#pro ai

Posts like “I’d rather read a million words of the worst, most OOC, badly spelled Mary Sue fanfic written by a human than a single sentence of AI slop” are honestly adorable in their naivety. That’s the kind of thing you say when you still live in a world where time is infinite and homework is the only thing on your calendar.

In adult life, attention is the rarest resource in the room. I’m a designer; the numbers have been getting worse for years. There was a time when you could assume a user would give a page maybe 12 seconds before bouncing. Then it was closer to 8. Whatever the exact current figure is, the curve is going in one direction: down. You fight for seconds, not for “a million words of charmingly bad human sincerity”.

So maybe somebody should tell Tumblr’s tiny martyrs of literature that once you add work, obligations, schedules, health, sleep, a life — your free time collapses into a very small pile of hours. And in that context, “write badly, it’s necessary” stops being wholesome and starts sounding like: please donate your limited attention to my unedited draft because I typed it with my raw, suffering fingers.

“Slop” can be human. That’s the part they never want to say out loud. The internet has been drowning in it for years, long before AI: fanfic that goes nowhere, essays that say nothing, sketches that should’ve stayed in the notebook. At some point people quietly stopped doing basic quality control on themselves and started dumping absolutely everything into public space. First it was “work in progress”, then “doodles”, then literal ballpoint scribbles on graph paper that hurt to look at. All of it wrapped in the same old story: I suffered, therefore it’s valid.

But here’s the boring, adult truth: I do not care how much you suffered over a dish if it looks awful and tastes worse. No restaurant gets away with serving garbage because “the chef cried into the sauce for four hours”. My taste buds do not owe him respect. Why should reading or looking be any different?

If you put a product into a public feed and ask for my time, then yes, I expect you to:

respect my limited attention;

run at least minimal quality control on what you’re posting;

use whatever tools you have — including AI — to make it less of a waste.

The real reason people panic about AI in creative work isn’t theft, or ethics, or “the soul of writing”. It’s much simpler: AI broke the old way we assigned status.

For a long time, the hierarchy in creative spaces was built on scarcity. Very few people could produce a readable story, a polished illustration, a working game. You paid for that ability with years of practice, failed drafts, burnout, night shifts. The suffering *was* part of your credential. You weren’t just “someone who made a thing” — you were “a real writer”, “a real artist”, because it had been objectively hard to get there.

AI walked in and quietly deleted the entry fee.

It doesn’t magically turn everyone into a genius. It doesn’t fix taste, or judgment, or the ability to say something that isn’t dead on arrival. What it does is level the technical floor. The distance between “I have an idea in my head” and “there is a presentable version of it on the screen” shrank from years to hours. You still need a brain; you no longer need a decade.

Once that happens, the old status markers stop working.

Nobody cares how long you wrestled with your tools. Nobody is impressed by how much you suffered for your process. The only interesting question left is: is the thing you put into the world actually worth anyone’s attention?

AI, in that sense, is brutally egalitarian. It puts everyone on the same board and says:

I don’t care how “real” you are. Just show what’s inside.

If what’s inside is sharp, strange, alive — the tools only make it easier for other people to see it.

If what’s inside is mostly the story of how hard you tried, how much you agonized, how “authentic” your struggle was… then yes, this moment is terrifying. Because suddenly effort is invisible and only output is real.

That’s why so many “protect the uniquely human space” speeches sound more like anxiety than ethics. They’re not defending art from machines. They’re defending a world where the sentence “I worked very hard on this” could substitute for “this is actually interesting”.

Because it's not like natural intelligence is always all that great either. - Linus Torvalds

https://lore.kernel.org/linux-media/CAHk-=wi4zC+Ze8e+p3tMv8TtG_80KzsZ1syL9anBtmEh5Z40vg@mail.gmail.com/

While anti‑AI folks are busy compiling blacklists of creators who “dare” to use AI, they might want to flip the logic. At the rate adoption is going, the only scarce resource soon will be people who don’t touch it.

In companies, AI is no longer a fringe experiment, it is the operating system. By 2025, about 78% of organizations were already using AI in at least one business function, up from roughly 55% just two years earlier. Some surveys put it even higher, at around 78% of companies actively using AI in daily operations and 90% either using or actively exploring it. Among knowledge workers, generative AI has gone from novelty to default: around 75% of global knowledge workers reported using AI tools regularly in 2024, with nearly half of them starting in just the previous six months. This is not a vibe shift, it is a mass migration.

Private users are not exactly waiting politely on the sidelines either. Microsoft’s 2025 data suggests generative AI tools are used by roughly one in six people worldwide, with adoption in the Global North approaching a quarter of the working‑age population. In more targeted national surveys, like the St. Louis Fed’s 2025 study, over half of adults aged 18–64 reported using generative AI, up from about 45% the year before. Put differently: if you still think AI users are a niche, the niche is now larger than the non‑niche. (1, 2)

And then there’s the products people already live inside. Major consumer tools have spent the last couple of years quietly swallowing AI: productivity suites, search, design tools, CRM, code editors, support platforms. Analysts estimate that by 2025 about three quarters of organizations globally were already using or piloting AI in core functions and actively baking it into their offerings. On the consumer side, surveys show that for two in five people, AI integration in products is now so normal it does not even significantly change their purchase decisions. AI is no longer the feature; it is the plumbing.

Maybe that is why the “AI bubble will pop” fantasy sounds so emotionally satisfying and so historically illiterate. Yes, bubbles burst. The railway bubble popped in the 19th century; a lot of investors got burned, a lot of speculative lines stopped being built at insane speed. But trains and railways did not vanish with the bubble, they became infrastructure. The frenzy died; the tracks stayed. If the AI bubble “pops”, it is far more likely to look like that: overfunded nonsense collapsing while the actual rails — the models, the tooling, the workflows — stay bolted into everything.

There’s something unintentionally poetic about people “taking their like back” from AI art.

On paper, it’s a nothing action. A free heart icon appears, then disappears. No one loses money, no one loses housing, nothing material shifts in the world. But inside their head, it’s a grand moral gesture: sanctions imposed, sin revoked, cosmic balance restored.

That’s the real trick: the platform gives them a button and they supply all the theology.

A like was supposed to be a reaction. “I enjoyed this.” “This hit my brain in a pleasant way.” Now it’s treated like moral currency. They don’t just click; they invest.

“I endorse you as human.”

“I withdraw support because you used AI.”

“I will not let my sacred heart icon be seen near this impurity.”

Meanwhile, you can’t buy groceries with that thing. You can’t pay off your loans with Tumblr notes. Nobody has ever walked into a bank, dropped a screenshot of their “unliked AI post” and gotten better interest rates. The only balance that moves is the one in their own nervous system.

And that’s where it gets Pavlovian.

First, you teach yourself:

AI tag → disgust → must un-like.

AI confession → anger → must reblog a sermon.

AI image I initially enjoyed → shame → must rewrite my reaction retroactively, so I wasn’t “fooled.”

Repeat that loop a few hundred times and you no longer see a picture, you see a trigger. The art doesn’t pass through your senses anymore. The word “AI” hits your cortex and your body does the rest: outrage, revocation, performance, relief.

They talk about “addiction to AI.” But watch the anti‑AI crowd and tell me you’re not seeing classic conditioning:

bell rings, mouth waters;

tag appears, moral panic.

The funniest part is how tiny the stimulus actually is. It’s a tag. A disclosure line. A sentence in the caption. From the outside, it’s nothing. Inside their system, it’s a full alarm: purity compromised, identity threatened, must purge the timeline.

That’s why the “I took my like back” posts are so revealing. They’re not flexing ethics; they’re documenting their training.

Look, I saw something, I liked it, then I learned it was AI, and I punished myself for my own reaction.

Look, I’m a good dog. I barked at the right bell.

And they will call this “critical thinking”. They will call it “protecting art”.

What they are actually doing is outsourcing their taste to a label and then violently policing their own nervous system to match it.

The machine in this story isn’t the model that generated the image.