To the critics of AI

Some people saw an AI generate an image with a few visual hiccups—warped hands, odd reflections, slightly cursed proportions—and immediately declared, with the confidence of someone who once updated an app, “AI is shit.” Case closed. Progress cancelled. Please return all future thinking to the front desk.

This is a wild standard when you zoom out. A toddler draws a person with floating limbs, melting faces, and a dog that may also be a chair, and we don’t declare crayons a failed technology. We hang it on the fridge. We say it’s “expressive.” We nod like we understand modern art.

But an AI—an infant intelligence that learned by inhaling a frankly alarming amount of human imagery—gets tripped up on hands, reflections, text, or perspective, and suddenly it’s garbage, dangerous, and proof that civilization peaked when DVDs had menus.

These glitches aren’t even “mistakes” in the human sense. They’re artifacts of probability colliding with the weirdness of reality. Hands are complicated. Mirrors are liars. Text in images is chaos. Perspective is a suggestion, not a rule. Humans break these things constantly, then act betrayed when a model trained on our visual chaos occasionally produces something… creatively incorrect.

Judging AI by early image quirks is like dismissing aviation because the Wright brothers’ plane was drafty. “Sure, it flies, but I noticed a vibration, so walking forever seems safer.”

And then we get to the real outrage: fake content. Deepfakes. Misinformation. People doing sketchy things with shiny new tools. Suddenly it’s not just “AI is bad,” it’s “AI did this to us.”

Except it didn’t. Humans did.

AI didn’t invent deception. We’ve been beta-testing that feature since rumors required face-to-face delivery. We gave Photoshop to tabloids, the printing press to propaganda, social media to trolls, and comment sections to people who confuse confidence with correctness. AI didn’t wake up and decide to ruin discourse. It got handed to humans, and some of them immediately asked, “How can I make this worse?”

Blaming AI for fake content is like blaming a camera for lying in court. The tool didn’t commit the act. A person aimed it, edited it, posted it, and hit “share” with purpose.

The comparison that really sticks is this: saying “AI is shit because it makes visual errors and people misuse it” is like saying “the internet is useless because spam exists.” The noise was never the point. What came after was.

History is remarkably consistent about this. It doesn’t remember the early glitches. It remembers who mistook growing pains for failure—and who couldn’t tell the difference between a powerful tool and the humans determined to misuse it.