Egoclip is the first RSS aggregator I know of to apply machine learning techniques to try to automatically identify posts of interest.
I don’t know exactly how it works. I don’t even know roughly how it works.
What I do know is using it is frustrating. It happily imported my feeds from sharpreader with their folder hierarchy, but when I tried to reorganize them, I quickly found that there is no obvious UI for creating new folders.
More annoying, I can’t use the arrow keys to navigate through posts because the focus switches to the post reading pane.
There are other frustrations too, which is a shame, because taken together, they will probably be enough to keep me from investing the time in using egoclip needed to train it.
I’ll check in on egoclip in another month in the hopes that it’s good enough at being a basic aggregator to be worth experimenting with.
All of this illustrates a basic issue with any product that requires the user to make a commitment beyond purchase/download to take advantage of the product — you need to give the users something good up front to make them stick around for the grand finale. It’s really no different from telling a good story. The real payoff might come at the end, but you’ve got to keep the audience interested.