Over 20 of us gathered at our latest 106 Comms breakfast in our Alice in Wonderland themed offices. Navigating through rabbit and teapot furnishings, everyone, no matter the industry, financial services, engineering, law, mining or advertising, arrived with one thing on their minds: AI.

A frenzy of networking that caused a babble of noise to fill the room had to be cut short, as we got down to the nitty gritty of the event.

Firstly, everyone shared their AI challenges, which included implementing Copilot, ethical concerns, frustrations with prompting, and the desire to use AI to unlock siloed teams and share information more widely across organisations.

With our compass and direction set by the crowd, Henry, Founder of 106 Communications kicked things off.

Some 106 AI input

Henry began by touching on the clash of culture and comms with AI – our first episode of our new podcast ‘The Brand and Culture Podcast’ covers this very topic, so if you want to find out more, you know where to go. The rest of the episodes aren’t bad either…!

Next, there was a brief demo of some of our consumer/client personas for the attendees. Imagine being able to have a focus group at your fingertips, or a stakeholder you could just summon out of your pocket. That’s what this is.

We asked our ‘train driver’ persona what would make him read a newsletter. His blunt advice? Make them shorter. Some good, honest feedback and not too much of the politeness filter!

Frank’s journey into AI – learn from history

Frank then shared how he became an AI consultant, starting with a cautionary tale about social media. He noted that when social media first emerged, comms dismissed it as a fad, something that would pass, that comms didn’t need to engage with.

But of course, we’ve seen how spectacularly wrong that was. Social media now dominates not just our lives but the world.

So, after seeing how slow comms had been to react when social media came out, when he came across AI for the first time, he saw something familiar. Something that would grow.

And grow it did – whereas Facebook took just under a year to hit 1 million users (Britannica), ChatGPT took just five days (DataLook.io).

Predicting this trend, that AI was here to stay, he decided to get ahead of it this time, and to help comms professionals learn to use AI well, before it became too big and they were left behind again.

After spending most of his career in financial services, Frank moved to a company in an entirely different field, armed with an early version of Gemini. This was his first step into what he calls ‘the AI sandpit’, giving him crucial insights into how the AI landscape was evolving, and having a place to trial new things, particularly for comms.

AI as transformation

The landscape has shifted dramatically since that point. AI now appears predominantly in transformation roles, closely linked with buzzwords like change, shift, and evolution – it’s not a fringe experimental tool, instead being baked into future strategy as we speak.

We must see transformation as AI and AI as transformation, they’re inextricably linked.

Taking this as a starting point, the important thing now is using experience to guide you.

Have a deep look at your processes to see where AI can automate, not just assist on a surface level, and really transform how you work. Make sure to invest time in learning, having deep knowledge of the uses and limitations of the programmes you’re using is extremely useful. Know your tech stack inside out, learn it, and use it flexibly – also AI will be a continuous learning journey, so don’t let up, there’s always more to learn.

As you integrate AI increasingly into your work, turn that lens on yourself too. What are your limitations? Your knowledge gaps? This can help you identify where AI can plug the holes in your capabilities, and where you need to develop your own skills further.

The sandpit

As Frank noted, the majority of us are still in ‘the sandpit’, playing around with AI, using it for minimal tasks. In many ways this is not surprising. The way AI has been introduced across many organisations is: ‘Here’s a tool, use it. And be productive with it! Bye.’ It’s just been dumped on us like a new toy without instructions.

However, it’s vital we get out of the sandpit and start looking at what we’re doing in a deeper way, and we must look forwards, to the future of AI usage.

Comms has been doing the same tasks in the same way for years and needs to change. AI offers us a new shift and a genuine opportunity to change as a department – to ride the wave rather than be swept away by it.

Three states of AI adoption

Frank outlined three states you can find yourself in when facing AI adoption:

Staying still – not changing at all

Here, Comms will be enveloped by AI. Content production, metrics analysis, sending newsletters, reviewing surveys – all the manual writing work that has historically been the bulk of what comms does is the low-hanging fruit for AI. It’s ripe for the picking and will all soon be automated.

Reading the tea leaves – getting there

This is assessing the current state and reacting to AI transformation on a day-to-day basis. It’s better than staying still but nonetheless is reactive rather than proactive.

Horizons – where you want to be

This is all about constantly thinking about and planning for the future. It’s also one of the best ways to process, consider and manage your AI transformation. AI is moving so quickly that it demands a different way of thinking.

Look to the future and define your destination. Where do you want to be and how do you want to work in an AI future?

Now, set your destination, then reverse engineer the steps to get there. Then you’ll have a clear pathway. This way you’re not just coping with the day-to-day change but moving forward with purpose, using AI to deliver strategically rather than just to complete tasks.

Victim to disruptor

This requires a shift in mindset. A shift from victim to disruptor.

Many of us feel at risk from AI at the moment, but having a victim mentality won’t help. It’s sink or swim, so swimming with it is the way to go.

Here, Frank drew parallels with ‘Planes, Trains and Automobiles’, a film featuring two men. One does the same thing every day, travels home the same way every day – a man of routine. The other is chaos incarnate, trains get cancelled around him, cars set on fire, nothing is ordered.

We are the first man; AI is the second. Something crucial happens in the movie – the ordered man begins to trust the chaotic one, and eventually everything works out.

If you can absorb chaos, trust the process, and accept that we’re going through a big moment of global change, and trust the very tool (AI!) that’s driving this change, you’ll come out stronger with greater knowledge and insight.

Use cases and prompting

We then explored use cases from attendees.

One attendee shared how she used Copilot to analyse pulse survey notes, which gave a more neutral overall sense of sentiment, which was valuable for getting another perspective. However, she’d noticed Copilot tended to stress the strongest opinions, which were often just outliers.

When she asked it to be ‘more neutral’, it returned something completely different.

Frank suggested setting a baseline.

Another attendee jumped in, making the point: What is the baseline, and where is it? If you keep asking AI to adjust, it keeps generating different things. So, what does ‘more neutral’ even mean?

Frank’s reply? This is user error, not machine error. AI doesn’t know what ‘neutrality’ is. We need to articulate clearly, in a way it understands, exactly what we want, down to a tee, often in minute detail.

This is all about prompting, and it’s a new skill we need to learn. Like riding a bike.

Now, this may take some training and reading, but it doesn’t have to be as hard as you think. Who’s one of the best people to teach you how to prompt AI? Well… AI itself.

You can ask AI to help you with prompting. Frank also unveiled a very handy tip he uses. At the end of his prompts, he encourages AI to keep asking him questions about the task until it’s 90-95% sure it understands exactly what it needs to do. This way you can mitigate too much fast and loose generation of ‘AI slop’ and get something back that’s actually useful.

Showing Comms’ value in all this?

A further question then arose. How can Comms show their value as a department, and can they use AI to do this?

There was consensus across the room that a pure strength of the Comms function is knowing how people feel, being human, bringing empathy, care, and that nuance that ensures messages land.

However, the question remains – how does Comms show this? Ultimately Comms is a support service to employees and businesses, so we need to get involved with business partners, with the C-suite, and show our presence.

One suggestion was that Comms can tie itself to other data points, much like marketing does with viewing metrics. An example was given of a company whose frontline workers can upsell products. If their upselling increases by even the tiniest percentage, it could boost the company’s bottom line significantly. Having Comms target this, communicate the essentiality of upselling, then track the effectiveness, is a powerful way of showing clear monetary impact.

It’s all about aligning communications to business motivations – what makes the business tick? What do they want to achieve? Align your communications to that.

Use AI to support this, to automate the more mundane and free yourself up to focus on the strategic.

We’re moving into the era of ‘radical redesign’.