Artificial intelligence (AI) generators have taken the world by storm. AI has been part of our Hemisphere toolkit for a while but it’s great to see it level up — moving beyond simple automation and testing.
We’re increasingly using AI to collect and use data about consumers — their interests, their demographics and their shopping habits. We can then use AI to customise messaging for each individual.
AI supports ambidextrous thinking
At Hemisphere, we’ve made use of AI along with virtual and augmented reality for different campaigns. But it still needs the strategic creative lens that only comes from human experience and expertise. We’re confident there will be even more opportunities with AI as it evolves further.
Creating data and insight led advertising means being able to make the best use of data. Innovative AI tools are helping us analyse data faster and make robust predictions — and quickly measure the effectiveness of our work.
Hemisphere Media has a fantastic track record of targeting consumers when and where they’re most receptive to our messages. AI is supporting us to do this more effectively.
Good data also informs our creative approach. We can greater understand our target audience, create more personalised messages, and quickly flesh out creative concepts on a wider scale.
It’s good, but it’s not there yet
AI can be used to write copy, but it currently lacks a real New Zealand flavour. And yes, it makes mistakes. Our managing director was devasted to discover Google Bard thinks he’s a Taylor Swift fan (he’s in the minority in our Wellington office).
With access to a vast amount of information and data on an unlimited number of topics, AI is also great for researching. Of course, there are some limitations. As we know, not all information on the web is accurate, but AI can be extremely helpful for establishing a framework or springboard for your work.
While AI makes a fantastic proofreader, it doesn’t cope well with our use of te reo Māori!
Good copywriting is based on a solid understanding of the audience — AI, even when we provide examples, isn’t ready for unique New Zealand audiences. AI draws on the genericness of the web and while we can train it, more work is needed. And drawing on the web means drawing on the biases of the content that’s out there.
Garbage in, garbage out
It’s great to see ChatGPT-4 and Google Bard bringing AI to the live web and pushing the boundaries of what AI can achieve.
But to use the classic coding phrase — GIGO, garbage in, garbage out — even AI can only work with what it’s given. We all need to learn prompt engineering, to develop the skills to get the best out of AI generators.
Our favourite AI tools (at the moment)
Whether you need help formatting a spreadsheet, or some inspiration to write that article, here are some of our favourite AI generators to help you with your creativity and productivity.
Our views are likely to change as the various tools evolve. We’re particularly excited by the evolution of the Adobe Creative Suite tools.
Know the limitations
AI is a fantastic tool, but there is much to consider before jumping on board. Make sure you understand the pros and cons when you’re making use of AI:
Pros:
Cons:
AI generators can be of huge value. They can support content generation, work productivity, business operations, and more. At Hemisphere we embrace AI and the way it can enhance our mahi, but we’re aware that it has its limits.
We believe AI can only augment and amplify human capabilities, not replace them. Human creativity, empathy, and logic are essential for creating compelling and engaging work that resonates with audiences.
We encourage you to try out AI, to see how it inspires and elevates your work. With a range of AI tools and services available, you’re bound to find one that fits your needs — but don’t be too caught up on its ability to produce whole pieces of work from scratch.
The rising power of AI lies in its ability to enhance your results, not replace you.
The image at the head of this article was created using the Adobe Firefly text to image creator (which is not yet released for commercial use). However, it struggled to cleanly combine the two sides of the brain so the final image required human input and is two AI generated images combined.