Michael, I think your self-driving cars analogy is a good one to consider. We're closer to that stage than most people realize and I doubt many people would have guessed that would be the case five years ago. The story doesn't go away in the model I'm describing. The model is less useful in the fiction world but the stories that need to be told in in the non-fiction space, including things like "here's how such-and-such is accomplished", are just as useful in the model I outlined. Just look at that Narrative Science example I also provided; they're able to spin up a story based on nothing more than names and numbers in a box score. David, I didn't say books and other containers will be totally eliminated. In fact, I noted that they'll fade into the background but they'll still be around for many years (see my opening paragraph in the original article).

You make an excellent point regarding the complexity of this task, Brian. I'm assuming this parsing tool would have some inputs (e.g., known keywords and perhaps synonyms for those keywords) as well as the ability to get smarter as it processes more text. It's easy for me to say and hard for someone to develop, I know, but I'd like to think it's possible.

Vourkos, you are absolutely right that models like this exist. As I mentioned in my article, I was inspired by Google's service as a standalone app. I haven't seen the model applied to ebooks though and that's why I wrote this piece. I'm curious to learn whether you feel your Pollfish service could be integrated with an ebook application. Drop me an email and let's continue the conversation there as it would be great to experiment with you on this. I can be reached at jwikert [at] gmail [dot] com.

Hi Ken. You could certainly implement what you described above but my biggest concern is scalability. You'd ideally want a model where the ads are more fluid and require no manual intervention (or selling). However, in the travel agency example you mention, I could see where maybe that agency might be interested in sponsoring and being the exclusive advertiser throughout that book. If so, and if the financials make sense, those survey questions that appear from chapter to chapter could be for the agency's questions and they'd have full access to all the answers. Ads in print are another option but they obviously won't have all the dynamics and flexibility of digital editions. Plus, you won't be able to get much data from them other than discount code conversions you insert in the ads.

Thanks for your thoughts on this, Mike. Perhaps a real example will show you that this doesn't result in throwing the retailers under a bus. The ebook is generally created at the same time the print book is about to go off to the printer. Let's say that date is March 21. I'm suggesting the ebook becomes available exclusively on the publisher's website on March 21 while the print book is being produced. In all likelihood, that book going to the printer on March 21 doesn't hit physical bookstore shelves till at least two to three weeks later, typically April 4 or 11. In reality, it often takes even longer than that before it first appears on shelves. In the scenario I just described nothing much happens between March 21 and April 4/11. Sure, print backorders accumulate, but that process could continue in the scenario I'm describing. For those two to three weeks after March 21 the publisher could make the ebook available only from their website and print distribution is not affected. I suggested a monthlong period of ebook exclusivity in the article but it could just be two or three weeks. It depends on the publisher's schedules and how they decide to leverage this sort of approach.

You're quite welcome, Brian. I'd like to clarify something... The vision I described in the article isn't mutually exclusive with an index in the back of the book. I can see where it would still be beneficial to include the back-of-book index as well as the inline one, mostly for the points you described.

Thanks for adding your valuable comments to this discussion, Jan and David.

You're right, Thad, and I believe our points of view are not mutually exclusive. The end-of-book index can co-exist with what I described in the article.

Hi Bob. Thanks for sharing your Baen experience. Your point about letting customers send the books themselves is where I see an opportunity. It can still be automated but as long as the files are sent by the customer, not the publisher, there shouldn't be a problem. Still, as I noted before, Amazon could shut the entire program down but that would seem a bit over-the-top, even for Amazon. I realize what I'm suggesting here is different from what I originally wrote but my point is that this still represents a service that has potential.