The IndieWeb is Wonderfully Dionysian · brennan.day

🌻 Welcome to brennan.day! I respect your decision to not use JavaScript. You can read here for more information about what functionality is disabled and why.

Skip to main content

'Trompe l’Oeil of Posted Notices and Prints' by Martin Cerulli. 1735/1765 | Art Institute of Chicago (edited by the Author)

A few months ago I wrote about the Piss Average problem, a term I coined to describe the phenomenon of the Internet being filled with the average, mediocre work created by genAI until every major platform becomes saturated and poisoned by it. I named it after the literal yellowing of images generated by ChatGPT—which sadly seems to have been fixed, and most have probably forgotten about this quirk by now.

I wrote that one of the biggest, existential issues with genAI is the crisis of faith—having no idea if the work you're looking at was produced by a person or an LLM. Eventually, this calcifies into a vigilant paranoia and questioning anything that's too standardized, too neat and tidy. I've gone on record saying this lack of trust is an awful way to interface with the web, creativity, and art as a whole. A compulsive, gnawing knowing.

Luckily, I've found that as time goes on, this is by-and-large a non-issue in most of the Internet I inhabit. There are really only a few platforms where I see obviously genAI cookie-cutter content: LinkedIn, Meta ads, and Medium.

LinkedIn , I think, would be a surprise to nobody. There has always been an uncanny, hollow culture on the platform driven by the worst impulses of late-stage capitalism. People are dedicated to maximizing reach and leverage, especially if they're unemployed and in sincere financial need for a job. Jobseekers would rightly do and say anything that might boost their chances. A 2024 study by Originality.AI analyzing nearly nine thousand long-form LinkedIn posts found that as of late 2024, over half were likely AI-generated. By 2025 that number was still holding, spread across industries from tech to healthcare to finance. The same study found that genAI posts received 45% less engagement than human-written ones.

Meta is a little more interesting, since I've completely stopped seeing genAI images or videos except for advertisements. And these are always awful—low-quality renders with uncanny skin, monotone voicework, and swimming with artifacts. In 2024 alone, more than 15 million ads were created using Meta's AI tools by over a million advertisers worldwide. It befuddles me. It seems as though genAI ads deliver ROI lifts. But a higher click-through rate on a visually debased product is still spam with better aim. The difference between a phishing email that works and a phishing email that doesn't. The metric succeeded and the commons became uglier. The entire platform (and the Internet, of course) would be far better off if AI-generated advertising were banned outright.

But Medium ? That's the most disheartening. Despite having a clear AI content policy, I have seen people time and time again violate and offend. It's one thing to use genAI imagery throughout your article—this is allowed, argued as supplementary garnish to the writing. Regardless, it is in poor taste. But it is another thing entirely to write articles with genAI.

Even worse is when I get vague, pseudo-intellectual comments rife with em-dashes and the same robotic colloquialisms cycling through like a washing machine. Maybe it's because I've been working with genAI since GPT-2 in late 2019, but the smells are strong and obvious. I'm going to share a few examples, names redacted:

There's a particular kind of quiet that comes after writing out a truth I was too scared to say out loud, and I've found that rhythm and meter often bridge the gap where ordinary logic fails.

Sir, this piece on loss is profoundly moving and deeply human. I was struck by the honesty in your words and the quiet strength that runs through the narrative. You've captured grief not just as pain, but as something layered, lingering, and strangely reflective. The way you revisit loss without exaggeration makes it even more powerful. It feels intimate, almost like a conversation with memory itself. Your writing carries a rare emotional depth that stays with the reader long after finishing. I truly appreciate the sensitivity and clarity you've brought to such a difficult and universal experience.

Beautifully written, there's a quiet honesty here that makes "home" feel deeper than a place. It captures belonging, memory, and identity in a way that lingers after reading.

In truth, out of everything, this is the most heartbreaking to see. Do you see what I mean?

Perfectly calibrated warmth. The brief feeling of being understood without understanding anything. I've seen people write about their divorce, their late mother, or a home that can't be returned to. And the machine replies and says "rare emotional depth."

I write this with as much urgent honesty as I...