Given that RSS is primarily a one-way information tool – we get information without letting the publisher know anything about us – it might seem strange to think about RSS in terms of privacy. However, this may (some would say must) change for RSS to expand its usefulness. This requires a little background.
A point I made several times at this year’s Gnomedex is that we need to stop thinking about RSS in terms of subscribing to individual feeds and start thinking about how those feeds can be aggregated to provide us the information we really want. RSS feeds contain individual items, so why subscribe to an entire feed when you’re only interested in a few of its items?
The first step in this process already exists: bloggers who act as editors by gathering and reporting links of interest to their readers. These editors enable us to subscribe to a single feed yet still keep up with the interesting posts from multiple feeds.
Take my own blog as an example. Let’s face it, I rarely have anything interesting to say. I’m usually nose-down in software development and don’t have time for extended blog entries like this one. So rather than subscribe to my feed, wouldn’t you rather subscribe to a feed which lets you know when I actually have an interesting post, especially if that feed includes only the interesting items from several other feeds? Blogs such as Boing Boing are obvious examples of this, as are aggregate feeds such as the ones Daypop provides. Link blogs such as del.icio.us where everyone acts as an editor perform a similar function.
That’s the first step. The next step is for your aggregator to let you know when an item of interest to you has been posted, regardless of whether you’re subscribed to a feed that includes that item. In some ways, FeedDemon’s search channels perform this task, but these are keyword-based rather than usage-based. You can also do this by viewing the Technorati Link Cosmos for an item you find particularly interesting, but that’s a manual, one-item-at-a-time process. And online aggregators such as Bloglines provide recommendations based on your subscriptions – which is great – but this is a one-feed-at-a-time process.
What I’d like to see – and what I envision FeedDemon becoming in the future – is an aggregator that acts like Windows Update. You tell it to go out and find what you’re missing, and it comes back with items that it thinks you’ll find interesting. You’d still be able to subscribe to individual feeds, of course, but unless you’re Robert Scoble, you’d only subscribe to a few feeds and let the aggregator do the rest of the work.
I’ve seen experimental aggregators which operate along these lines, except that they require you to rank items, and these rankings are used to make recommendations. Nice idea, but given all the interesting items you read in a day, would you really want to have to rank them? It seems to me this would have to be seamless, and wouldn’t require extra effort on your part.
For this to really work, your aggregator needs to silently collect attention data based on your reading habits. That attention data would then be provided to a service which recommends related items. And that’s where the privacy issue comes in.
Any service that expects to receive your attention information must do so anonymously. It should be absolutely impossible for anyone – or anything – to discover who you are based on this information. That seems obvious enough, but so many companies have got this wrong in the past that it bears repeating: if you want my usage information, you can have it provided that it’s 100% anonymous. And that, of course, boils down to trust.
I should add that this isn’t going to happen right away, and even after it happens, aggregator developers such as myself will still need to provide filters and other tools to find interesting items. But wouldn’t it be nice if a trusted source could let you know about things you’ve missed?
Related Link: Steve Gillmor: RSS for Food