Has anyone used something like this for non-fictional contexts? I'm really enthusiastic about the future of interactive journalism, particularly in terms of tailoring journalist narratives to each reader.
For example, I think coverage around the announcement of a new tax plan should adjust to the given user (ex. "Under Clinton's plan, your taxes will increase by 5%, to $14,000 annually."). Unfortunately, there just don't seem to be many good tools for allowing non-programmers to encode manipulable facts into their narratives.
I was just thinking about this today because of the Germanwings crash. What if instead of a news article you were presented with a page which somehow led you through a knowledge graph of the topic based on questions or actions by the reader?
When news is just breaking there would obviously be lots of dead ends, but perhaps if people kept inputting the same sort of question it would get flagged and become more important as the story develops. For example in the case of the plane crash, after hearing the initial reports of the pilot being locked out my immediate thoughts were:
- What is the lock mechanism like? Can the door be opened with force?
- How is the lock controlled?
- How is the altitude for autopilot set? Is it easy to set 96 feet instead of 9600 feet?
- Is there any record of the copilot's vitals before the crash?
- Why wasn't there another person in the cabin while the pilot stepped out?
...and so on. Some of these were answered, but the only way to really get the information is to read (or listen to) the same kind of reports over and over, hoping for new information and trying not to mistake an over-eager rephrasing as new facts.
What would be great is if I could see a list of topics, select one, and then either highlight sections/tags to follow up on, submit a question, or follow a link to an existing question. The UX here would be a very interesting study I think.
Inform> Scream expletives at reporter
That's not a verb I recognize.
---
The simplest way to do this would probably be to get a team of interns to trawl the net looking for what people are talking about after a story breaks (such as top voted Reddit comments), rephrasing them as necessary, and then turning them into tags which people reading the story can then subscribe to. Readers could select existing ones or see them fuzzy completed as they input their own questions. Hmmm.
Yup, I think this is actually one of the more interesting problems on the web today. We have access to more and more information than ever, but it's not organized in good navigation or narrative structures.
Interestingly, I actually think where the innovation is needed is on the input side. It's possible to produce a structured and personalizable story today, but it requires long programming lead times. We need to create systems which make it possible to build such stories at the speed of news.
---
(Shameless plug, but if anyone is interesting in this problem space I'm currently hiring. Email morgante@cafe.com if interested.)
For your Clinton Plan example, you might want to check out the Tangle JS framework for reactive documents. It allows readers to change a variable in a text and have other numbers change accordingly.
Yes, I'm quite familiar with Tangle and Bret Victor's other work.
While it provides an okay experience on the reader side, it's jut not usable on the writer side—in terms of composition, it's not much better than any other JS framework.
What I really need is a way of letting non-technical writers create interactive stories. Probably the most successful example of this is Excel.
For example, I think coverage around the announcement of a new tax plan should adjust to the given user (ex. "Under Clinton's plan, your taxes will increase by 5%, to $14,000 annually."). Unfortunately, there just don't seem to be many good tools for allowing non-programmers to encode manipulable facts into their narratives.