I don't read 1000/day, but I do step through 1,000 - I scan a headline/summary, hit "j" and move on for the vast majority of them. Some i'll star and come back to later, for others the first sentence of the summary is really enough to learn something about what's going on.. eg "vulnerability found in XYZ" is useful to tuck away in the back of my mind, but that's all i need.

Others are the same story reported by multiple sources; obvious from the headline.

I leave it open in a pinned tab and when i want to kill a minute or two, i'll scan through a few more posts and open a few to actually read.

The reason i like Bazqux is that it can keep up no matter how fast i hit "j". It's no prettier than Google Reader does, has few more bells/whistles, but if you really want to speed through items, save and search, it's great.

In reality, I would like something that combined Bazqux with some ML to recognize that an item is a dupe, or something i'm really not likely to be interested in and filter those thousand items down to a few hundred, but i've yet to see anything really achieve that.

Thank you!

> In reality, I would like something that combined Bazqux with some ML to recognize that an item is a dupe, or something i'm really not likely to be interested in and filter those thousand items down to a few hundred, but i've yet to see anything really achieve that.

About a decade ago - when the state of the art really wasn't up to it yet - I invested some money in a start-up that was attempting to achieve just that. They eventually turned to greener pastures and are still alive but I am still hopeful that someone will manage to put this together. I'd be more than happy to pay for such a thing.

Which reminds me of another dead product I used to use to do exactly this: Yahoo Pipes.

Reading 1000 things a day would probably make me suicidal