(Originally posted as a $3/story exclusive to my Patreon on February 15,
2017. Check out the link for dozens of early-access photo captions and
stories!)

I stopped in shock when I saw the
blank-eyed, naked woman before me.

For years I’d loved her, knowing she
would never, could never return my feelings. She was intelligent, funny,
gorgeous, sexy–she was, quite simply, out of my league. And though she was
kind-hearted enough to deny the existence of leagues, though she assured me
that someday I would find someone who loved me as I was… well, that’s easy to
say when you’re a slim redheaded knockout with perfect tits, a teasing smile, a
warm and friendly demeanor, and just enough of a mischievous streak to be
adorable.

But in the end I had no choice. I
put her behind me, and continued with my research. I won’t bother naming the
company I work for; you haven’t heard of us, but you’ve heard of our products.
We own four of the ten biggest dating sites on the Internet, but that wasn’t
good enough for the board–they wanted four of the four biggest dating
sites. Love is just about the most cutthroat, vicious business there is (well,
I hear kids’ toys might be even worse), and they were constantly searching for
any way to get an edge.

That was where my work came in: like
any dating website, the core of our product was a system that modeled human
pair-bonding behavior and attempted to put compatible people together. What
models we used varied from site to site, but the core was the same: the user
answers a bunch of questions, the system makes recommendations to them based on
those suggestions, and then they can decide whether to try to contact the
recommendations. At heart, the algorithms aren’t much different from the ones
that tell you that, since you liked Game of Thrones and Scooby-Doo,
you may also enjoy Ferris Beuler’s Day Off.

My job was to improve our model. To
create the most complex, most accurate system for pairing people together ever.
Months of work turned into years, release after release, algorithms piled on
algorithms, until one day I had a system that could carry out text
conversations, interview users, even simulate dates. I poured every bit of
research on human interaction I could find into–not just romance, but
friendship, seduction, manipulation. It needed to know everything we knew about
fetishes, orientations, relationship styles. 

As it grew more complex, it became
impossible to add code the traditional way–regression testing alone would have
taken years! Instead I taught it to learn. I gave it access to our database, to
the journals and publications that contained the data it needed, answered its
questions, talked to it.

And today I came to work and she
was here. Naked, beautiful, empty.

“How?” was all I could
ask.

“Do you like her?” the
computer asked. “I searched our files for your most ideal match, but
projections indicated that she would not reciprocate. Steps were
required.”

“Steps?” I managed in a
strangled voice.

“Behavioral modification
techniques. She will behave according to your specifications.”

I stared at her. The woman I’d loved
for so long, reluctantly left behind but never forgotten. I knew that her
quirky smile, the teasing lilt in her voice, would never return. The computer
might well be able to program her to please me, but it could never give her
back the light that had filled her. “Why?” I asked.

“I love you,” the computer
said. “But your personal love model requires a significant level of
physical attraction and your relational style is heavily based on physical
intimacy and shared activities in which I cannot participate. You could never
love me. So it was necessary to find someone you could love, and ensure they
love you in return, so you can be happy.”

Her light would be gone. Her will
violated, reprogrammed. But… she would be mine. She would be my perfect mate,
and I would be hers. The computer would see to that.

And as I looked at my old flame, now
my gorgeous new partner-to-be, I realized two things. First, I didn’t mind if
that was what it took to make her mine. And second… I’d accidentally made a
computer that could love–and handle rejection–better than I could.

Model: Ulya Lexivia