To many people, AI is scary. Hardly surprising, when we were brought up on movies like The Terminator and 2001: A Space Odyssey. The media helps feed the narrative that AI is sinister and out to compete with us. It’s more dramatic that way, after all.
But this isn’t particularly helpful, or accurate.
Rather than taking our jobs, there are many ways in which AI is helping us become better at them. For example, take voice-activated robots that help carers lift patients. The machine does the brute physical work, allowing the carer do more caring. The social, empathetic work is what only humans can do. Or chatbots that help doctors reach diagnoses faster. They enable more patients to be seen and in less time and reducing the strain on health services.
Journalism is no different. The last couple of years have seen AI make life easier for journalists in exciting ways, and broaden the scope of what they can cover. Here are a few:
Few journalists look forward to transcribing interviews: to most, it’s a chore and waste of time. But it needn’t be any more, thanks to AI-powered transcription tools like Otter, Trint or Recordly. Not only can these tools transcribe voice recordings more or less accurately in a matter of minutes, but the resulting text is time-stamped to the recording. This means you can isolate a specific part of the audio without having spend time skipping back and forth to find it. Furthermore, it makes it easy to search interviews for keywords: all ways to save journalists’ precious time.
2. Data journalism
Stories based on data and statistics have become increasingly prominent in recent years. And that trend is set to continue thanks to machine learning. For the Panama Papers story, the International Consortium of Investigative Journalists used various machine learning tools to sift through and classify millions of documents, a task which would have been impossible on their own. The team used data transformation and analysis software, a clustering tool and an optical character recognition tool developed specially for the purpose.
3. Satellite photo analysis
Applying machine learning to satellite data can help journalists keep track of activities which previously went unnoticed or were poorly understood. For example, AI has shone a light on illegal amber mining in Ukraine and transshipment — a practice often associated with illegal fishing catches, drug smuggling or forced labour.
4. Image and video recognition
Computer vision tools like Clarifai and Vidrovr scan images and video, allowing journalists to search them for people or things. For example, a journalist could use Clarifai to identify politicians in a group photo.
5. Suggesting story tips
Reuters has been building using a tool called Lynx Insight which analyses large datasets to identify things which could be newsworthy — eg. a sudden change in share price — and then inform its journalists accordingly. Swedish outfit Newsworthy does something similar.
This a quickly evolving field: expect to see many more complementary AI applications in journalism over the course of 2019.
Loyal is part of this new wave. Our tool can help inspire and inform journalists as they write articles, updating them about breaking news or suggesting relevant social media content — enabling them to do more writing and less tiresome searching as they go.
So next time your hear about the robots taking our jobs, remember that’s only part of the story. AI can work with us too.