Monday, January 30, 2012

Automating the conversion of girl friends to girlfriends

Since I'm already married, and it looks like Levi is getting with Yang, I guess I can go public with this.

Levi Self and Shirley Yang should have babies

I have developed a semi-automatic system for converting girl friends (meaning friends who are girls) into girlfriends, using Facebook and some macros. I'm not going to release any source code, but I do feel that is important to share this kind of information with the community. To this end, a walkthrough of the process is detailed below.

. . .

Step 1: Make a list of all of the Facebook friends you would date. They can be celebrities, friends, acquaintances, or cousins, as long as you would date them. 41 of my 483 Facebook friends are girls who I think are cute. The ratio will be different depending on your Facebook popularity and your standards, but I do think that 10% is probably close to the median ratio.

Is it just that I'm easy, or do I have a lot of cute Facebook friends? 

Step 2: Write a macro (I suggest using AutoIt or AutoHotKey for this) to iterate through this list of girls, pasting the following into private messages:

Dear %Name%,
I can't believe I'm actually typing this to you (I've only had two glasses of wine :) ). I have always been attracted to you. Tonight I have been (for various reasons) thinking about how short life is, and what I would regret if I were dying, and I think that not telling you how I felt would rank pretty high on the list. So I've got to say this now: I like you. I want to get coffee with you. Write me back and let me know what you think.
Step 3: Sift through the responses and correspond manually with your most promising leads until you've finished the conversion. It should be noted that the conversion rate between girl friends and girlfriends is not 1:1. Your mileage will vary, but count on the permanent loss of about ten friends who are girls for every girlfriend you pick up.

. . .

***Edit: There are people on reddit who are not catching that the above is a joke, and are freaking out about how much of a jerk I am. So, be it known: know I do not actually advocate lying to your friends en masse and trading all your female friends for a shot at one or two romances :) I do not actually cautiously guard as my trade secret the source code of my esteemed girlfriend macro. The rest of this post *is* somewhat serious.

This method is actually an adaptation of the one I used to use to get textbook editing gigs. I built an OPML file (an aggregation of RSS feeds) from all the world's craigslist gig listings. I then ran searches over the gig posts looking for words that were frequently used to describe gigs I could do well. I manually reviewed the results of this search, and then sent the plausible ones this:
My name is Kenneth Myers. I'm an instructional designer, software programmer, and former professor living in north Texas. I happened upon your craigslist post today, and if our distance wouldn't be an impediment to our working together, I'd like to be considered for the job.
I'd be delighted if you'd consider my resume, and even more delighted to hear back from you, either at this address or at (210) 289-3858.
I wish you the best of luck in finding the best fit for the job,

I got lots of jobs. I made really good money some months. And I began to notice a strange pattern: even when I would apply for a gig in Iowa, people would assume that I had serendipitously been inspired to check out their out-of-state craigslist on a whim, found their job posting, and applied to it and it alone. Much as we do with romantic relationships, we humans by and large believe that careers should grow out of our lives organically and accidentally. The deliberate engineering of these things runs too much against the grain of our internal Disney life narrative, where storybook logic assures that we have the friends we were supposed to have, and don't get the job because it wasn't meant to be.

I think narrative is the meaning of life. But I prefer to write my story.

Directions in future research

Having already the best of all possible wives (not kidding), the facebook girlfriend converter is a solution for a problem I don't have. So I'd like to begin work on an algorithmic attack on the age-old problem of friendship. How many friends of your friends are ones that you grew up with, or went to school with, or worked with, or met through another friend? For most people, my guess is about 100%. Is there any reason for believing that the people who were assigned seats next to you in high school have the highest probability of bringing you the most joy?

With sweat and desperation, we miss sleep to apply our skills to server problems at work. But the most important stuff of life, we leave to chance.

No comments:

Post a Comment