ChatGPT Feels like 2005
By Ashley Graham-Wilcox
Remember 2005? Facebook was for college students, YouTube had just launched, and blogs were the most exciting thing on the internet (RIP Livejournal). While the internet was old enough that demands had to be met and there was expertise to be found, the word “algorithm” was for math majors, and every brand was figuring out how to be online…
From there came the first wave of data-driven marketing.
I was a brand new communications grad, working for a tech startup, wondering how Marshall McLuhan fit into click-through rates. The data became instructions, and it was so helpful: Everything could be optimized, from headline length to website content.
(Ironically, on that fork of my career, I was part of more than one conversation between startup founders and those who’d “bought them out,” where the most techy of tech guys – those who’d literally invented programming languages or spearheaded major brands’ first online presences – balked at their autonomy and voices getting lost in the newest wave of communication.)
The era I’m talking about was the dawn of CAN-SPAM laws; I remember the conference room I sat in to learn about them.
Grateful for the instruction manual, I followed the parameters as given by my data-informed supervisors and wrote well-performing emails and learned about search optimization.
Fast forward to now, and I can see similar formulas creeping back in. Scroll through your inbox or feed. You can practically hear the algorithm humming:
The clean headline with an emoji at the end
The “three reasons why” hook
The slightly-too-attuned, AI-generated tone (“Hey, fam!”)
We know a Canva layout when we see it.
Everything is starting to feel the same, and the formula feels … icky? predictable? annoying? helpful? Honestly, I feel all of those things.
It’s obvious when ChatGPT has written something: The rhythm is tidy, the texture scrubbed away. Did we accidentally train the internet to sound like a copywriter training to be a radio host?
So, the tools are faster, smarter, and way more affordable, persuasive, and available. ChatGPT, Copilot, Meta AI, take your pick. They promise to save time and spark creativity, and sometimes they do. But they also smooth out the edges that make communication interesting. They nudge everything toward the same voice, the same look, the same generic vibe. It’s 2005 again: Everyone chasing the same hacks, the same data points, the same “believe me” tone. Only now, instead of learning, we’re outsourcing to machines.
My 6th grader occasionally takes a break from saying “6-7” to mention “the chat.”
Back in the AIM days, “the chat” was the place to be. It was the highway of my social world, and how I presented myself. I’d toggle between windows, sustaining fifteen conversations with people who all somehow felt like my closest friends. And don’t forget the Away Messages. (As someone who’s now married to my long-distance college boyfriend, I will never shade early 2000s AIM.)
Now “the chat” means something else entirely: For my kid, it’s a group of people they want attention from. For me, it’s a WhatsApp thread about who’s bringing snacks to the soccer game, the text chain that plans the class Halloween party, the side group chat where three parents confirm what the main group chat meant.
There are so many chats, and none of them are about my voice.
The internet was supposed to simplify connection, but it keeps fracturing it. With infinite ways to communicate, we’re constantly figuring out what to say, and how.
Ah, there you are, Mr. McLuhan.
The technology changes, but the pattern stays the same. Each new medium promises clarity and connection, and brings new noise to navigate.
If the early 2000s were about learning to post, the 2020s are about learning to discern. The question isn’t can we use AI; it’s should we—or maybe, when should we? When is it helping us clarify, and when is it flattening our voice? When is it saving us time, and when is it stamping out authenticity?
The environmental and educational impacts of AI are undeniable. And honestly, I don’t know how to parent my child through middle and high school to become a curious college student and world citizen in the age of AI without just…forbidding using it. SparkNotes has never seemed so benign.
The internet — where we all live — doesn’t need more content. It needs more curiosity, more imperfection, more human fingerprints that we can all see.
So yes, digital marketing in the age of ChatGPT feels like 2005: Weird, thrilling, over-optimized, and full of possibility. I’m trying to believe that buried beneath all the sameness, the next version of authentic communication is trying to break through.
