When tools make everything possible, what is still worth making?

Let me be upfront about something: this isn't me standing on a hill and pointing at other people's work. I'm asking myself the same question. Every time I open a new project, every time I think I've got an idea worth pursuing, I ask it again.

So, what is worth making?

I'll start with what I think we can quietly stop building: the infrastructure layer.

Payment processing. Hosting. Auth systems. Commerce platforms. These problems are solved, and solved well. Stripe exists. Vercel exists. Supabase exists. The people who built those things spent years getting security, scalability, and reliability right so you don't have to. Building another version of that isn't a product strategy. It's a distraction wearing a product strategy's clothes.

This isn't a controversial position. But it's worth saying plainly, because I still see builders pouring months into problems that the market solved a decade ago, usually because the idea feels technical, and technical feels serious.

It isn't. Serious is finding a real problem and being honest about whether your solution is wanted.

Where do good ideas come from, then?

Not from staring at your desk. Not from going harder. Not from the productivity framework you adopted last month that was supposed to finally unlock your potential.

I know this because I spent a significant stretch of my career believing the opposite. Two burnouts inside five years. The "no weekends, no Netflix, nothing wasted" version of hustle. I was extreme. I was miserable. And nothing brilliant came out of it — except, eventually, the understanding that I was never going back.

The ideas that are worth something tend to arrive when you're paying attention to your actual life. Talking to people. Walking around. Being genuinely curious about something that isn't your startup. Living like a person and not a content machine.

When you're desperate to find the next idea, you won't find it. That's not pessimism, it's just how attention works. You can't force this one.

What has changed, and changed fast, is the cost of testing an idea once you have one.

Building is cheap now. Faster than it's ever been. Which means the validation process should move just as quickly, and in the same direction. You don't need six months and a co-founder to figure out if something is real. You need a prototype, some conversations with the actual people who'd use it, and maybe a landing page with paid traffic to see if anyone clicks before you write a single line of product code.

Think of it like a buffet. You can try everything. But you still have to decide what's worth eating. The options being plentiful doesn't mean the judgment disappears. If anything, it matters more.

My process: record the idea, write out the thesis and what success would need to look like, show it to someone who matches the ICP, handle distribution before you build the product. Set kill criteria before you start. Know in advance what result would tell you to stop.

This isn't pessimism either. It's respect for your own time.

And when you're building: get out from behind your desk.

Talk to your users. Go to the community events. The feedback loop you need is not in your head — it's in the room with the people who'd pay for what you're making. Your instincts are useful. They're not sufficient.

And the cost of validation has dropped just as fast as the cost of building. You used to hire a market research firm for this. A product strategist. Run a focus group, wait three weeks, get a deck. Now you can pressure-test the same idea in an afternoon, if you're willing to be honest about what comes back.

I have what I call a product reduction room: a standing session with my AI product strategist where I bring a vague idea and a set of assumptions, and we pull them apart together. Market signals. Stress-testing the business model. Cutting the feature list down to the actual core. I tend to go big on scope from the jump, so I built the system specifically to stop myself from doing that. The system holds me accountable when enthusiasm outpaces sense.

It's not glamorous. But the products that come out of it are cleaner. The ones that don't make it past the kill criteria, I'm glad they didn't. Saved months I'd have spent building something nobody wanted, for the second time.

Which brings it back to the original question.

The tools make everything possible. That's true. It's also a little bit terrifying if you let it be.

The answer isn't to build everything you can. It's to get sharper about which problems are worth your attention, and to stay curious and alive enough that you notice them when they show up.

That part hasn't been automated yet.