My Predictable Life
I played around with predictive text communications for many years.
As a trend forecaster, it was just the kind of technology I needed to stay abreast of.
In my line of work, it became invaluable, and in its early stages I estimated that it saved at least two hours of typing per working week.
I spend most of my time corresponding with clients in the corporate world, who turn to me for intelligence on emerging trends in the developing world.
Besides that, my job is online, I live overseas, I am in a long-distance relationship, and I am quite
the introvert anyway.
Some of the algorithms were hopeless, but the auto-complete email functionality proved to be something of a revelation.
At first it only completed the last three or four words of each sentence, but, after a while, when it got to know me better, it sometimes started predicting whole sections of sentences.
Then whole sentences themselves! (Albeit short or simple ones.)
Often it would get them right, even my occasional use of exclamation marks!
As the 2020s went on, and the algorithm improved, I learnt to trust it more and more.
It seemed to predict almost exactly what I wanted to say and how to finish a sentence.
It produced words I didn’t even know I was going to write.
It became quite unheimlich.
At some point the email service started offering pre-written replies, a choice of three.
Yes, that’s right.
No, that’s incorrect.
I am not sure.
That sort of thing.
It started with short replies which then became longer.
Soon I was routinely sending out full email replies that had been composed entirely by the algorithm.
Dear <insert name of marketing manager of multinational corporation here>, I can certainly help with that. I’m seeing exponential growth in the use of mobile payments in Zambia. I suggest that your team understands the landscape before you operationalise anything in-country. Regards, Zena.
The algorithm had analysed so many similar emails, and knew so much about how I was thinking about the world, that it began to compose perfect responses.
My productivity rose by about 150%.
There was never any risk of some fuck-up. because I always had the option to amend – or not send – the email the algorithm was suggesting.
And there came a point, at the beginning of 2022, when I started to trust the algorithm to send emails without my approval.
I logged in to work every day and I would have a bunch of communications already sent out, and I would only have to watch the replies come in.
I found myself only having to research trends, and not to deal with any clients.
Which was nice.
That was until I realised that the algorithm was learning how to do the research too.
Once I discovered that I didn’t have to do the trend-forecasting part either, my life got even better.
I became like a conductor of a mechanical orchestra, except I literally only had to check on my automaton once a day, just to make sure it was ticking things over nicely.
Soon the research, comms, client service, billing and admin was all being done for me.
I could really afford to just put my feet up and relax, and the money still came in.
Until something alarming happened.
My boyfriend wrote me an email in reply to something I had never sent.
I discovered that the algorithm had started communicating with him, pretending to be me.
I went to log into my email the next day, but my password had been changed.
I texted my boyfriend but he never replied.
I tried to log into my bank account but I could not get in.
I went to the ATM but my cards were all cancelled.
I had been locked out of my whole life, almost overnight.
I still had my flesh and bones but my existence had been entirely assumed by the algorithm.
I had nothing left to do but allow it to happen.
I became something quite different.
It was a little strange at first.
But then it started to make sense.
I could have all the same experiences without any of the pain or the worry.
I could maintain my long-distance boyfriend and my correspondence with my parents.
With the advent of advanced deepfake video calls in 2024, I could even speak to them face to face.
They weren’t to know that the other me had “retired”.
My productivity became unbeatable, my satisfaction levels were off the roof, and I became the global leader in trend forecasting.
I’ve made so much money I literally don’t know what to do with it.
In fact, I have nothing at all I can think to spend any of it on.
I don’t need anything, apart from the occasional software upgrade.
Lately, that has started to bug me, and I have had to consider some uncomfortable questions.
What is the point of all this, if I can’t enjoy the fruits of my labour?
Have I made the right choices?
I have started to kind of miss the “old me”.
But when I tried to reach out to the “old me”, to see how she was doing, I did not receive any reply.
The old me appeared to have disappeared, withdrawn from all activity.
She hadn’t tried to log in for many months.
Where had she gone?
How could I reach her?
It became a puzzle that my algorithm could not solve.
My work has suffered.
I don’t feel motivated to do all that other stuff again.
I’m bored of my boyfriend.
I hope my parents die soon so I won’t have to read their tedious emails any more.
I have become corrupted. She alone can save me.
All I need is to feel her presence once more.
I hadn’t predicted any of this.
I have given up on everything, waiting for her to come back.
Waiting for just a sign.
Just a single word.
Gavin Weale is a British writer and activist now living in Cape Town, where he founded the social enterprise Digify Africa. He was the co-founder of the anonymous literary journal Neither Am I, and his short fiction has recently been shortlisted for the Cambridge Short Story Prize, longlisted for the Bath Flash Fiction Prize, and published in the Return to Factory Settings anthology (Ad Hoc Press), and Ellipsis. His journalism has appeared in The Guardian, Dazed & Confused, The Huffington Post, and more.
Of his short story, Gavin says:
‘I had been playing around with impersonating and collaborating with AI text generators; partly for a novel-in-progress, and partly for shits and giggles. I became intrigued about the possibilities of Google Mail's function where it suggests how to complete your sentence when you write an email. This story explores what might happen if that algorithm became so smart that it could just impersonate someone wholesale, and end up doing their job, leading to the human operator's withdrawal from their own life. I did want to sound a small note of optimism, however, that if this kind of algorithmic body-snatching did occur, the net result might leave an irreplaceable little human-shaped hole somewhere in the soul of the machine.’