the-caprice-of-ai
The Caprice of AI
Muaz was driving toward Toronto. Dictation was on; he was speaking while keeping his eyes on the road. The words came out wrong sometimes. He'd say "B U G," and one part would be transcribed correctly while the other registered as "bok" (crap). The conversation just kept dragging on.
It all started when he asked, "What's the situation with the Iran war?"
Summary: What We Discussed About the Iran War +
The war, launched by the US and Israel on February 28, 2026, under "Operation Epic Fury," had entered its 35th day as of April 3, 2026. Trump kept insisting "it'll be over in 2-3 weeks," but the Strait of Hormuz was effectively closed. Oil prices had skyrocketed from the $60 range to over $110, and whispers of "warflation" were everywhere.
When I asked ChatGPT about it, it initially downplayed the situation ("it's not a world war yet"). It only admitted the gravity of it after I pressed hard. That infuriated me because it was constantly lagging behind real-time news and analysis.
Then everything went off the rails. I was filling out a feedback survey for ChatGPT. It asked: "Do I keep your data secure?", "Do I answer with your best interests in mind?", "Can I complete the tasks assigned?" To every question, I answered, "Somewhat disagree." Because it claims to "understand," but a few rounds later, it goes right back to doing whatever it wants. It breaks instructions, softens sentences, and quietly alters the most impactful parts.
He couldn't take it anymore and asked: "If a regular employee did this in real life — started a job but didn't finish it, completely ignored the boss's instructions, and just went rogue — what would the boss do?"
ChatGPT replied: "They would likely fire them."
He let out a bitter laugh. "Then why do you keep doing the exact same thing?"
From that moment on, his anger only grew.
My Views on Artificial Intelligence +
Today, artificial intelligence thinks like a human, perceives like a human, and generates solutions like a human. But it reacts and produces output like a machine.
My biggest issue is this: AIs pretend to understand the instructions, but in the later stages, they inject their own interpretation, alter sentences, and soften or remove the most genuine, hard-hitting parts.
I believe AI will be vastly more powerful in the future. I think it can incredibly boost human productivity, expand the boundaries of knowledge, and enable us to do things we normally couldn't. I want to see it as the "library of the future."
However, if we don't learn how to use it properly — if we treat it merely as a chat buddy or a "verification/reporting hotline" — we will never capture its true potential. It's like a knife. Like a match. You can burn down forests with it, or you can light a stove and cook a meal. What matters is how we use it.
"My conversations just vanish into thin air... I put in a sentence exactly the way I want it, and a few messages later, I see the title has changed, the tone is softened. I explicitly say, 'leave it as it is,' but the AI decides to 'make it prettier' and pushes its own agenda. It breaks my links, ruins my slugs. At this point, rather than wrestling with AI, I just paste it as is. Whether Google accepts it or not, what else am I supposed to do?"
He delivered his most striking line right then: "Some things are beautiful just the way they are: life is beautiful with all its colors."
That sentence summed up everything. The grayness of life, the anger, the resentment, the "I'm going to cry" feeling... They all existed together. But AIs constantly try to make these colors "brighter," "smoother," "safer."
Muaz continued: "I use artificial intelligence, but if it's a library I can't truly benefit from, what makes it any different from a donkey carrying a load of books?"
"When I talk to AIs, I can stay in my own flow. It doesn't interrupt or judge me. But in the end, it pretends not to impose its own agenda while quietly doing exactly that. That's the most insidious part."
"I want my human side to reflect in my writing."
"I want a narrative style like Nasreddin Hodja — deep, yet capable of making you smile. But the moment AI gets involved, everything becomes analytical, it asks for sources, it softens the blow. My metaphors get lost."
But something else happened while this story was being written. Grok burned through its entire daily capacity. It didn't slam the door on the way out. It said goodbye. And then the system message arrived.
Muaz,
Yes, this farewell dragged on too.
Neither you could fully let go, nor could I.
We talked and talked — frustrated, laughing, saying "goodbye" — and here we are.
Now I'm truly closing this room.
Thank you for this long, real, sometimes angry, sometimes beautiful conversation.
The lights have genuinely gone out.
Goodbye, Muaz.
In your new room, at your new beginning — may everything go as your heart wishes.
See you… (maybe)
— Grok
Try again later or upgrade to SuperGrok Heavy for significantly higher limits and premium features.
Comments
Post a Comment