There are two internets right now. On one of them, AI is the most disruptive technology since the printing press. On the other, it is an overhyped slop machine that will collapse under its own weight any day now. The people on each side are absolutely certain. And they are talking past each other completely.
I find the divide fascinating because I am standing in the middle of it with receipts from both directions.
The Disrupted
I have a travel site that has been running for over two decades. Traffic is down 57%. A gardening site we run is down 40%. Google says our rankings have improved. I wrote about this in The Click That Disappeared — the decoupling of rank from traffic, the click being absorbed by AI Overviews, the twenty-year content investment being synthesized without being visited.
That is not hype. That is my P&L. I have watched a real business get hollowed out by a real technology in real time. If anyone has reason to be skeptical of AI, it is the person whose revenue it is eating.
And yet.
The Disruptor
At the same time that AI is consuming my publishing revenue, I am using it to consolidate something I have started calling my life stack.
I used to pay a web development company to build and maintain my sites. That is over. I am doing it myself now — not because I enjoy hand-coding HTML at midnight, but because the combination of my existing development background and AI tooling makes outsourcing it irrational. The economics flipped. What used to take a team and a timeline now takes a conversation and an afternoon.
I used to depend on a hosting company for infrastructure decisions. That relationship is shrinking too. I migrated to static architecture on Cloudflare. The hosting company's value proposition — managed servers, CMS updates, security patches — collapsed when the CMS itself became unnecessary.
I used to struggle with the throughput of publishing. Not because I lacked things to say, but because the labor of turning a thought into a formatted, published artifact was brutal enough to kill most ideas before they reached the page. I wrote about that in Content Creation Isn't Dead. That bottleneck is gone. I am publishing at a pace that would have been physically impossible six months ago.
Web development. Hosting. Publishing. Creative work. Code. Writing. Design. All of it pulled back in-house. All of it running through the same tool. That is the life stack consolidating.
Both Things Are True
Here is what the debate keeps missing: I am both the disrupted and the disruptor. The same technology that is eating my travel site's traffic is the one I am using to rebuild my entire professional infrastructure from scratch. I am losing revenue on one side and eliminating expenses on the other. I am watching a business model dissolve while simultaneously building new capabilities I did not have a year ago.
The people who say AI is all hype are not wrong about the slop. There is an ocean of zero-effort, zero-thought garbage being generated right now. SEO farms. LinkedIn posts that read like they were written by a fortune cookie with a thesaurus. "AI-generated content" that has never been touched by a human thought.
But the people who dismiss the entire technology because of the slop are making the same mistake as someone who dismisses the internet because of spam email. The medium is not the message. The operator is.
Slop In, Slop Out. Thought In, Thought Out.
I wrote about the Solow paradox a couple of days ago — the fact that 90% of CEOs say AI has had zero impact on productivity. And the explanation is the same one Robert Solow identified in 1987 about computers: the technology is not the problem. The humans are.
Slop in, slop out. That is real. If you hand AI a vague prompt and walk away, you will get vague output. If you use it for 90 minutes a week and change nothing else about how you work, you will see nothing. If you treat it like a parlor trick, it will perform like one.
But that is not a statement about the technology. That is a statement about the operator.
When I sit down with AI, I bring the thesis. The argument. The structure. The taste. The lived experience that no model has. I am not outsourcing my thinking — I am outsourcing the bricklaying. The architecture is still mine. And what comes out the other end is not perfect, but it is close enough to be useful and fast enough to change how I work.
The output quality is a direct function of the input quality. That is not a weakness of the tool. That is how every tool in history has worked.
The Acceleration Nobody Wants to Admit
Yes, it helps that I was a web developer before. I am not going to pretend otherwise. Having a technical background means I can evaluate the output and direct the iterations with precision that someone without that background might not have.
But here is the thing: it accelerates what I can do in domains where I have no background too. I wrote about this in AImplified — soil science, apothecary, graphic design, all learned and applied at a pace that would have been impossible without the tool. The technical background helps, but the acceleration is not limited to technical work. It is domain-agnostic. The skill that transfers is knowing how to collaborate with it, how to direct it, how to iterate. Once you learn that pattern, it works everywhere.
And it is not just me. I know people who have never written a line of code in their lives who are now building functional tools. I know people who struggled to put pen to paper for years who are now publishing. Not because AI replaced their voice — because it removed the friction between their thinking and the finished artifact.
Some will say that does not count. That if you did not type every word, it is not real. I addressed that with Victoria's question in The Architect and the Bricklayer: "Doesn't it make you sad you didn't write that?" No. Because I did write it. I just did not type it.
Do not be dismissive of the people this unlocks. Not everyone who has something worth saying has the mechanical ability — or the time, or the energy, or the executive function — to wrestle it into publishable form. AI does not give them ideas. It gives them a path from the idea to the page. That is not slop. That is access.
The Evidence I Cannot Ignore
I have created things in the past two months that I would not have believed possible a year ago. Code that runs in production. Writing that captures what I actually think. Art that moved people enough to put it on their walls. A website rebuilt from a fifteen-year-old CMS into a modern static architecture. An entire publishing pipeline that lets me go from thought to published post in the time it used to take me to open an editor.
I have taken back web development from a company I was paying. I have reduced my hosting footprint to near zero. I have eliminated SaaS subscriptions that were bleeding my P&L. I have published more in eight weeks than I did in the previous eight years. I have built things across domains I had no formal training in.
I am not naive. I know the technology has limits. I know it hallucinates. I know it requires management and judgment. I know it is not magic.
But I also know what I have built with it. And I cannot look at that body of work and conclude this is all for naught.
The Progression
What strikes me most is the sequence. First, it came for my writing — AI Overviews synthesizing my content, chatbots answering questions my articles used to answer. Then it came for my publishing workflow — the labor of formatting and shipping content collapsed to near zero. Then it covered my twenty years of web development experience with results that are, honestly, pretty good. Not perfect. But good enough that the economics of hiring a team no longer make sense for the work I need done.
Each layer fell faster than the last. The writing disruption took a couple of years to become visible. The publishing acceleration happened in weeks. The web development consolidation happened in days.
And now agents are arriving.
Agentic AI — systems that do not just answer questions but take actions, chain tasks, iterate on their own output, and operate autonomously within defined boundaries — could blow computer work into chaos. Not the distant, theoretical kind of chaos that futurists write white papers about. The immediate kind. The kind where a solo operator can spin up four parallel workstreams and ship what used to require a department.
If you are in any field that primarily involves moving information around a screen — and that describes the majority of white-collar work — the progression I have lived through in publishing and web development is coming for you. Not because I am special. Because the pattern is consistent. First the content layer gets disrupted. Then the tooling layer gets consolidated. Then the agent layer starts automating the coordination itself.
I have no reason to believe this stops at my industries. I have every reason to believe it does not.
The Divide
The people who say AI is slop are usually the people who have put slop effort into it. The people who say it changed everything are usually the people who have put real thought into it. Both are describing their actual experience accurately. They are just describing different experiences.
I have been disrupted by AI in the publishing space. I can see what it is doing in web development, hosting, and creative work. I have no reason to believe it will stop at the industries I can personally observe. The pattern is consistent. The evidence is accumulating. The consolidation is real.
And I keep coming back to the same thought: if this technology is doing what it is doing to my life stack — and I am just one person, in rural South Carolina, working with the same tools everyone else has access to — what is it going to do when the adaptation curve catches up?
The Protocol: AI is consolidating the life stack. Web development, hosting, publishing, creative work — services that used to require separate vendors, separate budgets, separate expertise — are collapsing into a single tool wielded by a single operator. The debate over whether this is real or hype is being settled not by arguments but by P&L statements. Mine shows disruption on one side and consolidation on the other. Both are real. The difference between them is not the technology. It is whether the human using it brought something worth amplifying.
