How I actually decide what to take on
Once you start figuring things out and realize you can basically build almost anything at any level of complexity, the real question shows up:
what to take on how many projects you should run and which ones are worth pulling off the ground and which ones aren't
And nobody can tell you that. You decide it yourself.
I tried to figure out the criteria I use to pick projects.
Honestly — I don't know.
I don't have a clear system that I run them through and logically decide: this one I take, this one I don't.
But one thing I understood very clearly:
you can't take on everything you want to
because you'll just drown
In practice it looks like this.
You only have to talk to people you know and explain what you can do — and right away it starts:
this can be automated this can be simplified this can be redone
And then you don't account for the fact that one thing pulls in another, then a third — and suddenly it's a full-blown project.
And where's the money — who the hell knows. Maybe there will be. Someday.
There are projects you latch onto just because the idea is great.
You start digging in — and the deeper you go, the more you see: how many hidden traps there are, how complex the logistics get, and how long the whole thing will take to pay off.
And then there's another situation.
You talk to people and you can see they already have a working process, but it's crooked.
People just can't keep up. They turn down clients.
And you realize that if you tweak it just a little — the person will simply start earning more.
Not because sales went up. But because they stopped losing money.
And things like that, you feel them right away.
In parallel you spot a ton of places where you could help.
You look at sites, processes, analytics — and you see it's a complete mess.
But at the same time you don't want to push yourself on anyone. If people don't need it — they don't need it. You move on.
And at some point you realize one more thing:
you can't pull all this off alone.
You start thinking: I need to bring people in, explain things, delegate.
And right away you hit another problem: everyone who knows how — is already busy. Everyone who doesn't — needs to be taught. And that takes time too, which you don't have.
There's also a separate type of project.
You can see there's no money in it. None at all.
But something inside still pulls you in.
That's where I am right now with video generation. The logic there is complex, you yourself don't fully understand what's happening, the agents live their own lives, and you can't always keep up with them.
But you keep coming back and polishing it. Because you feel there's potential in it. Even though no one guarantees anything will come of it.
And there are the "quick and dirty" projects.
You throw it together in a day or two, play with it, take a look — and realize it's a toy. And you just let it go.
I don't pick projects "the right way." I'm learning not to grab at everything. And it looks like that's the first skill you actually need.
The second — understanding that you can't pull this off alone.
If you're in the same place — you already know what I mean.