Leave your email address and receive the latest developments in software, AI and Mendix.

There is a good chance that low-code will become the standard way of developing apps before long. Meanwhile, AI technology is more powerful and more accessible than ever. ChatGPT grew to 100 million users in record time. What do these developments mean for the way we will develop apps in the future? Will all traditional developers be out of work? We do not think so. In fact, we think their work will become much more enjoyable. This article explains why.
AI technology has of course been used inside apps for a long time. It is why Siri gives you a fitting answer, why Netflix recommends series you then genuinely want to binge, and why your phone can recognise your face.
But these days AI plays an ever larger role in the process of developing apps too. AI is then no longer the functionality: AI builds the functionality. (Will AI soon be able to create new AI as well? Inception?) Tools such as GitHub Copilot and ChatGPT are already being used to write logic or to refactor existing code. And the results are astonishingly good. Sometimes the code you get back is usable straight away, and sometimes the AI at least points you in the right direction. That saves an enormous amount of time, especially for developers with a bit less experience.
Low-code helps to simplify, speed up and improve the development process. AI contributes to those same goals. No wonder, then, that low-code platforms are already building AI into their solutions to make things even easier for developers. A powerful combination.
Mendix introduced Mendix Assist a while back, for example. Mendix Assist currently consists of two virtual co-developer bots: MxAssist Logic Bot and MxAssist Performance Bot. Both are already widely used by Mendix users.
In Mendix you build logic in the form of microflows. MxAssist Logic Bot was developed by analysing more than twelve million of these microflows. They were built by Mendix users over the past 10 years. That allowed Mendix to work out what the best practices around microflows were. When a Mendix developer builds a microflow now, this bot makes suggestions based on those best practices. Around 40% of all microflow activities are now created with the help of the MxAssist Logic Bot.
MxAssist Performance Bot inspects your project in real time and checks it against performance best practices. The bot then makes recommendations or fixes the issue for you straight away. MxAssist Performance Bot was built using statistical analysis of thousands of anonymised Mendix projects. That is how it learned common anti-patterns and best practices for developing microflows, domain models, pages, security and more.
We expect Mendix to add more smart, AI-supported features in the time ahead. Perhaps we will see more of that in a next version of Mendix?
Some people get a little nervous about the current developments. We are mostly just very enthusiastic about the future. AI already makes our work as developers easier, and that is likely to take off considerably in the near future. This is what we predict:
Eventually we will probably be able to tell AI what we want in natural language. The AI then gives us the logic back in record time, with fewer and fewer errors. That means we will barely have to write code ourselves any more. Does that mean developers who can write code will no longer be needed? We do not think it will go that far. Ultimately, the code AI generates still has to be checked and adjusted. That requires developers. So AI will not take over developers' work. But do you cling to the traditional way of developing? Then you will probably be overtaken quickly by the developers who do embrace AI. So at some point you will have to jump on this train.
At the moment, developers' work involves a great deal of repetitive work. That work is being taken over by AI. Which means developers can spend their time on far more enjoyable and challenging work. A large part of a developer's work will consist of giving instructions to AI. A good developer will soon be a developer who knows how to get the most out of AI.
AI makes app development accessible to a far larger group of people. Low-code has already brought about a rise in 'citizen developers', and that will only grow further.
You can already see it in the MxAssist Performance Bot from Mendix we mentioned earlier. AI makes it much easier to enforce best practices, which prevents errors. On top of that, AI can be put to very good use in testing and debugging apps. With the help of AI, the apps of the future will be more secure and of higher quality.
Some people are afraid AI will make apps very monotonous. We think the opposite will happen. Because the basics are in place much faster, you can give far more attention to the extras. And AI can produce a great many variants in a short space of time. AI-generated images and sound also open up possibilities. Adding animation to your app, for example, becomes much simpler and cheaper as a result.
We get very enthusiastic about the developments in low-code and AI. We think they will bring mostly very positive things, and that before long we will be able to build even cooler things in less time.
Want to know more?Ask Armando
Privacy, GDPR, contracts, leave, time tracking, invoicing: as an organisation grows, so does everything you need to arra...

Anyone who knows JAM-IT knows that we believe strongly in structure, quality and ownership. Not as buzzwords, but as a f...

In a professional application, good observability is not optional: it is your lifeline when something unexpected happens...