Just imagine, if 100 years ago the Wright brothers would have tried to build an ‘artificial bird’ rather than calling it an airplane? Or Henry Ford an ‘artificial factory’ rather than a car manufacturer? We would have had endless discussions around engineering this machine to emulate birds in every detail. Would you feel comfortable to fly on an artificial bird? Makes it emotional right? Does an artificial factory produce better cars than a human-centered production environment?
Artificial intelligence was invented as a term in the 1950’s when computer science professors at Dartmouth College convened a conference in order to teach computers to think during one summer – a slightly optimistic view as it turned out. While the researchers did not succeed in one summer, their core principles and main ideas have had a lasting impact.
Instead of emulating the human, we (the authors of this article) believe brain machine learning is best applied when it focuses on the items that human brains aren’t ideal for: endless repetitive tasks, constantly scanning vast volumes of data, broad mathematical simulations and rapid decision making based on big data. So, what does this have to do with media and publishing?
Let’s look at AI in other industries first to know how the wider market is affected by it:
• According to a recent study from Oxford University, AI will replace up to 45 per cent of jobs within 20 years globally
• It doesn’t just affect blue-collar jobs; think of how Uber is aiming to replace all human drivers with autonomous vehicles. But also white-collar professions like insurance agents as 34 employees at the Japanese insurance firm Fukoku Mutual Life found out when they were replaced with a computer programme in January; basic legal procedures face a similar development.
• Chatbots are taking over customer support functions, by not just using pre-defined answers but actually ‘thinking’ and helping customers like ordering at Taco Bell or get tickets from Ticketmaster while scheduling doctor appointments and providing medical info
• Wal-Mart is already applying AI and blockchain to better master its supply chain safety by spotting anomalies and reacting to them pre-emptively
• Artsy has redefined how we find, look at, buy and sell artworks by implementing a massive big data system to catalogue artists, artworks, museums, exhibitions and galleries around the world; the next step is combining this setup with chatbots and automated art consultants – a blueprint for publishers catering to the art aficionados?
The media industry has already experimented using AI in various settings.
• Back in 2014, one of the first examples of robo-journalism was applied when a machine for the LA Times wrote an automated article about an earthquake only 3 minutes after its appearance.
• The Associated Press working with data from the baseball association NCAA, is now using automation technology to provide thousands of stories about college sports it would have previously not covered.
• German newspaper publishers have begun using services like AX Semantics to quickly publish dozens of fully automated Dritte Fussballliga and Regionalliga reports every week about soccer games based on match reports taken from national databases.
• Similar to how Netflix used data to help fine-tune House of Cards, news organisations will have the opportunity to adjust editorial narratives to make stories more engaging
• AI can be used to suggest the right articles based on your reading habits in real-time, 200-year-old Ebner Media Group even has gone so far to combine this with a predictive commerce engine to monetise audiences automatically
• Companies like OpenTopic use IBM’s Watson technology to provide hyper-segmented ‘cognitive content’ landing pages in order to convert website visitors into ‘magazine’ subscribers
• Even in the esteemed world of peer-reviewed science publishing, the use of AI is coming. Too many examples exist of published articles that contain plagiarism and wrong facts. AI could run every submission through a plagiarism detection software system and a robust image detection software system
Get stories like these every week directly in your inbox. Subscribe to our (free) FIPP World newsletter.
What does this all mean?
Like any other disruptive technology, AI will shake up production, business development, marketing, sales, administration and finance of every firm in every industry. Only because we have this traditional mental image of journalism and publishing will not make it different for the publishing industry. We already use automation to produce cars (i.e. products), predict the weather, report on stock markets, trade natural resources, print 3D objects, create VR environments, and recruiting software is targeting the right candidates using precise algorithms – why would the impact on the underlying procedures of the media industry be different?
That’s why we think AI in media will just help to turn this industry into a normal industry where products are enhanced with AI, marketing and distribution optimised with computer learning and customer engagement boosted with chatbots.
Instead of sitting back and letting this AI revolution drive by you, your firm should actively create an AI strategy to think about in what parts of your value chain AI can save costs, improve sales or enhance your production of content.
This is the right time to have (inside or outside your company) your first AI manager. He will lead your strategy based on the aforementioned principles, he will leverage AI technologies to provide your business a competitive edge.
By Dominik Grau & Simon Schneider
Dominik Grau is the CIO of Ebner Media Group.
Simon Schneider is an entrepreneur and UK Director of ECSI Consulting.
More like this