Each month we turn the spotlight on new skills and new approaches our team are bringing to their work. This month, we’re discussing the industry’s hottest topic, Generative A.I, with Highwire’s EVP of Digital, James Holland.
How do you use generative A.I. at work?
By following our Framework, of course! I often have ChatGPT, Claude, MidJourney, Adobe Firefly and a couple of other tools open at once. For me, they’ve become a sidekick. I’ll pinball an idea between them, and rapidly come up with a range of options to explore. I’ve also been using MacGPT to bring A.I, into any text field too, so I can use it wherever I can type.
It doesn’t replace human creativity, but I find Generative A.I. has vast potential to quickly catapult past the obvious, trite or derivative - I use it when writing everything from presentations to emails and briefs for the team. It helps me move quickly from knee-jerk reactions into second-order thinking and on to something more substantive. Integrated properly, A.I. makes me a better thinker, and a faster doer. But it doesn’t stop me from being in control.
How can the use of generative A.I. help marketing and communications teams? What are three tips you can offer?
With the caveat that you need to be clear and transparent about your use of A.I. and avoid giving it any information that is confidential, proprietary or protected, I’d say there are three areas where A.I. can give marketing and communications teams an edge:
1. Creativity
The conversational, iterative nature of Generative A.I. is really great for solo brainstorms, punching up a concept in a pinch and adding a touch of flair to otherwise routine work, simply by exploring some alternative thinking very quickly. In a group setting, it can help get all the obvious ideas out in the first instance, so the team can move rapidly to working on less predictable solutions or adding their unique perspectives and lived experiences. It’s an accelerant, and a catalyst, for original thinking but not, in and of itself, a substitute for it.
2. Analysis
At the opposite end of the spectrum, there’s the ability to upload files to ChatGPT or Claude, and get a detailed breakdown of their contents, often in ways which would be impossible or impractical otherwise. I’ve used A.I for everything from tone analysis on an executive’s social media to informing a style guide or byline brief by analyzing a publication’s accepted voice. I’ve dug into coverage at scale, going beyond clip counting and keywords to reverse engineer key messages, topics, third party validators necessary to gain coverage. I’ve even begun to segment individual writers and beats, unearthing the differences which make a particular reporter sit up and take notice, where others might not.
Tools like Claude are great here. It has a larger context window than ChatGPT - meaning it can retain and analyze much more content at once. Some of this analysis wasn’t possible with the first generation of Gen A.I., but we’re already speeding past those limitations.
3. Visualization and diverse thinking
I’m dyslexic, so any tools that help see the world differently are valuable to me. Sometimes I find one that fits the way my brain works naturally, and that helps me to unlock more insights or work more comfortably. Generative A.I. definitely fits into that category.
I use tools like Voices or MacWhisper (which use OpenAI’s Text-To-Speech / Speech-To-Text models) to listen to words, instead of reading them, or vice versa. And did you know ChatGPT can make charts and graphs? You have to double check them, and providing good quality/well organized source data helps, but it’s way faster than Excel for proving a point or exploring some numbers very quickly. I’m looking forward to Google rolling Bard functionality out to Google Sheets.
And then there’s Gen A.I. graphics. I’ve been a MidJourney addict for a long time. Now we’re inundated with options to create A.I. graphics, from Dall-e within ChatGPT to Ideogram and now video with tools like Runway. I find myself using AI to illustrate proposals and pitch decks around 90% of the time now - particularly at early stages of a conversation. There are fewer dangers in using AI this way - with no limits on proprietary data or public use, and while prompting for image creation is a true art in itself, creating a custom image (rather than relying on stock photography) can really help create a mood or belief in the subject being discussed. It’s transformational.
What challenges will marketing and communications teams face in 2024 related to generative A.I. and how can they overcome them?
It feels pointless to predict too much in this field with any certainty, given how fast it’s already moved. With that said, I’m sure 2024 being an election year will bring a new complexity to Generative A.I.’s ability to personalize content at scale.
That’s one area we’ve heard promise, but seen little evidence of change: Truly personalized, customized experiences which are unique down to the individual, and which keep getting more fine-tuned are tantalizing. I’m excited to see this technology make channels like email marketing more powerful than ever (and potentially re-invigorate a whole sector).
Am I nervous about the impact A.I. could have on political or issue-based campaigns? You bet. But there’s never been a more interesting or exciting time to be involved in tech, communications or marketing.
