Share
MidReal Story
The latest GPT model starts writing predictive texts about users
Scenario:The latest GPT model starts writing predictive texts about users' futures.
Create my version of this story
“David, Emma, you’re both very familiar with our data model, and you understand the potential behind it,” our CEO said, breaking the silence.
I glanced at David, who was sitting next to me.
He looked as surprised as I felt.
Our company had been struggling for years, getting by on the few projects we managed to land, but nothing we did seemed to stick.
We were the definition of a sinking ship, and it was only a matter of time before we went under.
“What is this about, sir?”
“Emma, David, I’ll get straight to the point.” Our CEO leaned forward in his chair, his hands clasped together.
“I’m going to show you something that I hope will change everything for us.”
That’s a big promise for a software demo, I thought.
The screen behind him flickered to life.
He opened the web browser and navigated to our website, which looked like any other e-commerce site, full of product listings and discount banners.
But then he clicked on a product page and scrolled down to enter the chat feature at the bottom.
“Hello, how can I help you today?”
A chat bubble appeared.
He typed a message: “What will I buy next?”
A few seconds passed, and another message arrived: “Based on your recent purchases and browsing history, here are a few items you might be interested in.”
The screen showed a list of recommended products with direct links to their product pages.
“That’s an AI model we developed in house,” he said with pride.
“It generates predictive texts based on recent online activities and personal data.
It’s still in the experimental phase, but as you can see, it’s already generating accurate product recommendations.”
My jaw dropped.
I looked at David who seemed equally impressed.
“I’m sure you’re both wondering why I’m showing you this,” he said.
“I believe we can use this model not just to predict what our users will buy next, but also to forecast their futures.”
He paused to let that sink in.
“What do you mean by 'forecast their futures'?”
David asked after a moment.
“Imagine if we could predict what our users would buy in the next month or even year.
We could adjust our inventory to meet their needs and optimize our sales revenue.
We’ll be ahead of the game, leading the market instead of following it.”
His words sounded too good to be true.
There must be a catch somewhere.
My mind was already racing ahead to think of all the possible ways this could go wrong.
If our users knew what they were going to buy in the future, wouldn’t that be a self-fulfilling prophecy?
Wouldn’t that take away their agency?
And how could we even predict what someone would buy?
The more I thought about it, the more skeptical I became.
“Sir, with all due respect,” I said.
“This sounds like a fantastic idea in theory, but what you’re suggesting could have serious ethical implications.
It’s one thing to use AI to predict what products our users are likely to buy.
It’s another thing to predict what they will do in the future.”
The CEO’s smile wavered for a moment, but he quickly regained his composure.
