An illustration of data and AI being held in a ball of light in a man's hand

No, AI isn’t the End for PR and Marketing

Some might suspect this blog has been written using AI. That’s the extent of its current influence – especially on written content – across the many industries that deliver ‘comms’ of some kind.

It’s not difficult to see why. There has been near-constant press coverage since OpenAI launched ChatGPT-4 for public use in November 2022. It can – quite convincingly – draft lawsuits, write sophisticated code and pass standardised tests that require years of study from mere mortals.  

But is it going to replace PR and marketing? I’m not so sure.

The Shock of the New

ChatGPT is a legitimately disruptive technology. But how and where it fits into current and future working practices is a still topic of debate, not least because some of those responsible for its creation are now urging for tighter regulation.

As with any new powerful technology, the conversation around AI has split opinion. Some are feverish with excitement; others are understandably concerned for their job (or the future of humanity, but that’s another conversation altogether).  

By now you’ve probably read about how the tractor replaced the traditional farmhand and the word processor’s hard-fought victory over the typewriter – changes on a level not dissimilar to the advent of natural language processing.

These examples keep cropping up because they serve as good reminders about the nature of work. New tools are there to help us become more efficient, freeing up time for more complex and creative tasks, and even if some jobs are lost, plenty more are created along the way.

ChatGPT is slightly different. It’s clearly quicker than a human at producing copy, and there’s also some element of creativity involved. As the Guardian points out, to be very good at creating text turns out to be “practically similar to being very good at understanding and reasoning about the world”.

From here, it’s easy to understand the disquiet within PR and other professions. If generative AI can interpret events and explain them clearly to others, where does that leave storytellers, journalists and other creatives?

Case closed, right?

Placing Too Much trust in the Machine

As a writer, it’s hard not to sound like a luddite when writing about generative AI, but there are good reasons to place some doubt in its remarkable powers. Take hallucinations. These emerge from the machine’s inherent bias and lack of understanding of the real world. In an effort to produce something useful for the person asking questions, ChatGPT can issue answers that sound plausible but are actually incorrect or totally unrelated.

Why is this a problem? Firstly, hallucinated outputs run the risk of perpetuating harmful stereotypes and spreading misinformation dressed up as fact. Secondly, they may lead some people to make bad decisions in situations where critical knowledge is needed.

These issues are not theoretical. The US Federal Trade Commission, for example, has already issued a warning to companies using AI to manipulate people’s behaviour for commercial gain. This includes the use of chatbots to persuade people into purchases when they ask for what’s assumed to be impartial advice from a third party.

Generative AI is also self-referential. This means its knowledge is derived from what’s already out there, leading to copy that is functional but often uninspiring. Along with this is a hard cut-off date, which sees the machine ‘learn’ from data sets that are often outdated. OpenAI even warns about this on its own website, stating that ChatGPT’s knowledge is limited to events and news that occurred before September 2021, and anything after this point is not guaranteed to be reliable.

The implication for PR and marketing should be clear. Those that rely too heavily on AI run the risk creating unoriginal content that fails to engage with its intended audience. This applies as much to SEO as it does to feature writing, blogs and other more ‘traditional’ forms of communication. In a world with so much noise, why rely on commoditised copy when the reason for producing it in the first place is to stand out from the competition?

The risks are even greater for businesses that see ChatGPT as a handy cost-cutting tool, especially in sectors that require care and good judgement. You could use ChatGPT to begin drafting statements during a crisis but, for the reasons explained above, it’s going to need experience and oversight at some point. If not, those issuing the statement run the risk of making a delicate situation much worse.  

And this is really the key point. AI is clearly an incredible tool that should be adopted in PR and marketing, but its flaws emphasise why businesses seek out guidance from experts and consultants who know their industry best. Businesses ultimately sell to people, not machines.

About the author: Content director Tom Went is a content specialist in B2B PR and publishing. He turns technical subject matter into accessible campaign content that helps to generate leads and position clients as industry thought leaders across WPR’s B2B portfolio.

Want to
know more?

Contact us