Danny, a freelancer I’ve worked with over the years, writes sagely on automated journalism – the idea of algorithmic “writers” interpreting standard corporate output (financial statements, press releases etc) and interpreting them automatically:
Narrative Science, a startup in Evanston, Illinois, wants to do just that, with data-intensive stories. Its technology uses natural language algorithms to craft rudimentary news articles about data-intensive subjects, such as sports and financial results.
I find this fascinating as a concept. Of course there are limitations, but given the concerns the industry has about the decline of ‘proper’ journalism – investigative reporting, in-depth analysis – basically anything beyond ‘churnalism’ – and the challenges new media presents in creative storytelling – demanding video, data visualisation and beyond – I hope this concept develops.
In practice, the market for human-written news (even churned) stories will remain, keeping that industry afloat (I think, if they can monetize well enough) – but imagine if a virtual writing assistant helps draw correlations and interesting facts out of a decade worth of financial reports to add some colour to the latest story, or automatically trawls through to get you the aggregate views the Internet has on a product you’ve been sent to test… Would make that industry more efficient and their output more meaningful.
Maybe Public Business should talk to these guys… Although the method of supporting the media is different (investment in training and specific types of journalism vs. creating an algorithmic automator for reporting the news) some of the goals will overlap – in terms of interpreting and presenting data back in a useful way that informs more insightful reporting.