Interview with Mattia Peretti, manager of the project ‘Journalism AI’ at the London School of Economics and Political Science (LSE). The project is funded by the Google News Initiative and started in April 2019.
What is the aim of the project ‘AI in journalism’?
From what you have seen in your own research until now, how can journalism benefit from AI?
Mattia Peretti: AI can free journalists from some of the most trivial and boring tasks. Think about transcribing the audio recording of an interview. Think about writing a short, factual news article that is not much more than filling in a template with new data. This can already be automated. Human journalists can save this time and do something more creative and useful.
How are you doing the research in the project?
Mattia Peretti: I do the project together with project leader professor Charlie Beckett and one research assistant. We are not AI-experts ourselves, but we are going to research how AI is already used in the newsroom. We have developed a survey that is being sent out to a selected number of newsrooms around the world. We ask them questions like: What are the consequences for the editorial process? What are the financial consequences for newsrooms? We will look for the best case studies and later share them with newsrooms that have not started yet.
When can we expect the results?
Mattia Peretti: The aim is to present the conclusions of the survey in an online-report in the fall of 2019. We hope to develop ideas and guidelines that will help newsrooms to think about AI. We will put a specific emphasis on the ethical side. How can we make sure that AI in the newsroom is implemented properly and ethically?
For AI to be successful in journalism there should not just be attention for AI but also for the interaction between humans and machines. What is your take on that?
Mattia Peretti: Yes, that is a key area for our research as well. Newsrooms are very much aware that AI is going to change the work of journalists. They are trying to take a more holistic approach and not just look at the AI. For a good cooperation between humans and machines new training needs to be developed. We will also look this year at what training is needed. Like many other sectors, journalism can’t escape the influence of AI-technologies. Let’s then find the best way to get a positive impact.