Get this: do away with of all of the content creators for the various travel and leisure websites and replace them with robots that generate content in “real life” language – complete with local and regional dialects.
This was the view of Kevin Ashbridge, vice president for global travel and leisure solutions at translation technology company SDL, when he made the proposition at The Phocuswright Conference last week.
According to Ashbridge – who admitted that while the technology wasn’t perfect, it could be perfected “if given time” – automating the creation of content would be able to bridge the challenge of delivering personally and contextually relevant content at the rate of technology adoption.
Simply put, the rise of the smartphone and the tablet has made it essential to provide an integrated customer experience, but that content cannot be personalized – unless robots are deployed.
Subscribe to our newsletter below
Other industries have already integrated AI into content generation – for example, Siri or Google Cortana use NLP, or natural language processing, which is technology that understands human language by a machine.
For the travel industry, though, programmer Julia Matyunina of Code Tiburon suggests that while AI can create content for travel sites at an enormous pace – saving both time and money – there are still a lot of “bugs” that need to be ironed out when it comes to deploying it on a large scale. What’s more, these “bugs” create an issue of trust in the content, thereby making it unreliable.
“AI is good for time-consuming tasks that do not require creativity. However, it still needs human supervision and decision making. And it’s hopeless at creative and more demanding tasks where a personal touch is a game-changer,” she says.
And this, too, is a position that Ashbridge has taken. “We haven’t gotten rid of the human editors behind the content creation. But, in terms of the day-to-day stuff – in terms of the general output of content to then be edited – we’ve cut out the ‘middle man’ and turned it over to the machines,” he says.