The End of Programming as We Know It

It just hit me and I couldn’t stop thinking about it.

Remember those threads about how we’ve forgotten how to make good software? The ones where people marvel at programs written 20-30 years ago and complain that development isn’t what it used to be, engineers are getting weaker, everything’s being simplified, and the levels of abstraction just keep growing.

I think this trend will only intensify. In 50 years, people will stop programming altogether. Well, not in the way we’re used to. It’ll be more of a toy or hobby for fans of retro aesthetics. It’ll be some mix of visual editors with flowcharts and a dash of text prompts to describe the overall task and possible nuances. Writing code by hand? What nonsense.

Humans will just control the result, tweak the automation, make sure everything connects properly.

And they’ll completely stop thinking about where bytes go, which framework is being used, which language, syntax, and so on. That was always just extra information needed only to be able to write what you envisioned. Objectively just applied knowledge with no other value.

And knowledge of algorithms will be devalued too. There just won’t be any use for it. The robots do the heavy lifting.

We’ve already invented so much stuff that even a dumb machine can create apps at a junior-to-mid level. What can we expect from a smart one?

All the necessary knowledge is already there. The technology to use any knowledge will come too. As well as the ability to create new knowledge.

And that’s a good thing 💪

Lopaka Gallery Launched and a UX Surprise

We recently added a gallery to Lopaka where all public projects are displayed.

There are a ton of awesome designs there, check it out: https://lopaka.app/gallery

It’s super useful to observe exactly how users are using the editor. Really nice to see that people are actually using our product to the fullest.

One interesting observation: all user graphics are stored in projects, and each project consists of screens. A project can have a name (shown on the main page) and each screen can also have a name (shown in the screen list).

And here’s the thing: nobody gives a damn about screen names! It’s just there, and nobody (95%) ever edits it.

First Customer, Giant PR, and a Beautiful Day

This morning I opened GitHub and merged a giant pull request.

I’ve been working on it for the last two months. Had a great time, got a ton of enjoyment from engineering all the cloud features for Lopaka.

Today, for the first time in my life, someone bought a subscription to my software! Unreal.

Huge thanks to everyone who supported and gave feedback.

Special thanks to Denis aka deadlink for his huge contribution to the project.

Thanks to my beloved wife, who helped research the market.

Thanks to Pasha Zhovner for the reposts and support.

What a beautiful day!

I Cloned My App in 30 Minutes Using AI

Recently stumbled upon an internet simulator: you type in any (even non-existent) URL or search query and get something generated by a neural net: https://websim.ai/

You can check out a ton of stuff on the homepage. Works really well.

The coolest part is that you can generate any app there (with enough patience). For example, in half an hour and about ten follow-up prompts, I made a copy of Lopaka: https://websim.ai/c/URr4umb4PKmu44Jr2 It’s not perfect, but that’s just THIRTY MINUTES OF WORK 🤯

Video review: https://www.youtube.com/watch?v=a4nXGnumD1U

By the way, I’m not stressing about AI replacing us. It’s fine, we’ll find something more interesting to do. Online schools, though — they’re gonna have a rough time.

Or that everything around us will be flooded with generated content. Also don’t care. My mom (65) happily watches generated YouTube videos and listens to AI retro covers of Yesenin, not even suspecting how it’s all made. She gets a lot of joy out of it, forwards them to her friends. And that’s how we’ll all be.

The Best Startup Advice I Got: Monetize First

The most useful advice I got:

1. Explore monetization opportunities If you quit your job to work on your own project, as tempting as it is to start cranking out a bunch of cool features, pay yourself first (as Kiyosaki would say).

My hypothesis at the time was that solo indie students and hobbyists wouldn’t pay. So I needed to find small engineering teams and study their processes.

I was advised to explore both directions.

What to do For B2C — based on past interviews, pick the most needed features, create teasers and pricing descriptions. Instead of payment, ask for an email for a waitlist. This way we can tell if there’s real demand, experiment with pricing, and calculate conversion rates.

For B2B, they suggested outreach via LinkedIn and cold emails, finding more people for interviews, sending them questionnaires.

What happened Two weeks went into updating the site. We switched to Tailwind+DaisyUI — modern layout libraries. I was skeptical, but turns out it’s super convenient and saves a ton of time. I named the paid subscription Lopaka PRO, set the price at $10, and waited.

Then my wife joined me and started actively pushing LinkedIn. We picked about 30 companies that make electronic devices with screens. Got about 80 contacts. It’s a very slow process, people are reluctant to accept connection requests, rarely reply, but we got maybe five constructive conversations going. We even got a response from the head of platforms at Teenage Engineering 🤩! (Though he thought it was another pitch looking for investments, so it didn’t really go anywhere.)

My assumptions were wrong…

B2C: In a month of experiments, 50 people wanted to buy and filled out the form on the site! Some even filled out an additional survey. That’s absolutely insane. I could already be making $500 a month (hypothetically 😁)

B2B: It’s complicated. Only managed to collect a few responses, never got on a call with anyone. There’s a set of features that could be interesting, but it’ll be super hard to replace a whole bunch of established, if suboptimal, processes all at once. It’s a lot of work, we’ll get there gradually.

Next piece of advice:

2. Sell the most needed feature you can build fast I was researching demand by offering a huge set of different things for graphics design: team collaboration, asset library, prototypes, and diagrams. But I’d be building all this forever. It’s very hard to do well and quickly as a one-person team. And I ideally need to roll out some features and start making a living in less than a year.

They suggested I focus on the bare minimum, with a small price tag. A basic feature set—Lopaka Plus!

And so for the past two weeks I’ve been rewriting the beta and integrating cloud tech into the current project. There will be a database, a project list, and auth via GitHub and Google. Just need to hook up payments. So far I’ve only registered with LemonSqueezy, haven’t figured out how everything works yet.

One more small insight:

3. Talk to people who are ready to pay All this time I’ve been conducting interviews without paying much attention to money. I was trying to understand how people’s processes work and what’s missing in their workflow.

When project survival is on the line, you need to focus only on those whose pain is so strong they’re willing to pay just to stop suffering. They’ll be much more eager to share their experience and demand solutions to their problems. Exactly what you need!

The advisors also confirmed that an editor is not really a venture story at all. Pure bootstrap, you can launch it on your own.

You’d think I’ve been reading about startups, product launches, and all that for years. And the advice is super simple. Like, stupidly basic. And yet I struggled to get my thoughts together and pick the right direction. Over the next couple months I’ll be testing new hypotheses.

Maybe you’ve got some advice too? I’d love to hear it, maybe something will work.