neuron-cell-system

Grappling with AI: staying true, accountable and human

Our guiding belief remains simple: business should amplify humanity, rather than diminish it.

AI tests that belief in small ways every day. It can make almost anything sound finished: confident, tidy and plausible, whether or not it’s true, useful or ours. Then it lands in seconds, looks done and gets waved through, with no one stopping to ask “is it good?” or “what is good?”

Over the past year or so, our team have been grappling with what AI’s arrival means for our role as leaders and creators. We’re also watching how it’s being used in brands, communications and organisations. Both raise questions that don’t fit into one neat answer:

  • When we use AI, what are we asking it to optimise for?
  • Who decides whether the output sounds like us, reflects our values or is right for the people we are trying to reach?
  • And if something looks polished, how do we know it’s actually useful and truthful?
  • What adds real value, and what just adds to the noise?
  • What helps us be more human, and what takes something away?
  • How do we develop the judgement and ethical discernment AI now asks of us, especially as its use scales?

Perhaps you’re having some of the same conversations in your own organisation.

We’re thinking out loud here, and we invite you into these questions in the hope that however you end up using AI, you choose intentionally and choose well.

Looking for the story behind the signals

Every organisation is surrounded by information.

Customer feedback. Team conversations. Website behaviour. Search patterns. Market shifts. The emails people send when something has gone wrong, and the words they use when something has mattered more than they expected.

AI can help bring those signals into view. It can notice recurring themes across hundreds of comments. It can spot inconsistencies in a message. It can help a team make sense of a body of information that would otherwise take weeks to work through.

That is useful. But a pattern is not yet an insight.

For example, a spike in website exits cannot tell you what a person felt. A drop in engagement cannot explain whether your audience is losing interest, or whether your message has stopped sounding like you. A report may identify what people are clicking, without telling you what they need.

That is where the human work begins.

We listen for the story behind the signal. We look for the tension beneath the trend. We ask what the information means in the lived reality of the organisation, its people and the communities around it.

The answer is rarely jumping to start a new campaign. Sometimes it is a harder conversation about a promise the organisation is not keeping. Sometimes it is an opportunity to make something clearer, kinder or more useful.

The question is not only what you ask AI

A good prompt matters, yes, of course it does… The question should be shaped by what you understand about your organisation, your audience, your culture and the decision you are trying to make. A vague question tends to invite a vague answer. A prompt that ignores people, context and values can give you something efficient, but not necessarily something wise.

But good use of AI does not end when the prompt is written. The real test often comes afterwards.

We’ve all been there, when the answer arrives from our AI, particularly when it’s outside of our field it sounds really plausible. It is neatly formatted and reassuringly certain. It may even be very good.

Then someone needs to ask: is it true here?

Does it sound like us, or has it borrowed a voice that could belong to anyone? What has it assumed? Whose experience is missing? Does this advice fit the relationships we are trying to build? Would we be comfortable standing behind this decision when the people affected by it are in the room?

That is discernment. It cannot be reduced to a checklist, because it depends on the people, the stakes and the context in front of you.

Before turning to AI, it can help to ask:

  • What are we really trying to solve?
  • What would a good outcome look like for people, not only for a metric?
  • Whose perspective might not be represented in what we have supplied?
  • What values need to guide this decision?
  • What needs to remain a human judgement, rather than an automated recommendation?

And once the response is in front of you:

  • Is this accurate enough to act on?
  • Does it sound and feel like our organisation?
  • What has been simplified, flattened or left out?
  • Is it useful, or merely polished?
  • What needs a conversation before we move forward?
  • If we acted on this, what would people experience differently?

The slow drift towards sameness

The risk with AI is not always obvious.

It is not necessarily a glaring error, a strange image or a wildly inappropriate response. It can often be really subtle. It can be found in a piece of copy that is “fine” or a campaign line that sounds right because it sounds like everything else in the category. Safe (but ordinary) reassurance.

Then another. And another.

Before long, a brand that once had a point of view begins to speak in the safe, polished language of everyone else. Its personality is softened. Its edges are rounded off. Its values are mentioned, but no longer felt. Sadly, we’ve seen it happen too often: brands that had a voice, a purpose, something they believed in, reduced to ordinary.

This is just one way brand drift can happen (we have a whole article that dives deep into this).

Drift is rarely through one reckless decision, but through a series of reasonable compromises made by busy people trying to get good work out the door.

AI can accelerate that drift if we let it become the main source of the work. It is trained to recognise what is familiar. It is good at finding the middle of the road. But a brand is not here to blend in with the category.

A brand needs to do more than meet category expectations. It needs to express its own point of view, personality and history. It needs to respond to the real needs of the people it serves. And it needs to hold the bigger picture: the values it is trying to live by, the relationships it wants to build and the contribution it hopes to make. Most of all, your brand needs to be authentic to you.

A brand is a promise people learn to recognise over time. It lives in the language you choose, the experiences you create, the choices you make when there is pressure to take an easier path. It is shaped by what you believe, what you say and what you do.

Those things need to point in the same direction.

A moment to reflect

As AI becomes more familiar in the way we work, it may be worth returning to a few simple questions:

  • Is this recognisably us, or could it come from anyone in our category?
  • Are we treating a confident output as a finished answer?
  • What assumptions have we inherited without noticing?
  • Does this strengthen the connection between what we believe, what we say and what people experience?
  • Are we creating more noise, or more understanding?
  • Are we using the time AI saves to create more content, or to make better decisions?

None of these questions are anti-technology. They are questions about responsibility.

They help ensure efficiency does not become an excuse for disconnection. That speed does not replace care. That a well-formed answer does not stop us from asking the better question.

Staying curious, staying accountable

We don’t need to turn away from AI, but we do need to think carefully about how we use it. That means staying curious about what it can make possible, while remaining clear-eyed about its limitations.

To keep testing what it can make possible. To be honest about where it is useful. To be careful where the stakes are high. To make room for the human work that cannot be automated: noticing nuance, holding tension, recognising harm, weighing competing needs and deciding what is fair. And most importantly, to keep asking discerning questions as the technology evolves.

For us, that means using AI to support our research and insights, and to create new efficiencies, while keeping relationships, people, our values and the purpose in our work at the centre. It means being open about the fact that our approach will continue to evolve. And it means remaining accountable to the belief that guided us before AI became part of the conversation.

Business should amplify humanity, not diminish it.

The question for all of us is how that belief becomes visible in the choices we make next.

A place to begin

If you have questions about bringing your brand to life, a Brand Check-in is a good place to start.

It’s a low-key conversation about what is working, where your brand feels clear or unclear, where you’re headed and what it needs to remain true to as AI becomes part of the work. Reach out, we’d love to inspire and guide you on your journey.

Author
Jess Aquilina - Mezzanine Team
Jess Johnson
Brand and Communications Strategist

Sign up to our 'Brand Insights' email

We know how hard it can be to stay aligned, consistent and strategic when managing your brand day-to-day. We provide updates, tips and advice straight to your inbox to help you ensure your brand builds long-term value for your organisation and has a positive impact.

It starts here!

The best way to get started is to complete these questions to help us better understand your organisations needs. We look forward to meeting and working with you.