You know what's ironic about the most recent AI revolution? It's pushing us to be more... human.
I've spent decades watching SEO evolve from keyword stuffing (1997 Seth is 100% guilty, and 1999 Seth would like to formally apologize to the entire internet) to link farms, to content mills, to whatever "guaranteed page one ranking" trick was making the rounds on Black Hat World or WebmasterWorld. Half of it read like a recipe for success, and the other half read like a recipe for malware.
Then generative AI showed up and started producing content faster than any team of humans ever could.
And that's when the weird thing happened: authentic human expertise got more valuable.
Not as a theory. I watched it happen in real time with clients. Rankings got harder to hold with generic "SEO content." Meanwhile, the messy posts written by real humans - warts, opinions, and all - started pulling more weight. It felt like the internet got tired of being politely lied to.
The pattern we keep repeating
This has happened before, just with different costumes.
Small businesses used to have a human answer the phone. Then automated phone trees arrived, promising "efficiency." We quickly learned that:
- "Do y'all got olive oil back in stock?"
- "We're closed because the power went out."
- "The delivery guy just backed into the sign."
...are not things a phone menu handles very well.
AI content is doing something similar. It's great at scale, but it's often bad at the messy, real-world edges that people actually care about.
From Yahoo directories to robot overlords
Back in the mid-90s, Yahoo was a curated directory. Humans read websites and categorized them. It didn't scale, but it worked. Really, how could the internet ever possibly scale to be bigger than a few thousand domains, right?
Then Google and PageRank changed the game. Suddenly, search was math-driven, link-driven, and optimized by people who enjoyed algorithms a little too much. The SEO arms race became predictable:
- Google updates the algorithm
- SEO folks find a workaround
- Repeat
Now AI (LLMs) has landed, and the web flooded with articles that were correct but useless. Accurate and boring. And when everyone uses the same models, the sameness becomes a signal. Google's job gets easier.
Here's the old playbook vs the new one, no fake code required:
- Old SEO: write for robots - win rankings - hope humans tolerate it
- New SEO: write for humans - earn trust - robots notice
Humans are back (but still use headings, please)
"Optimizing for humans" doesn't mean technical SEO is dead - I believe in this so much I even maintain TWO technical SEO tools. It means it's back to its original job: helping people use your site.
Keep the basics:
- clean structure and headings
- fast pages
- useful internal links
- readable URLs (not the ones that look like your cat walked across your keyboard)
Then focus on what AI still struggles with:
- Real experience: the "gotchas" you only learn by doing.
- A consistent voice: bots can't clone your weird brain yet.
- Fresh reality: AI knows what worked in 2024. You know what's breaking today.
- An opinion: AI loves to ride the middle of the road, humans are more opinionated. Yay Python. Boo NodeJS. Hiss.
And yes, there's a dark side here too.
I've watched quite a few sites "solve SEO" by shipping thousands of AI pages in a short amount of time. I still feel like this may work, but you'll attract more bots visiting pages - and filling out forms - than humans.
If you need Google's blessing in writing, they've published guidance on helpful content here:
Creating Helpful Content
SEO is dead. Long live SEO.
The paradox is simple: once content got cheap, trust got expensive.
So publish less junk. Say real things. Share scars. Update old posts when you learn something new. And if you're going to use AI, use it like a power tool-not like a personality replacement.
-Sethers