Finally. Two weeks into 2020. I’ve finished the New Year predictions about the future of journalism.
The good news is the almost unanimous agreement on business models which require deeper engagement with the communities journalists serve. What was missing was the same strategic clarity about the technology that will make or break engaged journalism.
The broad spectrum of innovation we call artificial intelligence is playing a role in decentralising the information economy from the network to the individual. It turbo-charges the shift from social media to personal media, where people have come to expect digital experiences they can customise, curate and privatise.
For the past two years, Kinzen has been trying to figure out how a shift towards personalisation could help enhance journalism’s communal purpose.
Using lessons learned in the development of a news curation app, we developed a software product with a uniquely purpose-driven approach to artificial intelligence. Now our partners include broadcasters, regional newspaper groups and aggregators with remarkably similar pain points. We’ve been helping build personalised experiences which empower the user and make content smarter and more discoverable. Our goal is to become standard kit for any news organisation generating revenue through deep personal engagement and trust.
We’re enabled by technology, but we’re also well placed to see the danger ahead. Even in the most optimistic scenario, a force like machine learning will be deeply disruptive. Charlie Beckett summed it up well in 2019. “If we value journalism as a social good,” he said in a landmark report on artificial intelligence by Polis at LSE, “then we have a window of perhaps 2–5 years, when news organisations must get across this technology.”
So, as a chaser to that long list of New Year aspirations, we offer up some lessons we’ve learned about new rules of engaged journalism in an AI world:
Journalism needs to take back control of its relationship with citizens from the social platforms. That means news publishers will need to become technology companies in their own right, building personalised user experiences with their own content, on their own platforms.
In its forecast for 2020, the Reuters Institute for Journalism found more than half of news publishers are making AI initiatives a key priority, but “smaller publishers worry that they could be left behind due to the complexity, expense, and scarcity of skills”. That’s where Kinzen can help, by delivering a single affordable software solution for any publisher who wants to use AI to drive personal engagement.
AI can help journalism rebuild daily routines that were destroyed by endless scrolling newsfeeds and 24/7 alerts. Every internet citizen now expects news to be aligned with the rhythm of their day, as they move through locations, moods and needs (Sarah Marshall has some great pointers on such ‘news moments’). Done badly, personalisation of news can detach an individual from shared reality. Done well, it becomes an essential tool for active citizens, seamlessly synching their daily lives to the multiple communities they belong to.
The first generation of personalisation technologies were creepy and opaque largely because they were built to meet the needs of advertising platforms. But with a focus on new revenue sources based on deep engagement, we expect to see a shift from an exclusive reliance on behavioural tracking to ‘personalisation with a purpose’. The focus of this model is empowering the individual to express their intentional preferences through interactive content experiences. The user is given agency over outcomes. Trust is earned. Loyalty is built. The interactive user experiences Kinzen is pioneering are a brand new feedback loop for publishers, and a brand new source of data to inform editorial and commercial strategy.
AI does not replace journalism. It expands its scope, creating a need for new editorial roles. At Kinzen, we’ve given publishers the ability to shape the editorial elements of the user experience, and apply a layer of human curation to the natural language processing. We are experimenting with user experiences that engineer serendipity and burst filter bubbles. For example, a newsletter format that would help users mindfully monitor their information routines, and provide a role for ‘editor’s picks’ to balance the diet. For more inspiration on AI-enabled journalism, check out this wonderful piece by Sarah Schmalbach of the Lenfest Local Lab.
Journalists tend to underestimate the unsexy plumbing work their organisation needs to do to integrate a hot technology. Before executing an AI strategy, newsrooms need to fix the plumbing of their content system. For the user to be involved in personalisation they have to be able to easily navigate through the topics and people that interest them. But most news organisations struggle to structure and tag their content in a way that maximises user interaction and engagement. Through advances in NLP and content clustering (and a simple software integration), we can radically improve the chances users will build routines around the right content. For more on how AI can make your content smarter, watch this this great presentation by Culture Trip’s Ana Jakimovska.
A newfound respect for the sanctity of personal data could mean the end of third party cookies. That’s both an opportunity and headache for news publishers, adding urgency to the shift towards a business model which relies on deeper engagement (we love this excellent primer by Bloomberg’s M. Scott Havens). Personalisation is a rich new source of first party data for publishers. And richer still are the signals from interactive content experiences powered by purpose-driven personalisation.
I’ll leave you with a final challenge that intrigues us at Kinzen. Not every media company produces the variety of quality content that meets the wide range of interests in a personalised newsfeed. Publishers will increasingly need to consider the role of quality third-party content in their digital experiences. If publishers are trying to engage the Netflix generation, they have to get used to curating content that is not exclusively their own, but does meet their own high standards (check out the work of the Distributed Media Lab for more inspiration).
This may well be one of the most significant impacts of AI on journalism, to turn each publisher into an aggregator, the anchor in a bundle of personalised content. As Charlie Beckett observes, the journalist should not expect AI “to re-establish its preeminence as the information gatekeeper. But instead to find new ways of being the creator and curator of credible information”.
When I think of what journalism can gain from artificial intelligence, I’m reminded of that old Buzzcocks line about “nostalgia for an age yet to come”. If we choose to engage with confidence and clarity, AI is the road to the future that leads us back to our original purpose.