One thing I find tricky when designing AI products is deciding when the product should just act and when it should ask the user first.

Asking for confirmation feels safer, especially when AI is making decisions or changing something for the user.

But if every action needs confirmation, the experience quickly becomes annoying. At some point, the user is basically supervising the AI instead of benefiting from it.

I don t think the answer is simply confirm important actions, because what counts as important can depend on the user, the context, and whether the action can be reversed.

I've always worked remotely, always in small teams. For years I struggled with the same things:

• staying productive

• staying consistent

• doing the right tasks instead of the ones I enjoy

• getting my habits back after travel or a messy month

I thought I was alone in this, so I kept hacking my way to solutions. Here's what has actually worked so far.

If you've tried registering a domain anytime recently, you know how rare this is. Usually every decent idea is either parked for thousands or forces you into compromises. Finding a clean, 6-letter .com sitting right there felt like winning a tiny lottery.
It got me thinking about how much energy goes into naming in the early stages.
Question for the founders and creators here: How do you usually go about finding domains or experimenting with names?
Do you use AI tools, churn through registry searches, look for expired domains, or just stick to standard TLDs like .com or .ai no matter what?

When I launched I just picked a price below a couple of competitors without really looking at long term math. Six months in between hosting costs and hours spent answering tickets I'm essentially working for free.

Grandfathering old users or adding higher tiers makes sense on paper but actually communicating the change to people who signed up early feels extremely awkward.

If you had to fix underpriced software after launch what made writing that announcement email easier?

This partner model was actually tried by a few earlier startups but it ran into massive operational friction.

Beyond inventory mismatches speed is the biggest bottleneck a kirana owner has to manage walk in customers cash payments and chit-chat while dark store staff exist solely to pack orders in under 90 seconds. Plus, quick-commerce platforms make better margins when they buy bulk inventory directly from brands instead of splitting profits with a middleman shop.

Have you seen any local apps successfully blend both models in Delhi?

Working towards launching my app.
It's too early for meaningful data, growth trends, or any real signal on what's working, and I'm okay with that.

What I've noticed though is that the internet is full of milestone posts. First 100 users, $10k MRR, viral launches. And when you're pre-data, it's easy to accidentally use someone else's month 18 as your week 1 benchmark.

I'm not losing sleep over it, but it did get me thinking about how founders define meaningful progress before the numbers are there to tell the story.

My current approach is staying focused on qualitative signals are the right people finding it, are early users actually engaging, are conversations happening. But I'm curious what others have done:

So many indie makers launch like this:

• alone.

• no hype.

• no crew.

• no plan.

• just vibes.

They think it s noble. if it s good, people will notice.

But here s the problem:

I've been thinking about AI product economics, and I think "cost per API call" is only part of the problem.

An AI feature can look cheap from the infrastructure side and still become expensive to operate.