NYTimes: 'Vibecoded' apps flooding the app store

Software
1

Interesting report on the current vibecoding thing.

Katie

1 Like
2

Whatā€™s the TL;DR for people that donā€™t subscribe to the NY Times?

PS: You might want to correct the typo in the title? ā€œVidecodedā€ ā†’ ā€œVibecodedā€

3

Here is a concise summary of the article:

The ā€œVibecodingā€ Boom

  • Definition: ā€œVibecodingā€ is a trend where individuals build custom websites or apps by describing their ideas to AI tools, completely bypassing the need for traditional coding skills.
  • Surge in App Creation: After years of stagnation, new App Store releases jumped 30% last year to approximately 600,000. In the first half of this year alone, new apps doubled to around 560,000.

Developer Experiences

  • Profit-Driven Success: Marco PĆ©rez, a Chicago software engineer, used AI to build and release profitable apps like InfoDrizzle and Stampa in just weeks. The success is prompting him to quit his job to develop apps full-time.
  • Passion Projects: David Svensson utilized AI to release 11 highly specific niche apps (such as tracking fire risks in Sweden) over five months, driven by personal interest rather than financial gain.

Challenges for Appleā€™s App Store

  • Stagnant User Demand: Although app creation is booming, consumer interest is not matching the pace. App downloads only grew by 3% last year and 2% in the first half of this year.
  • Review Bottlenecks: The massive influx of new apps requires heavy manual review by Apple. This has led to some developer complaints regarding delayed approval times, though Apple maintains it reviews 90% of apps within 48 hours.
  • Quality and Revenue Risks: Experts warn the trend could clutter the App Store with ā€œjunkā€ or inappropriate content. Additionally, because vibecoded apps often rely on ad revenue rather than complex in-app purchases, Apple may miss out on its standard 15% to 30% commission.
1 Like
4

Thank you.

Which AI made this summary?

5

This should be a functioning gift link to the NYT article.

1 Like
6

Good question. Itā€™s a pretty good summary. Useful.

7

Is the link from Lisa indeed the full article?

(It only mentions the two apps from PĆ©rez)

If so, the AI that Kirk used hallucinated a lotā€¦

8

It is the full article as it exists 7/20/26 7:50 PM EDT, but is not the full article that I read this morning when I first posted. It appears about half the article is now gone. Interesting. Maybe some editor wanted it short, or maybe someone else objected to the content.

Kaite

9

In the last couple of months, I have vibecoded several apps. Ranging from something that could be generally interesting (a reminders+calendar combo), niche (a feed reader for MiniFlux backends), to esoteric (a screen remote viewer for a Korg Kronos synthesizer).

The commond denominator, though, is that these are not finished apps. Thereā€™s a lot of polish involved so that they do not feel generic, making a native app does not automatically mean it is a good mac-assed app, by default they are a little bit like Visual Basic apps of yesteyear: yes, they work. But are not nice.

I suspect we are going to see a lot of these types of apps published. Which will mean the good ones will become buried (unless they pay) as the App Store is not good for discoverability.

Another effect: if you build an app that happens to be successful, the differentiating value will quickly drop to zero as anyone will be able to basically clone it for free. As an example, if you need something that Fantastical puts in their paid tiers (for example, coalescing the same event happening in different calendars, or hiding an event without deleting it) itā€™s basically trivial to implement in your own app now. And once you have it for you, why not sharing with the world?