ANTicle
ANTicle revolutionizes content creation in digital newsrooms by providing journalists with high-quality texts, headlines, and SEO content, allowing them more space for research and creativity.
ANT (AI Newsroom Tool)
ANTicle is evolving the way online media creates content. While the necessity to swiftly publish news dominates the operation of digital newsrooms, journalists often have limited time for research and creativity. This AI-based tool, specifically designed for newsrooms, generates drafts for high-quality journalistic articles, headlines, SEO titles, and texts.
The Open Source Solution
Problem
In newsrooms, every second counts, especially during breaking news. Editors must quickly transform news agency reports into articles to attract readership without neglecting thorough research. In times of wars and global crises, it is crucial for journalists to have enough time to verify photos, consult local sources, and obtain statements from regional actors. This often requires delegating some of their current tasks.
Idea
KI-Systeme wie Chat-GPT haben gezeigt, dass sie in menschenähnlicher Weise schreiben können, vorausgesetzt, sie erhalten präzise Anweisungen. Journalist:innen nutzen diese Sprachmodelle jedoch nicht direkt für ihre Arbeit, da der Prozess zur Texterstellung mehrere zeitaufwendige Schritte erfordert. Die Lösung besteht darin, ein Tool zu verwenden, das diese Modelle für Journalist:innen einsetzt. Diese geben lediglich den Anfangsinput und nutzen das schnelle und unsichtbare Ergebnis, wodurch sie wertvolle Zeit für Recherchen gewinnen können.
Solution
The solution is remarkably straightforward. Users utilize a (web-based) interface, inputting their data consisting of agency reports, social media posts, etc., and click "Go." The information is then sent to the chosen Large Language Model (LLM), the results are generated, iteratively refined through several loops, and subsequently presented back to the users.
Fellows of the project
Simon Uhl
Simon Uhl is a data journalist, developer, and lecturer. In his lectures and work, he focuses particularly on the intersection between editorial work and computer science. The central question guiding his efforts is: What could the future of journalism look like?