Imagine you had a personal assistant that you can task with sorting out a pile of messy documents, or ploughing through a mountain of spreadsheets to find what you are looking for.
Enter the Quartz AI Studio, a US-based project that helps journalists use machine learning to write better stories.
The initiative, launched in November 2018 with the support of Knight Foundation, is spearheaded by John Keefe, Quartz’s technical architect for bots and machine learning, who previously led the Quartz Bot Studio.
"Many of our bots do include artificial intelligence, and specifically machine-learning, in their operations,” he said.
“But that’s usually to help facilitate conversations — understanding and interpreting natural language, for example. Or learning when is the best time of day to talk to someone.
“But for the machine learning we think can help support reporters, the applications are far more specific. More like: ‘Computer, here are 100 examples of the pattern I’m looking for. Please find more of those this in this pile of 1 million records.’”
The good news is that machine-learning tools available via the platform can help journalists analyse data even if they have no coding or maths skills.
Keefe explained that computers can help journalists with various tasks, such as categorising (for example marking messages as ‘important’ or ‘spam’). Machines are also better at spotting patterns within huge datasets whether these are words, numbers or pictures. Finally, artificial intelligence comes with some capability to make predictions based on what happened in the past.
“Machine learning can help solve these problems and Quartz tries to bring these solutions to smaller newsrooms that often haven't got resources to do this themselves,” said Keefe.
"Many journalists don't even recognise what stories could benefit from machine learning. And when they do, they often wonder how to go about it.”
Quartz AI Studio is planning on running some workshops and training courses in autumn 2019.
If you are not able to attend in person, the publisher has also launched a dedicated Slack bot for journalists, aptly named Quackbot, that can do some handy tasks, such as grabbing screenshots of websites or pointing out clichés. More tools and content will be added to Quackbot as more data-driven stories are published.
“Journalist-programmers are an especially sharing lot,” wrote Keefe to introduce the bot.
“Sure, they’ll work night and day to scoop each other, but once the story’s published they’re happy to share how they did it — even sharing the tools they built. As a result, there are many dozens of useful tools available to programmers in newsrooms everywhere.”
Keefe added that the objective of the project is to share the tools with as many journalists as possible.
Have you got a story that would benefit from machine learning? Get in touch with the Quartz AI team here.
Free daily newsletter
- 'Speculative journalism': why telling fictionalised stories can help us understand reality
- What does content verification mean in the age of deepfakes?
- Five surprising journalism disciplines all students should know about
- Bellingcat launches investigative podcast about the Malaysia Airlines Flight 17 crash
- A journalist's guide to open-source tools