Generative AI is a great gift for our creatives and art directors. We use artificial intelligence for mood boards, image and video editing, image creation, voice adaptations, voice generation and optimization and multi-channel content creation, among other things.
AI as an enabler and efficiency driver
AI makes our processes much more efficient and enables us to achieve much more within the same budget. On the one hand, our clients benefit from the fact that we can act in less time and with leaner budgets and, on the other, from the fact that unique campaigns and designs are created that are only made possible by AI.
Image creation with artificial intelligence
As a technology pioneer, we already generated images with AI in May 2022. The basis for this was open source software from Open AI, which we installed on our server and fed with input. For example, we were able to create visuals for a software conference that represent the various stages of a mountain range to be climbed and then play them out in different styles and angles at the click of a mouse.
A few weeks later, Midjourney came onto the market and has developed rapidly. A light fantasy look was typical of the pictures, which was only suitable for selected projects. Nevertheless, Midjourney was great for us, because we were able to create moodboards for design projects or storyboards for films much faster and also build a funny meme without much effort… Prompting has become a new competence for us.
You could read about new image generators almost every day and each one had different strengths. We have tried many of these text to picture applications and were particularly impressed by Firefly. In our view, the beta version of Firefly has led to much more realistic results than the current commercial Firefly Image 2. Fun fact: in our AI workshops, we play the game “Fake or Real?” with our customers and if we’ve done a good job of prompting, most participants are unable to recognize which images are real and which are generated with AI.
Create vector graphics with AI
Even vector graphics can already be created with AI and recolored at the click of a mouse. Most of our customers have a specific illustration style that they want their new illustrations to reflect. But even that is no problem for AI systems. In addition to the textual description of what should be shown on the illustration, we can load sample illustrations into the system and predefine colors and styles. The results are good templates for our designers, which are then adapted and redrawn.
Generative AI in image processing
We use artificial intelligence by far the most in the area of image and video editing. We expand images so that they are suitable for different formats, delete elements from images or replace them by describing in text what should be shown instead. The AI program generates several suggestions for the extension or the new objects to be inserted based on our text input. This AI-supported image processing speeds up our work considerably and enables complex image processing even for small projects. Images can thus be used in a wide variety of formats and flexibly redesigned.
In addition, we can use artificial intelligence to optimize the resolution of images so that images can be used larger or are sharper.
Video editing and generation with artificial intelligence
Editing and cutting films is much easier with AI, as the spoken word is automatically transcribed as text and changes in the text are transferred to the cut. This allows you to change the length and order of the clips with copy-paste in the transcript or eliminate unwanted pauses and filler words text-based. In addition, we can also expand the video image section or replace elements in the video in video scenes with constant camera settings by combining AI-supported image expansion and video elements.
Video generation with artificial intelligence is still a small niche, but will certainly be very interesting in the future. We have already brought images to life with AI or converted text to video applications for our clients’ social media projects.
Voice generation and optimization with AI
We optimize voices with AI almost every day. For example in podcasts or videos. The artificial intelligence removes unwanted background noise and optimizes the voice so that the recording has a significantly better sound quality. Another interesting application for our customers is voice generation with AI. Text can be generated into voices in all possible accents, languages and intonations. Your own voice can also be cloned and then translated into any language. We see three particular areas of potential for companies here:
On the one hand, a corporate voice can be created that can then also be used across national borders. On the other hand, videos or audio content can be played in different voices depending on preference. Or the voice of a person – for example the salesperson or CEO – could be cloned and then played out in different languages, for example. Data protection and ethical issues have certainly been the biggest stumbling blocks so far.
AI consulting and workshops
If you would like to gain deeper insights into our AI processes or develop a strategy for dealing with generative AI for your company, please feel free to contact us. We advise companies on the integration of AI into their corporate strategy and design workshops on generative AI.
