Part 2. Workshop: Image Creation With AI
This blog post is the second part of the workshop report on the “Future of Marketing” conference. In this article, we will dive deeper into image creation with AI. You will learn more about how you can use already existing image creation AI tools. If you missed the first part of the workshop report, you can find it here.
Let’s have a look at tools that create different types of images with the help of AI.
1. Logo creation with AI
An interesting service for the generation of logos is the Looka service. You choose the designs, colours, and lettering you like and within a few minutes, the tool generates a logo based on AI. For start-ups, it’s a fast alternative. For companies that already have a successful standing in the market, the service seems a bit simplistic to us, especially since the topic of logos deals with further questions such as branding and positioning. But of course, these topics could also be addressed on the basis of AI in the medium term and perhaps even work great, as AI can analyze and process different dimensions.
2. Face creation on the basis of AI
An interesting service for the creation of faces based on AI is This person does not exist. Here you can create an infinite number of faces that do not exist in reality. The service is exciting because the naked eye typically cannot discern that the faces do not exist in reality. How exactly this works can be read here. The question remains, however, for what purpose one would use these services beyond illegal or deceptive intentions. We cannot immediately think of many reasonable purposes that are not based on deliberate deception. For the workshop, however, it was an exciting experience to see what AI is capable of.
3. Picture composing
With the Icons8 -tool, you can create different images completely masked with AI to a composition. This is an interesting service that can be used to create halfway usable image compositions.
Aswell worth to mention a direction where the image creation is moving to is Image generation based on provided text.
Runwayml is experimenting already in this direction. This service is in fact still in the experimental phase. Hardly usable for creating useful content. If at all, then rather in the field of art to settle.
The same applies to the tool called Dreamscope. It creates and filters for uploaded images. These types of amazing images we’ve seen already on Instagram and possibly on other social media channels as well.
In the field of image processing, artificial intelligence already brings together some skills. Nevertheless, we are still in the experimental phase with the tested tools, and their value in regard to the practical use of content creation is limited. But one should not underestimate how fast the development phase progresses here.
Just think of moving cars that are able to process hundreds of thousands of visual impressions every second and convert the three-dimensional models.
As already mentioned, the results of the text tools look somewhat better. In particular, a project from the United States that has shown fake news can be created on a large scale in the shortest possible time using artificial intelligence should be mentioned here.
This means that soon we won’t be able to be sure texts were actually written by humans. Artificial intelligence is already being used today, especially for structured data such as weather reports and sports reports. The same applies to texts used for online shops, i.e., structured texts such as product descriptions.