In November 2022, a new era began when Open AI made its ChatGPT tool available to the general public. We spoke to Sandhya, Creative Consultant at Habegger, about her experiences and insights into how her working life has changed.
In the Creation team, Sandhya is responsible for developing the concepts and ideas for client projects, understanding the client’s needs, translating them into a clear story and, finally, designing the event content based on this. Sandhya has been working with the tool since the early stages of ChatGPT.
How do you work with ChatGPT?
Sandhya: I use ChatGPT in different ways, depending on the phase. When analysing needs, it is like an extended research tool. We can generate a large number of ideas during brainstorming sessions, as the AI assistant can now think for itself. However, clients never receive ChatGPT concepts from us; it’s just part of the process. I am ultimately the one who develops the ideas further so that they fit the project and the needs of the target group.
What about data protection?
That is still a grey area. In-house, we have a ChatGPT working group that creates instructions for the AI application and updates rules on an ongoing basis. The team also keeps abreast of the latest topics and is always at the cutting edge of technology by continuously experimenting with new approaches. We all benefit from the expertise we have built up through internal training courses, which also ensures the transfer of knowledge.
Our guidelines take into account values, cultures and interpersonal aspects that must be incorporated into our work. We also ensure that the tone of voice, language level and our standards are always in harmony.
Your lessons learned & experiences?
In my opinion, AI will be indispensable in the coming years because it offers powerful application possibilities. It is simple and practical, new ways of working can be created, saving time. However, there is also a danger of getting lost in it.
I can now create unique mood boards with ingenious visuals that don’t yet exist, which gives me new food for thought. The AI assistant is good for making a rough draft, but not suitable for the final draft. The more specific the requirements, the less accurate the result.
The biggest challenges?
Where ChatGPT ends, our real work begins. It is merely an aid. When creating content, I always have to check whether the statements are correct, whether there are studies or high-quality sources.
The use of images with people remains a challenge. The faces were initially extremely artificial, but the quality has now improved, although it is still not perfect. ChatGPT is constantly learning and developing, but it still requires a lot of personal commitment and work to achieve optimum results.
Creating storyboards with an AI tool is also challenging, as the image sequences are not always seamless. A draughtsman, on the other hand, can sketch them effortlessly on paper.
The biggest fails?
The biggest fails occur when a picture appears perfect, but there is a tiny detail that isn’t right and cannot be easily fixed with ChatGPT. Photoshop is needed in cases like these, and the manual adjustments can be very time-consuming.
A look into the future – what’s next?
In the future, there will be a stronger focus on the creation of moving images. This area has not yet become optimally established. Nevertheless, demand is high, which is why I am firmly convinced that the next step in development will be from static images to moving images.
The quality of AI-generated content will also continue to improve, making it increasingly difficult to tell the difference between AI-generated and non-AI-generated content. The pressure for social responsibility will continue to increase to ensure that the tools are used correctly in every respect.
Subscribe to our newsletter and follow us on social media.
