At Liip, we actively address the opportunities and challenges of AI-based applications. We do this internally (read more in this blog post), but above all in projects and labs with our customers and partners. The openness to embrace and explore new technologies, while looking for real added value, is important to us.

Sharing these values ​​with others was the motivation for the training “AI in editorial workflows”.

We considered editorial teams broadly: We addressed small and medium-sized editorial teams and profiles from organizational communications. In a diverse group, participants can learn not only from us, but also from each other.

The pilot: 2 mornings

The first edition of the training took place in spring. We decided on the following setup:

  • 2 mornings of 3.5 hours each on site (Liip Office Zurich).
  • A break between the 2 sessions of 3 weeks so that what has been learned can sink in and questions can arise.
  • Supplemented by tips and tasks for self-study.

We realized that 3.5 hours of face-to-face time per session is short - at the same time we wanted to keep the face-to-face time to a minimum and the costs for the training to a minimum. Looking back, we would have liked to have had more time to work on topics together, so we will rethink this aspect.

The range of topics: broad

As we did not require any prior knowledge for the training, we wanted to offer a well-founded introduction and an overview. We combined input from different Liipers with applied exercises and group work. We covered the following topics:

  • Background and development of artificial intelligence, technical context
  • Different forms of representation, generating various content yourself
  • Focus on AI and text processing, ChatGPT applications
  • Prompt writing
  • Ethical aspects and editorial guidelines
  • The legal situation when dealing with AI

Through group work, we explored the following key points in more detail:

  • AI-based tools in the editorial workflow: In a first step, we went through the editorial work process and openly discussed where AI-based applications offer potential. In a second step, we critically reflected on these points: Where is the benefit and effort in a sensible ratio? What can we hand over to the system, but perhaps don't want to for work ethics reasons?
  • My team is getting AI-fit! AI is changing our everyday lives similarly to how digitalization once did. We have to face the challenges and want to use the potential for the editorial team. How do we make our team fit to deal with AI? How can we sensitize our colleagues to the risks and opportunities of AI? Where do we need offers and guidelines? Together, we exchanged ideas and discussed challenges.
  • The AI ​​tool of my dreams: Artificial intelligence is revolutionizing editorial work and offers great potential if the tools don't get in the way. How do we find the ideal tool that promotes our creativity and doesn't inhibit it? What functions should an AI tool have to support our editorial team and what specific requirements do our employees have? A small group has developed an initial sketch for an editorial tool.

Sharing resources and tips

A key aspect of the training was connecting with like-minded people and learning from each other. This included sharing helpful resources with each other.

Here are 2 resources we worked with:

Outlook

At the end of the training, all participants tried to define a realistic next step for themselves in order to stay on top of the topic. We also decided to stay in touch as a group for further exchange on the topics.

At Liip, we were pleased with the positive feedback from the participants. We are considering offering the training again in a further developed form in a few months. Anyone interested can contact Stephanie Grubenmann or Stefan Huber.

At Liip, we look at the potential of AI-based tools beyond editorial workflows. In the best case, we can work on this together with our clients - in accordance with one of our guiding principles, "Practice over Theory". In our blog – lex4youGPT – you can find a description of such an implementation. Or here we describe how we proceed in this work: AI @Liip: from initial trials to client projects.