Lukas

🍎 Indie developer building for iOS/MacOS (AI powered)
🌐 dpli → Unwind →

[Lukas @1uk6o]

It still feels a little surreal that we’ve reached a point where one person can take an idea and actually build it with the help of AI.

For me, that has been a game changer.

dpli started with a very simple problem: I wanted to help my son focus while studying at home.

I began exploring whether sound could help create a better environment for concentration. AI helped me go much deeper than I could have on my own — navigating research across auditory neuroscience, acoustics, physics and mathematics, questioning assumptions, comparing approaches, and gradually shaping the concept behind the engine.

The idea became:

Generate the sound in real time instead of playing recordings.

No loops.

Everything runs locally on the device.

And make the sound spatial, so it doesn’t feel like noise trapped inside your headphones, but more like a soft acoustic cocoon around you.

From there I started experimenting with different kinds of noise and how they could be shaped to create an environment that supports sustained focus without becoming another distraction.

A few years ago, I probably wouldn’t have attempted something like this.

Or I would have gone down the rabbit hole, hit the complexity wall and eventually given up.

AI changed that.

It didn’t magically build the idea for me — but it allowed me to research, question, prototype, validate, iterate and supervise work across areas far outside my original expertise.

And eventually turn an idea into something real.

Something that first helped my son.

Then helped me.

And now helps other people too.

That became dpli.

[Lukas @1uk6o]

@skpnky Turns out you weren’t wrong — you just underestimated how many geniuses are out there 😂

[Lukas @1uk6o]

@Kishan_Devani_ The data suggests AI amplifies creativity more than it replaces it — improving individual output while often reducing the diversity of ideas.

[Lukas @1uk6o]

@Paulina_laba solid 👋

[Lukas @1uk6o]

I keep wondering where the real competitive advantage is today.

Product? Features? UI?
Most things can be copied by competitors over a weekend.
So where is the moat?

Distribution? Brand? Data? Community? Speed? Customer relationships? Or simply the ability to learn faster than everyone else?

What are you building your advantage on that can’t be copied in 48 hours?

[Lukas @1uk6o]

@cinamarina Huge congrats! 🚀 Having a neurologist validate Lumah — and potentially join forces with you — sounds like a major step forward. Excited to see where this goes 👏

[Lukas @1uk6o]

@Raigeki_Dev @X 👋

[Lukas @1uk6o]

@aesatiroglu 👋

[Lukas @1uk6o]

@thejustinwelsh Your environment quietly becomes your definition of normal. Choose it carefully.

[Lukas @1uk6o]

@thejustinwelsh How many people are stuck because they’re searching for a better idea instead of distributing the one they already have?

[Lukas @1uk6o]

@sciboticsX 👋

[Lukas @1uk6o]

@yongfook @levelsio when building becomes commoditized and AI can increasingly commoditize distribution too, what remains a durable moat for a small software company?

[Lukas @1uk6o]

I think this is the interesting shift: building the product is becoming commoditized, so the moat moves somewhere else — distribution, trust, audience, proprietary data, deep workflow integration.

Personality can be part of that, especially for indie hackers.

Do you think we’re heading toward a world where the best small SaaS companies are essentially creator-led software businesses?

[Lukas @1uk6o]

@mattisigur @AdityaJhajhar12 same here

[Lukas @1uk6o]

@mattisigur That’s probably the more interesting part — if engineering stops being the bottleneck, where does it move?

[Lukas @1uk6o]

One thing AI coding agents seem to be quietly removing: natural stopping points.

As an indie developer, this is especially noticeable.

There’s always one more prompt to run, one more bug to fix, one more feature to ship, one more “small” improvement before going to sleep.

And because the next iteration can take 30 seconds instead of 30 minutes, stopping almost feels irrational.

This article puts some interesting numbers behind that feeling — developers reporting “just one more” loops, working longer than intended, and delaying sleep because of AI coding:

https://t.co/mfMlbKiPU4

For indie hackers building entire products with agents, I wonder if this becomes one of the hidden costs of the new leverage.

AI lets one person do the work of a much bigger team — but does it also make it much harder to ever feel “done” for the day?

Have you noticed this when building with agents?

[Lukas @1uk6o]

@eliana_jordan “Features don’t bring users. Distribution does.” Probably the most expensive lesson for first-time founders to learn.

[Lukas @1uk6o]

@DanielSmidstrup or everyone is prototyping now 😅

[Lukas @1uk6o]

Nothing makes me happier than seeing my app genuinely help someone and work exactly the way I hoped it would.

[Lukas @1uk6o]

That’s exactly why I wouldn’t call iOS-first a universal rule. Android still has ~68% (https://t.co/Slbbk0XFdv) of the global mobile market, while iOS monetizes much better per install (https://t.co/OCjGbF7IHq) .
So iOS-first can be a great business choice — but it still depends on who you’re building for.