Technology Solutions for Newsroom Problems: GENSummit 2018
I was fortunate enough to be able to attend the Global Editors Network Summit 2018 in Lisbon last week. I am currently working on improving our newsroom workflows using technology so the sub-title of the conference “Towards the Augmented Newsroom” really piqued my interest. Scanning through the programme, there seemed to be some interesting talks with quite a few mentioning technology solutions, such as AI and Blockchain.
Over the three days, a number of themes emerged from the sessions. I will describe each one here and the technologies being used to overcome the issues. The first of these, and possibly the most prevalent, was Trust.
Combating the loss of trust in media
Unless you have been living in a cave for the past 2 years you will have heard the term “fake news”. Donald Trump has used it when criticising some media outlets and articles. Coupled with the rise of social media and the ability for stories (false or otherwise) to be circulated rapidly around the world, this has lead to a perceived distrust of media brands.
At least two presenters at the conference called for people to stop using the term “fake news”, as it is now poisoned to mean anything that you want to discredit regardless of its veracity. The preferred term for something factually incorrect was “false news”.
Emily Bell, of the Tow Center for Digital Journalism, presented some fascinating statistics based on her research surveying newsroom staff indicating that among other things
- 83% have changed their production processes (in a minor or major way) in response to the rise of social media platforms
- 86% believe social media platforms have caused decreased trust in journalism
In contrast to this, in Friday’s keynote with Jimmy Wales (Wikipedia/WikiTribune) and Ed Williams (Edelman), Ed stated that trust in major brands has actually increased, the drop in trust is only on Social Media platforms. Whether this extends to major brands posting on Social Media is not clear.
So what can technology do to help this perceived lack of trust in brands?
Blockchain, one of the biggest technologies of the moment, is being put forward by a number of companies as the solution to this problem. Blockchain is a pretty difficult thing to understand fully and one of the points raised was that you don’t really need to understand it in order to utilise a technology. The Internet runs on HTTP and TCP protocols but 99% of internet users don’t understand how they work, they just use them.
Both Hubii and Civil have propositions both based on the Ethereum Blockchain technology. Hubii and Civil provide a decentralised content marketplace where content can be submitted and verified by the community, and then consumed directly by individuals or by content distributors. In the case of Civil, publishers can be whitelisted if they agree to the code of conduct. Anyone in the community can challenge a publisher if they believe they are in breach of the code.
Hubii also proposed a model whereby to assert a fact you would need to stake money against it, if your fact was found to be false you would lose the money, thus incentivising people to provide correct answers.
Both are based on a self-moderating community model and require consumers to use their platform rather than getting their news direct from a publisher. I find it difficult to believe that major publishers will voluntarily sign up for a model where they have to distribute content through a marketplace like this. Perhaps this is one for the smaller publishers or individual journalists.
Robots in the Newsroom?
The next big theme of the 3 days was Machine Learning and Artificial Intelligence and how it is changing the newsroom. I’ve heard a few descriptions of ML and AI but the one given by Benedict Evans, from Andreessen Horowitz, was possibly the most coherent I’ve come across. You can watch it here:
One part of his description that I particularly liked is how he described Machine Learning as being able to do anything that a million 10 year olds could do. This gets across the fact that you can use it to solve problems that require simple tasks to be done on very large datasets.
There were a few presentations about other uses of AI including one by RADAR (Reporters And Data And Robots), who were funded initially by the Google DNI project. They are using AI techniques to find stories in public datasets and then localise stories initially templated by journalists. They have run a pilot already with a number of publishers and have published content both online and in print.
Another interesting application was presented by La Parisien, who have built a system to provide contextual information on content about Paris. This ingests a lot of public datasets and understands which data is relevant to present on particular articles. One example given was to present train timetables on articles about train strikes. Other data that could be presented included weather updates, events near you, restaurant reviews etc. They are also able to use the time of day to further customise the users experience, eg. giving commuting information at the relevant times.
The main application for newsrooms currently appears to be for recognising patterns in data in order to discover newsworthy stories. Although, as Benedict mentioned, the full capabilities of ML will only become apparent as the space matures and people think of creative new uses for the tech. He also stressed that we have a responsibility to train these systems with good data to get reliable results and as with any technology there could be people willing to use these techniques for evil.
Alternatives to Advertising
The last theme is the one that has been troubling the news industry for some years now, how to shift away from reliance on the failing advertising-based business model. Understandably, this is high on the agenda for all traditional news businesses. A fundamental shift is underway and the industry is struggling to find an alternative.
Paywalls, in their various flavours (hard, metered, premium etc), have been the go-to option for most newspaper websites and there was no shortage of paywall technology on display.
A few variations on the classic model were Poool, Jamatto and Laterpay. Poool offer a simple interface to tailor the paywall experience to different user segments, allowing you to constantly optimise the journey for all types of user. Jamatto offer a micropayments system.
Laterpay lets a user past the paywall without paying (only agreeing to pay later) meaning they can get access to some content before they decide if they want to buy. At the point where a user has consumed $5 worth of content they are then presented with the paywall, they transact and carry on until the next $5 is consumed. They also offer subscription periods as well as the pay as you go option.
The only real revolutionary idea for a change of business model came from Tony Haile (Ex CEO of Chartbeat) and his company Scroll. They propose that news companies are not making the most of the vast majority of their website visitors, who actually don’t really care that its your website they happen to be on. What they do care about is the experience they get when they are there. People want fast webpages without ads. They suggest these people would prefer to pay $5 a month to improve their experience on a group of websites.
Their solution is to sign up publishers to implement their javascript-based solution on their website, which would prevent all advertising showing to any Scroll subscriber. They then deliver an amount of their subscription fee back to each of the publishers, they are convinced that even this small amount would actually equate to more than the publisher would earn from display advertising.
I see a few issues with this model, firstly the more publishers that sign up to this the smaller is the cut of the subscription fee each would get — so it could be a victim of its success — or may require a greater fee to the customer. Secondly, this model relies on websites continuing to have display advertising (despite its dropping value) as an incentive for the user to get around it. If display advertising is eventually so ineffective that sites opt to drop or reduce the amount of advertising then this model no longer works. Finally people can already use Ad Blockers to cut all advertising out of pages, so they will need to offer something more than just the ad-free experience to make this viable.
Rana Ayyub and #metoo
One last talk worth mentioning was regarding the #metoo movement and sexism. Rana Ayyub, an Indian investigative journalist, recounted her experiences of the past few months, having suffered an extended campaign of abuse and death threats through social media, including WhatsApp, Facebook and Twitter. Her passionate account received a standing ovation at the conference and left an impact on everyone there. A timely reminder that humans will always find ways to misuse technology. You can watch her account here (from around 14:30 minutes in):
