As ChatGPT and other AI systems displace traditional search, more and more buying decisions will be made by the contextual search results that are produced. Consider competitive searches such as “what are the most promising cybersecurity technologies in the market today?” or “who are the most promising photonic computing companies?” To rank and stack-rank technologies and businesses, AI systems need highly contextualized, unbiased, and truthful information. These head-to-head or stack-rank search queries would need to pull heavily from media coverage and journalists that are knowledgeable about companies and technologies or products within a certain sector, along with information from analyst reports or academic research.

I have checked with AI experts (and ChatGPT itself) and it is clear that journalism outweighs other sources of information on the Internet for such results. Even when the results largely reflect the go-to-market messaging on a company’s website, the order in which it is ranked against others would be valued on relevance, which is highly influenced by news cycles and unbiased media coverage.?

Over my career, I have had dozens of journalists or TV producers call me to interview a client based on a story from years ago that they had dug up during routine research. These are stories long forgotten about, not discussed in our meetings, not mentioned in budget negotiations. That’s the way it’s always been. AI is doing this every day – considering the entire history of messaging, media coverage, analyst reports, social media, etc. to frame and position your company within millions of query possibilities.

Are you ready? Are you adapting? To me, this is a fun new challenge and good news for comms professionals. Let’s go get it!?

P.S. Don’t just take my word for it.?