VibeCoding Vs. Flutter – App Developer Freelancer

Would we need a developer if you can develop an app with AI? In other words could LLM develop an app that is functional and scalable? With the constant evolution of technology with new hardware such as smartwatches and new frameworks such as Flutter, is VibeCoding the next big thing in app development?

What is VibeCoding? 

Vibecoding is slang for describing the process of having AI develop a program using prompts to have large language model (LLM) write the code for you. In other words prompting features for your app and have the code be written by AI based on the functionality you want without you writing a single line of code. It is often advertised as a way to have junior programmers or even people who have never written code in their lives to produce software including mobile apps.

What is Flutter Development? 

Flutter is a framework backed by Dart, which was introduced by Google in 2017. It allows for programmers to write one line of code and have your app be functional on both iOS and Android systems. As a Flutter Freelancer, more and more app development is done in Flutter. This is because the system is flexible, stable and has performance similar to native development. Recently it has evolved to be able to build game apps without the use of unity.

What is Native App Development? 

Native app development is coding in the language and framework that is specific to the operating system of the phone. For iOS you would use Swift, which replaced Objective-C apps. For Android you would use Kotlin or Java. The problem with native app development, is if you want your app to be on both Android and iOS platforms, you would need to write the same code twice. The advantage is that you would have a better performing app, however, in recent years Flutter has decreased that advantage. Of course if you have an app that was written in native code, and need to have maintenance on it or new features added, as a native app developer, I would be more than happy to help.

What are the risk of AI generated apps? 

The biggest risks of using AI to generate your app, is the security, debugging and scalability. Here are the main risks that currently exists with AI generated mobile apps:

  • Security: AI apps may not be designed in a way that keeps your data secure. It is not unheard of of having vulnerabilities in your code. So if you are developing an app to handle any personal data or any data or process in your app that needs security for financial apps or military apps, vibecoding is not for you. For example, as a financial app freelancer, most of quality control is spent on testing the vulnerability of the app, and even known packages in Flutter, may not be viable for banks to use due to its potential risk.
  • Lack of Explanation: Having your app as a black box, can have issues with heavily regulated spaces such as legal apps, financial apps, healthcare apps and even simple consumer apps. Not having control over the code or ability to explain why a function is performing the way it is performing could lead to legal challenges.
  • Inaccurate outputs: Due to its training functionality, some outputs can be wrong. For example, you may have an app that counts calories for a specific sport. If you were to use traditional programming you would have a formula that would calculate calories based on factors inputted by the user. This formula with AI may change all the time and provide different answers based on the same inputs. Therefore, creating an app that’s not stable.
  • Debugging: If there is a bug, it makes it hard to find the issue, especially for inexperienced app developers. As a mobile app developer of over 12 years, I have tried to use VibeCoding to create a simple MVP, what I found is without the knowledge of where the bug was, it would be impossible to launch an AI generated app. What’s more, I found lots of security breaches and errors that would arise when you try to scale as it was build without the proper infrastructure that would not scale.
  • App is more than just a code: App typically starts with an idea, AI cannot come up with a unique idea by its nature of learning from past data. It also needs design elements, and while it can generate AI images and mimic voices of famous people, it is often easy to spot and may look like other similar apps, which may infringe on copywrites. Finally, the app has to be launched in the app store and more importantly found. Working with an ASO marketing agency, we find that even suggestions given by Apple search ads or Google AdWords, decreases the results of doing your own analysis and app store optimization. These are just some of the reasons where AI just simply would not work.

When would you use AI in Flutter App Development? 

There are instances where AI can help speed up app development. When providing flutter app development services, your job is to build apps, debug and optimize performance. Sometimes it is faster to ask AI on some very specific code that is needed, which then can be tweaked to your specific instance. Typically, this can save you time, but should only be done in very specific use cases and checked prior to submitting the app for app review by AppStore or Google Play.

Another use case is to develop proof of concept app, if you need to present a prototype in order to raise funds, it maybe faster or cheaper to build your app with AI. However, for MVP (minimal viable product) that you would launch on the app stores it is advisable to build it the right way, where all code can be explained and easily de-bugged, with proper design and marketing this will ensure its highest chance of success.

About Brian Kayfitz

Brian Kayfitz, started his career 20 years ago developing games for Xbox, Play Station as well as Unity. Once apps came to your mobile phone he coded in Java and Objective-C. Transitioning into Kotlin and Swift, before finally developing in Flutter. After working for financial institution BMO developing their investment app, he worked for a quick serve restaurant developing apps for both eCommerce ordering as well as internal productivity. Finally, he has decided to mentor other app developers, build his own apps including popular travel app Visited, which has over 2.5 million users, and helped many businesses build their apps. As a flutter freelancer, he has worked on productivity apps, game apps, fitness apps, healthcare apps and more! If you are looking to develop your app in Flutter, Brian Kayfitz would be happy to help.

Leave a Reply

Your email address will not be published. Required fields are marked *