It's not that you have too many options now. It's that AI removed the friction that used to filter them for you.

Someone sent me a 94-page guide last week. Forty-eight ways to make money online, each one broken down with real startup costs, honest timelines, and a step-by-step blueprint. Ebooks, courses, templates, print-on-demand, custom GPTs, dropshipping, affiliate marketing, micro-SaaS, renting out your spare GPU capacity.

It was thorough. It was well-researched. Parts of it were genuinely useful.

But honestly, it made my head spin. When I finished it, I felt worse than when I started.

Not because the information was bad. Because for the first time in my career, all 48 of those paths felt equally within reach. I could picture myself doing any of them, starting this afternoon. That feeling should have been exciting. Instead, it felt like standing in a hallway with 48 doors and no sense of which one I'd actually walk through.

I want to be fair to it. Most guides like this are recycled hype with no numbers behind them. This one had real ones. It didn't promise passive income for no effort. It flagged scams. It told the truth about how long things actually take.

The problem wasn't the content. It was the premise. A guide that covers all of these business ideas with equal seriousness isn't really teaching you to build a business. It's teaching you that building a business is a menu, and the only decision left is which item to order.

Ten years ago, that menu would have filtered itself. You couldn't seriously consider dropshipping, audiobook narration, and micro-SaaS development in the same afternoon, because each one required a different skill you didn't have and would need months to build. 

The friction did the choosing for you, whether you liked it or not.

That friction is mostly gone now. A recent industry analysis put it plainly: AI has lowered the barrier to launch, but raised the bar to compete. The first half of that sentence is what everyone talks about. The second half is the part nobody sits with.

I don't think AI made 48 businesses possible for me. I think it made 48 businesses feel possible for me, which is a different thing entirely. Writing a script, designing a logo, drafting a sales page, building a simple chatbot, editing audio into a passable voiceover: AI can produce a rough version of all of it in an afternoon. What it can't produce is the year of showing up after the rough version stops feeling exciting.

Roughly half of small businesses now say AI pushed them to seriously consider a path they'd never have weighed before. That's not a small shift. It means the natural gatekeeping that used to happen before you ever opened a laptop. Do I actually have the skill for this? That question has all but disappeared.

The gatekeeping has to happen somewhere else now. It has to happen in your judgment, because it's no longer happening in your capability.

There's a famous experiment in consumer psychology, usually called the jam study. Researchers set up a tasting table with six varieties of jam on one day and twenty-four on another. The bigger display drew more browsers. But the six-jam table converted roughly ten times as many buyers as the twenty-four-jam table did.

More options didn't produce more decisions. It produced more looking and less choosing.

That's decision fatigue, and it's been studied for two decades in contexts far less complicated than "start a business." A guide offering 48 legitimate paths isn't generous. It's the twenty-four-jam table, dressed up as thoroughness. Every additional viable option doesn't add opportunity; it adds a reason to keep researching instead of starting.

I've lived this one personally, long before AI made it worse. I've sat with a list of possible directions and mistaken the length of the list for progress. It isn't. A list is just a list until you cross most of it out.

Interestingly, the guide's own closing page said almost exactly this. After 48 detailed paths, it ended by telling the reader to shortlist three, pick one, and ignore the other forty-seven for ninety days.

That's the right instinct buried at the back of a document built to work against it.

If I were building that filter for myself, it wouldn't start with the list of paths. It would start with three narrower questions.

Those three questions eliminate most of a 48-path list immediately, not because the other paths are bad, but because they were never actually available to you specifically.

A path you have no existing skill in, no audience for, and no real patience for is a path you were never going to finish, no matter how good the blueprint looked on page 40.

The advice showing up most often from people who've actually launched something with AI tools this year isn't about which tool to use. It's simpler than that: talk to real potential customers before you build anything.

AI makes building faster. It does nothing to make learning what's actually worth building faster. Those are two separate skills, and the second one is the scarce one now.

Most online businesses still fail in their first year, by a wide margin, and that was true before AI and remains true after it. Speed of production was never the bottleneck. Speed of building the wrong thing well is not an improvement over slowly building the wrong thing. It's the same failure with a shorter runway.

I closed that 94-page guide and didn't pick any of the 48 paths. What I actually needed wasn't inside it. I needed a way to filter the list down to one or two things that fit what I already know, what I already have an audience for, and what I could actually stay with past the point where it stops being interesting.

AI didn't create the problem of too many choices. It just removed the last thing that used to make the choice for us. The friction is gone. What replaces it has to come from somewhere, and it isn't going to come from a 49th path.

It comes from getting better at deciding, before you ever open a new tab to start building.

© Copyright David Wakeman.  All Rights Reserved