10.05.2023 at 11:00 AM

ChatGPT in newsrooms

How to achieve peak performance with Prompts & Fine Tuning

Artificial intelligence is fundamentally changing the world of work, and journalism is no exception. With the advent of ChatGPT and other large language models (LLMs), the question is how these advanced technologies can be optimally used. How can ChatGPT make the newsroom more efficient and optimize journalistic work? Our webinar gives you exactly the answers you're looking for!

Use prompts and fine-tuning for better editorial content

We explain prompts and fine-tuning using practical examples. A prompt is an instruction for an AI such as ChatGPT. An example: “Write a short article about the benefits of electric cars.” The quality of AI results depends heavily on the prompts selected. We'll show you the key tricks.

Fine tuning, on the other hand, refers to the process of specifically adapting an LLM such as ChatGPT for specific applications. For example, a specialized ChatGPT model can be fine-tuned for financial news in such a way that it generates more precise and relevant information on financial topics. In this way, you increase the quality of the texts you create and meet the requirements of your newsroom.

Our webinar provides you with in-depth insights into:

  • Basics of effective prompts for first-class ChatGPT results
  • Targeted fine-tuning: ChatGPT tailor-made for your publication
  • The potential and limits of LLMs in everyday journalism

Peter Dyllick-Brenzinger
Head of Product and Engineering at SPRYLAB
Peter Dyllick-Brenzinger has been working with and on content management systems for over 15 years. Prior to his current role at Purple as Head of Product and Engineering, he was head of product for internal headless CMS and editorial AI development at Axel Springer for BILD and WELT.

Interested in digital-first content creation and publishing? Please contact us.


Discuss your use case with our experts and have Purple shown to you in person - book an appointment directly via Calendly.

Or write to us briefly via the contact form and we will get back to you as soon as possible:

Thank you We'll get back to you as soon as possible.
Oops! That didn't work, please try again.