The rapid rise of AI in recent years has affected every aspect of our lives.
There are obvious possibilities for its use, including health care, scientific research And education.
The risks are also all too obvious. Deepfakes and disinformation AI’s contributions dominate headlines, but the potential impacts of AI’s use on society, government, employment and the environment environment are still being understood.
Cosmos has a long history of communicating science to a broad audience. Our science-trained journalists report on the latest research and interview those behind the stories. Cosmos prides itself on fact-based reporting, including on research and use of AI.
We decided to take the plunge and participate in a scientific experiment to answer important questions about the use of AI in the media. Our goal is to report on our experiences using AI in our newsroom and provide evidence-based examples of the risks and opportunities of science communication.
Our first lesson, in hindsight, is that we needed to communicate more clearly about the project before testing and publishing test papers. We could have done better in this regard. We have briefly paused the project while we review the feedback and questions received so far and remain committed to ensuring responsible and ethical AI practices. We appreciate your your feedback, positive and negative, as well as your questions.
What is the Cosmos AI project?
Our archive contains over 16,000 articles written by authors, including current and former staff members, as well as a range of freelancers who have been commissioned to write articles for Cosmos. This content has long been used by our journalists as part of their research for new articles or explanatory pieces such as How could we travel through time? Our archives are a solid source of information that allows us to create more explanations.
Explanations are important because they help our readers understand science and have been popular on our website for many years. They are an educational tool for all ages. They are fundamentally different from news articles and reports.
Our journalists have written explainers in the past, but it takes them away from their mission to cover news and interview researchers. We decided to test whether an AI system could help us create explainers to support our journalists’ work. Explainers should be factually accurate to promote facts over misinformation.
The Cosmos AI system uses a research augmented generation (RAG) model. It first identifies articles in the archive that are relevant to the question or topic. This saves the journalist from having to manually check the archive at the beginning (but there is significant human editorial involvement later in the process – see below). RAG reduces the risk of “hallucinations” where the AI can make things up.
The second step is the writing phase, which involves using a large language model (LLM). Open AI’s GPT-4 (GPT-4) was chosen after researching various platforms. GPT-4 helps with writing explanations. Our story archives are not used to train GPT-4 or any other LLM. We do not use ChatGPT, the publicly available product owned by Open AI that many readers are familiar with.
The last step in our process is the most important. We use trained science communicators to fact-check and edit the first drafts of explainers that were created with the help of the AI system. Nothing is published without at least two real people, including our journalists, reviewing, editing, and finalizing the explainers. We will monitor the time it takes to do this and compare it to a journalist writing and fact-checking the explainer from scratch.
Cosmos has received a grant from the Walkley Foundation’s Meta News Fund 2023 to carry out this project. The project runs from March 2024 to February 2025. The Cosmos AI system was designed by an external developer who has worked with Cosmos for many years. The system is not publicly available, but we will continue to share the results and experiences of our project publicly.
If we are to ask the question “what is AI for in media,” we need to explore how it can be used and interrogate the outcomes. As we cross this boundary, we must build on a strong commitment to ethical and responsible journalism practices.
