Um, folks, this is a non-peer reviewed paper written by some Princeton students. It's not, contrary to what at least three news organizations suggested, the considered opinion of the faculty of "Princeton's Department of Mechanical and Aerospace Engineering."

Now, the paper's conclusions may or may not be correct; I'm expressing no opinion here. My point is really about standards of journalism -- would it really kill reporters to include the word "student" somewhere in the text of an article trumpeting a paper written by students?

Vocativ (which I generally like, BTW) did this as well with this amazingly declarative headline yesterday: "Facebook Will Lose 80% of Its User Base by 2017" No question mark! Not even attribution. I noted this on Twitter and their reporter replied by saying: "So?" https://twitter.com/declanm/status/425699497092460545

For those of you who may not have been online in the early 1990s, Time magazine did this to CMU undergrad Marty Rimm's paper in its "Cyberporn" cover story. A U.S. senator subsequently waved it around on the floor of the Senate to justify an Internet censorship law: http://w2.eff.org/Misc/Publications/Declan_McCullagh/www/rim...

That was the good Sen. Chuck Grassley, in June 1995: "Mr. President, Georgetown University Law School has released a remarkable study conducted by researchers at Carnegie Mellon University. This study raises important questions about the availability and the nature of cyberporn."

The Communications Decency Act became law some six months later.

Update three hours later: An editor at Fast Company responded and said they'll add a note to the story. https://twitter.com/JessHullinger/status/426050602984935424

They're using Google Trends to gauge how many people are using Facebook. Since 90% of this is people typing in "facebook" on their homepage to get to the FB login page, this might actually be a surprisingly good metric to track facebook usage.

Then again, the recent dip in searches could just mean that people are finally learning to use bookmarks, or click on Facebook directly now that most browsers show you your favorite sites when they start up.

Or more are just using mobile apps to use facebook.

Or they are using the mobile apps more and more.

The study was not peer reviewed, and is nothing more than an interesting experiment, not deserving of the factual assertion of the headline.

Although at least it's a testable prediction.

They vastly underestimate the number of old folks that want to keep up with their grandchildren. This is how I and my children use it now. It is the easiest platform to put on an iPad to send the old folks updates.

Is it really easier than email?

Yes. I can post a single picture of my kids and all my relatives see it at once. Sure, Instagram would be more to the point, but most of my family is on Facebook. Consider that email means I have to type up everyone's address and not forget anyone. Facebook lets me just post and people may choose to view or not view what I post.

The prediction is based on how pathogens evolve over time with no basis on how this apply to social network adoption.

A comparison with Myspace doesn't help either since Myspace decline started when FB started growing rapidly, and currently apart from G+ (which isn't growing that rapidly) there's really no competition out there.

with pathogens carriers either die or adapter either way they stop infecting after a period of time. a better analogy would be telephones or the internet itself.

Unsurprising if true. Out of several hundred friends, I'm down to perhaps 10 that post to Facebook on a regular basis, and they'd probably be better served by moving their activities to Twitter.

Are those 10 people really the only regular posters? Or are those 10 friends that Facebook has decided to spotlight on your news feed? I never feel like I'm getting everybody's updates because Facebook is trying to only show me "important" updates.

How do infectious diseases have anything in common with social media?

Both spread by social contact and network effects. If you don't interact with anyone that has the flu, you won't get the flu. If you don't interact with anyone that has facebook, you won't get facebook.

I have no idea if their spread followed similar models, but it wouldn't surprise me.

I can see the correlation in adoption rate, but social media users don't get purged from networks like infectious diseases do from our bodies.

The reasons people use and/or leave social networks are completely different from the reasons diseases leave the body, which is the correlation needed for this study to make sense. It just seems like linkbait to me.

Interesting corollary: if you've got two smaller social networks, both trying to grow, of equal sizes, the one with a denser group bond (more edges between nodes) has a better shot at it than the sparser one.

Which is to say: start with a tight group and you'll tend to grow -- the benefits to members of participating (and costs of defecting) are higher. With a widely distributed network, not so much.

Facebook exploited this with its initial seeding: Harvard / Ivys / selective colleges / all edus / general public.

If I recall, a strong reason for Orkut's failure to launch was that it never really developed a tight inside bonded group outside of Brazil and Latin America.

Much as epidemics spread in densely populated areas.

Gives a new meaning to "viral media" and "social disease".

So going forward Facebook endless buys companies to try and prop up its user base?

I suppose that could work to some extent, since Facebook knows how to monetize.