I’ve Seen Every Tech Fad for 20 Years. AI Is Different.

I’ve been doing this for over twenty years now.

In that time, I’ve watched a lot of “next big things” come through. Every couple of years, something new showed up, and everybody was absolutely sure it was going to change everything. It’d be called a “gamechanger”. You’d get a wave of hype, a wave of new “gurus,” a wave of courses, a wave of people telling you that you’d be left behind if you didn’t jump in right now.

And then… most of it faded. Some of it disappeared entirely.

So I’ve gotten pretty skeptical of hype over the years. I’ve seen it so many times that I’ve become jaded by it. When everybody’s shouting that something is revolutionary, my first instinct is to wait and see whether it’s actually useful.

AI is different. Very different.

I don’t say that lightly. This is the first time in a long time that the hype has turned out to be mostly accurate. AI has changed how I work every single day. Research, writing, troubleshooting, planning, the tech side of running a business… it’s in nearly all of it now. And I think the benefits for solopreneurs running their own online business are strong enough that ignoring it, or putting it off for “someday,” would be a real mistake.

And I’ll be honest… I didn’t always feel this way. Back in 2021, I wrote a post literally titled AI Copywriting Tools: Why I Just Say No. I’d tried one of the early AI writing tools, found it clunky and pretty useless, and worried it would just flood the internet with even more junk content.

Some of that still holds. (AI still can’t originate anything… more on that below.) But the tools have changed dramatically since then, and so has the way I use them. So I changed my mind. That’s what happens when you actually test things instead of reacting to the hype in either direction.

That’s why you’re going to see me write about it more here on this blog. I’ve already touched on it in posts like The AI Hype Is Real… But So Is the Learning Curve and Blogging in the Age of AI, and I’ve been talking about it increasingly in The WP Edge newsletter.

If you’ve been reading Blog Marketing Academy for a while, don’t worry. I’m not turning this into an AI blog. 😇 The fundamentals I’ve been writing about for years still matter just as much… building an audience, creating content people care about, making offers, running a business that actually makes money. None of that changes.

What changes is how fast you can do it. AI is a tool for speeding up the work you’re already doing and taking a lot of the mechanical grunt work off your plate.

But I know that for a lot of people, AI still feels like a black box. When I talk with other solopreneurs about it, I keep hearing some version of the same thing…

“I know I should be using it. I just don’t know where to start.”

Or: “I’m not sure I trust it.”

I get it. Both reactions make sense. So consider this the starting point for the series. No jargon, no hype. Just what AI actually is, what it isn’t, and a straight answer on the concerns that make people nervous.

Let’s dive in…

So What Is AI, Really?

When most people say “AI” right now, they’re talking about tools like ChatGPT, Claude and Gemini. You type a question into it… and it types back at you. It is basically a chat session.

That’s how almost everybody starts. But as you’ll see in a bit, the chat window is really just the front door.

Under the hood, these tools run on what’s called a large language model (or LLM). That’s a fancy name for a system that was trained on an enormous amount of text… books, websites, articles, code, you name it. From all that reading, it learned how language works and how ideas connect to each other.

How it actually works (in plain English)

You don’t need to understand the technical guts to use AI. But a rough picture of what’s going on behind the curtain explains a lot of its quirks. Here’s the short version, minus the math.

It learned by playing a guessing game. During training, the model reads an enormous pile of text and does one thing over and over: it tries to guess the next word. It guesses, checks the real answer, and adjusts itself a tiny bit. Then it does that again. Trillions of times.

There’s no database of facts inside it. Everything it learned is stored as billions of tiny numeric settings that got nudged during all that guessing. So when it answers you, it isn’t looking anything up. It’s rebuilding an answer from patterns. That’s why it’s usually right… and occasionally wrong with total confidence.

It writes the answer one word at a time. At each step, it’s asking itself, “Given everything so far, what’s the best next word?” There’s a little randomness built in, which is why you can ask the same question twice and get two different answers.

It was taught to act like an assistant. A raw model just continues whatever text you give it. The helpful, polite, answer-your-question behavior comes from a second round of training, where people rated its responses and it learned what a good answer looks like.

It doesn’t learn from your conversations. Each new chat starts fresh, aside from memory features that some apps bolt on top. So if it seems to “forget” what you told it last week, that’s why.

The mental model that works for me

Here’s how I think about it…

AI is a very well-read assistant. Fast, tireless, and eager to help. It has read more than any human ever could, and it can write, summarize, explain, brainstorm and even write code.

Notice I said well-read. Not original. Everything it produces is built out of patterns from what it has already seen. It’s remarkably good at recombining those patterns, but it isn’t coming up with anything from scratch. (More on why that matters in a minute.)

And it just met you five minutes ago.

It doesn’t know your business. It doesn’t know your customers, your history, or how you like things done. Not unless you tell it.

And it has one quirk you always have to keep in mind. When it doesn’t know something, it’ll sometimes make something up… and say it with total confidence.

Once you think of it that way, most of the confusion clears up. You wouldn’t hand a brand-new assistant a vague one-line request and expect a perfect result. You’d give them some context. You’d check their work, at least at first. Same deal here.

What AI Is Not

A lot of the frustration people have with AI comes from expecting it to be something it isn’t. I find that too many people put this blind trust into AI… just thinking it is some all-knowing digital God where you can trust everything it says. But, here’s the truth…

It’s not a search engine. Many of these tools can search the web now, which helps. But by default, they’re generating answers from what they learned during training, and that training has a cutoff date. Ask about something recent and you may get outdated information. Grok is actually often more current in terms of information because it taps into the X firehose.

It’s not always right. The industry term for AI making stuff up is “hallucination.” It happens less than it used to, but it still happens. Fake statistics, quotes nobody said, a plugin setting that doesn’t exist. All delivered in the same confident tone as the correct answers.

It’s not a person. It can sound remarkably human. It’ll be polite and encouraging, and it’ll apologize when you point out a mistake. But it has no stake in how your business turns out. That’s your job. It doesn’t feel bad… nor does it have any personality. Any hints of humanity are just patterns adapted into its training.

It’s not creative. Not the way you are. This one gets lost in all the hype.

A human being can come up with a concept out of nothing. A new angle, a strange idea, a connection nobody’s made before. It’s why humans can create art, invent things, and literally imbue the environment with beauty. That’s something unique to human beings, and it always will be.

AI can’t do that. It’s a pattern recognition engine. It gives you the most likely, most reasonable response based on everything it has seen. That’s incredibly useful. It’s also a real limitation.

If everybody lets AI do their thinking, everything drifts toward the same common denominator. Same blog post structure. Same tired phrases. Same “10 tips” everyone else has. You’ve probably already noticed this online… a whole lot of content that all sounds vaguely the same.

The variety comes from people. Your creativity, your experience, your quirks, your opinions. In a world full of AI-generated sameness, that becomes more valuable, not less.

It’s not a replacement for your judgment. Or your voice, for that matter. AI is great at producing a solid first pass. Deciding whether that first pass is actually good, and making it yours, is still on you. You are the decider.

And it’s not one single thing. ChatGPT is made by OpenAI. Claude is made by Anthropic. Gemini is Google’s. Grok is now part of SpaceX. And there are plenty of others. “AI” is a generic term, kind of like “the Internet”… it’s really a whole lot of different tools and models, in a very competitive and fast-moving field.

What It’s Actually Good At

I’ll get much more specific later in this series, but here’s the short version for people running a small online business.

AI is good at getting you past the blank page. First drafts of emails, blog outlines, product descriptions, social posts.

It’s good at summarizing. Long articles, meeting transcripts, that terms-of-service page nobody reads.

It’s good at explaining things. Paste in a confusing error message and ask, “What does this mean, in plain English?” Frankly, this one alone can save you hours of forum digging.

It’s a good brainstorming partner. Headline ideas, product names, ways to reword that paragraph that just isn’t working. Remember, it’s recognizing patterns, not being creative… but it may spot patterns you missed. That’s what makes it useful for brainstorming.

And it’s good at organizing mess. Hand it a pile of scattered notes and ask it to turn them into a plan.

All of that is useful. But it’s still you and a chat window, copying and pasting back and forth.

And honestly? That’s just the beginning…

It’s Not Just a Chat Window Anymore

Here’s where things get really interesting… and where I think most solopreneurs need to end up.

AI can connect to your other tools and actually DO things. Not just tell you how to do them. Actually do them.

Hook it up to your email, and it can sort your inbox and draft replies. Connect it to your calendar, and it can find a time for a meeting. Give it access to your WordPress site, and it can update a post, check which plugins need updates, or dig into why a page is loading slowly. Connect it to your project management tool, and it can create tasks and keep things organized.

You’ll hear this called “agentic AI” or “AI agents.” Ignore the buzzwords. The idea is simple. Instead of AI telling you what to click, it does the clicking.

Honestly, I barely ever just “chat” with AI now. Sure, I’m using chat windows, but I’m having it DO things day in and day out. From a series of ongoing business tasks (like email triage, handling site ops, research tasks, etc.)… to directly debugging WordPress sites and even, at times, building out full features. That’s where the real time savings are… handing off the repetitive, step-by-step work that eats hours every week and doesn’t need my brain to get done.

It also raises the stakes a bit. An AI that can only chat with you can’t break anything. An AI that can change your website can. So the same rules apply, only more so. Start small, keep backups, and check its work before you trust it with anything important. Learning how to use AI without breaking anything or exposing sensitive data is a skill in itself.

That’s where this series is headed. We’ll start with the basics, but the goal is to get you from “chatting with AI” to “AI doing real work in your business.”

Not sure yet what you’d actually hand off? I built a free tool that walks you through your business and sorts your tasks for you… takes about 10 minutes.

The Concerns (and My Honest Take on Each)

Now for the part a lot of AI content skips entirely. There are real reasons people are hesitant, and they deserve more than a “don’t worry about it.”

“Won’t it just make stuff up?”

Sometimes, yes. See above.

My rule is simple. Anything that matters gets verified. I often tell it to run live searches to fact-check itself. And sometimes my years of experience just tell me the AI is plain wrong… and I tell it so. 🤣 When AI is working in my business, it can look but not touch. Nothing changes without my express approval.

This isn’t that different from working with a human, really. You don’t blindly publish what a new contractor hands you on day one either.

“What about the privacy of what I type into it?”

This is a fair question, and the answer depends on the tool and the plan you’re on.

Here’s the part that surprises people. On the personal plans of all three big tools, your conversations can be used to train future models by default. That includes the paid personal plans, not just the free ones. You have to go turn it off.

The good news is it only takes a minute:

• ChatGPT: Settings → Data Controls → turn off “Improve the model for everyone.” (OpenAI’s instructions)
• Claude: Settings → Privacy → turn off the model training toggle. (Anthropic’s announcement)
• Gemini: turn off “Keep Activity.” With it on, a sample of chats can also be read by human reviewers. (Google’s privacy hub)

Business and team plans are a different story. Those generally don’t train on your data by default. Either way, these policies change, so it’s worth checking your settings every so often.

Beyond that, my rule of thumb: don’t paste in anything you wouldn’t be comfortable emailing to a stranger. Passwords, customer personal information, financial account details. Keep those out. For everyday work like drafting a blog post or figuring out a plugin conflict, there’s really not much to worry about.

And if privacy really matters for what you do, you can run an AI model entirely on your own computer. There are open models you can download and run locally with free tools like Ollama or LM Studio, so nothing ever leaves your machine. 100% private. The catch is it takes serious hardware (and the price of computers that can do it well has shot through the roof), and the models you can run at home aren’t quite as capable as the big cloud ones yet. Not where I’d tell a beginner to start… but it’s a real option, and I may dig into it more later in this series.

“What about the environmental impact?”

This is probably the concern I hear most from thoughtful people, so it deserves an honest answer instead of a brush-off.

At the industry level, it’s real. AI runs in massive data centers that use a lot of electricity and water, and the International Energy Agency projects that data center electricity use will roughly double by 2030. Some perspective helps, though. Globally, data centers are expected to account for only about a tenth of the growth in electricity demand… less than electric vehicles, and less than air conditioning. (Here in the U.S., the share is bigger, so it’s a fair concern here.)

At the individual level, the numbers are small. Google’s published figures put a typical text prompt at about as much energy as watching TV for less than nine seconds, plus about five drops of water. That’s roughly the electricity it takes to drive an EV five feet. OpenAI’s number for ChatGPT is in the same ballpark.

To be fair, those are the companies’ own numbers, and they’re for simple chat questions. Having AI run multi-step tasks for you can use hundreds of times more. Even so, heavy use still adds up to far less than what most of us use driving or flying.

As for the power itself, many AI companies are building their own supply… nuclear, solar, and for now, a lot of natural gas. In the short term, that demand is pushing power bills up in some regions. Longer term, I’m optimistic. When the richest companies on Earth suddenly need enormous amounts of reliable power, they pour money into finding cheaper ways to make it. My bet is that ends up making energy cheaper and cleaner for everybody.

And the water? Most of it evaporates during cooling and comes back down as rain… just not necessarily in the same place. So the real issue is local water stress, like a data center in the desert. Not AI “drinking up the world’s water.”

So my take? The big-picture concern is valid. Your personal use for business tasks is a very small slice of it… smaller than a lot of things most of us do every day without thinking twice. If it still weighs on you, that’s a reasonable thing to care about. Use it deliberately for work that matters instead of for everything. That’s a good habit anyway.

“Isn’t using AI kind of cheating?”

Is using spellcheck cheating? A calculator? A template?

It depends on how you use it. If someone publishes generic AI output with nothing of themselves in it, that’s a problem. But it’s a problem because it’s lazy, bland content, not because a machine was involved. Readers can smell it either way.

Using AI to handle research and rough drafts so you can spend your energy on the thinking, the stories and the opinions? That’s just using a good tool. It’s how I use it.

“Is it going to replace me?”

For solopreneurs, I think this one flips around.

You’re not the employee being replaced here. You’re the business owner who suddenly has access to help you could never afford to hire. And the things that make your business yours, like your relationships, your perspective and your experience, are exactly the things AI can’t do.

“It seems too technical for me.”

If you can send a text message, you can use AI.

There’s no code involved. No setup, for the basic tools. You just describe what you want in plain English. The only real skill is being clear about what you’re asking for… which, frankly, is the same skill you use every time you delegate anything to a human.

Why This Matters So Much for Solopreneurs

Here’s the part I’m genuinely excited about.

When you run a one-person business, you wear every hat. Marketing. Content. Customer support. Bookkeeping. Tech. Planning. And you don’t just wear those hats… you personally do the work under every one of them, from coming up with the idea to executing every step.

Let’s be real. That doesn’t work very well. There’s only so much mental throughput to go around, so things slip. Some parts of the business get done halfway. Some don’t get done at all.

This is where I think AI makes the biggest difference. I think of it like the conductor of an orchestra.

A conductor doesn’t play every instrument. They probably couldn’t play half of them if you handed them over. And yet the whole orchestra follows their lead. The conductor’s job is the big picture… the interpretation, the timing, how all the pieces come together into one performance.

A solopreneur can work the same way. You write the song. You come up with the plan, the ideas, the direction. Then you orchestrate how it all gets played out across your business. And more and more, AI can handle the mechanics of actually playing it.

You’re still the only human in the business. You still wear every hat. But you’re no longer stuck doing all the grunt work under every one of them, so you get to spend more of your time thinking, deciding and creating… the stuff only you can do. That means a solopreneur can grow in ways that used to require hiring a team.

I’ve been building this into every part of my own business, across all seven functions every business has, and I’ll walk you through how I’ve set that up later in this series.

If you want to see how that plays out across your own business, the AI Delegation Map does exactly that. It walks you through all seven functions, sorts the tasks you actually do, and shows you the three best places to start.

Your First Step

Don’t try to learn everything at once. That’s how people end up overwhelmed and doing nothing.

Instead, pick one tool. ChatGPT, Claude or Gemini… they all have free versions, and any of them is fine to start. Then give it one real task from your actual to-do list today. An email you’ve been putting off. An error message you don’t understand. A blog post you can’t figure out how to start.

See what it does. Push back if the answer’s not quite right. Ask it to try again.

That’s it. That’s the whole first step. Connecting it to your tools and having it do real work comes later. First, just get comfortable talking to it.

In the next post in this series, I’ll go deeper into which of these tools is actually worth paying for… and when the free version is plenty.