If you’ve logged onto Facebook at any point in the past few days, you’ve almost certainly seen oddly on point non sequiturs filling your news feed. These are coming to you straight from a new app called What Would I Say, which trolls your Facebook presence to pick up your most used word and phrases, and then generates new statuses for you. The results are unbelievable.
The site uses what’s called a Markov model to figure out what to write. After sifting through a Facebook user’s status updates, the site picks one word and starts calculating probabilities of other words that are likely to follow. Eventually creating a sentence.
If you need some comic relief, check out “What Would I Say?” a new website that generates Facebook posts based on your past status updates.
The website goes through your past posts to get an idea about the things you talk about and how you write. Click “Generate Status” to create new posts, and you’ll likely come across a number of hilarious results. There’s also a good chance the website-generated posts might not make any sense.
Here are a few of the results I got:
“Losing happens, but these are still the Harlem Shake”
“Alas, I would have gone missing. #Apple”
“Use protection when you sell your soul to the HTC First, aka”
“Keep calm and kick BAM!”
At best, the statuses seem like something you would write, but at worst, they seem like they’d be something you write after a long night out.
The hilarious website was created at hackathon this weekend by Princeton students. The site took just 36 hours to make, and now, it is quickly making the rounds on Twitter and on news websites.
To use the app, simply head to what-would-i-say.com and hit the Facebook login button in the top right corner to give the website permission to access your past Facebook posts. For those worried about the site storing your information, “What Would I Say?” assures users that users’ Facebook status data are stored in their own Web browsers.
The students told The Times they were inspired by “the bots” users chat with on messaging services like AOL Instant Messenger. Those bots were automated accounts that were programmed to respond to messages sent by real users. Often, the results were hilarious.
We “realized that Facebook posts would be a great resource for training these bots because they’re so personal and funny,” the students told The Times in an email.
Play around with the app and see what it comes up with based on your past posts. If you see a status you actually want to share, simply hit “Post to Facebook.”
By Salvador Rodriguez
(c)2013 the Los Angeles Times
Distributed by MCT Information Services