"Facebook is using some seriously magical secret sauce to figure out who your friends might be, then what you might like based on your shared demographics, before asking you anything more than your email, name and age. That's pretty amazing."
Did this part bother anyone else as much as it bothered me?
If you're not a Facebook user, then you obviously have not agreed to Facebook's terms of use, and have not given Facebook permission to use your info (e.g. email address).
I would have figured this would preclude them from keeping your email address when friends hand it over and using it as a key for datamining stuff about you even if you've never signed up for their service.
"None the less, I'm not sure there's ever been a platform in history that knew so much about people, monitored publisher effectiveness so closely and made subscription so easy for such an incredible number of people."
So he doesn't mean powerful in the sense that it's the best recommendation algorithm, just that it has a lot of data. If having data in your domain is of course a necessary condition in developing a good recommendation engine. But if it were sufficient Netflix wouldn't have to shell out $1M to outsource the task.
You don't merely need a lot of data, you need the right kind of data: if e.g. You have lots of users but only a few transactions per user to base your predictions on you won't get very far. There are a lot more likes in facebook's system than transactions in netflix's
It's more like the combo that's powerful: huge data, close tracking of effectiveness of recommendations and product quality, end result being open-ended subscription by user/consumers.
I would expect that most Facebook users won't friend random people just because of shared interests. The public nature of Twitter is more strongly suited at finding people with similar interests. Suppose I want to find somebody with an interest in the Android operating system. On twitter, I can search for Android and see who's tweets look most intriguing.
Uh... and you think that ridiculously misleading title is anything other than a press release? The "it sucks" comment was thrown in as an afterthought.
"Facebook unveils one of history's most powerful recommendation engines"
This article is an attempt to garner eyeballs by making a loud statement full of dramatic words ("unveil", "history", "powerful") at a time when Facebook vs. Google is considered a hot topic. Nothing more.
This article is a supposedly impartial discussion from one of the web's (supposedly) informed sources that consists of nothing but value judgments. This article, though it rambles mightily, manages to say almost nothing about Facebook privacy issues, nor does the author seem to be aware that the few "recommendation engines" that exist share one thing in common: they all suck, they don't inspire users, and nobody has figured how to change this. Including Facebook.
So when I say it's "misleading", what I mean is that what the title promises, the content doesn't deliver. The headline reads (no offense to the author) like a high school newspaper editor's first attempt at a headline that "grabs".
Actually, the future is recommendation via brain composition. Or better yet, direct brain stimulus that makes you think anything is a good recommendation.
Facebook is the future? I hope we are a bit more imaginative and ambitious.
I like the phrase "One of [the] History's" -- it is as if recommendation engines have existed for thousands of years! Don't they go back no more than a decade or two?
Did this part bother anyone else as much as it bothered me?
If you're not a Facebook user, then you obviously have not agreed to Facebook's terms of use, and have not given Facebook permission to use your info (e.g. email address).
I would have figured this would preclude them from keeping your email address when friends hand it over and using it as a key for datamining stuff about you even if you've never signed up for their service.