The Citizen Developer

Preparing to be the Citizen Developer of Tomorrow

• By Howard M. Cohen
• March 28, 2024

The world of the citizen development continues to evolve at an ever increasing pace. Yesterday we were talking about the possibility of creating useful applications with low- or no-code (LCNC) tools, and about how this approach would put app development into the hands of those who knew best the processes that needed automating: the users. Drag-and-drop, point-and-click, moving icons instead of cutting code; it all sounded great, and indeed it was and is great.

According to a recent survey, 82% of organizations are saying they "can’t attract and retain the quality and quantity of software engineers [they need] and are replacing them with citizen developers." The industry watchers at Gartner are saying that "by 2026, developers outside formal IT departments will account for at least 80% of the user base for low-code development tools." With so many turning to citizen development, it has become absolutely vital to begin looking ahead and envisioning what our citizen developers will need to know in the future.
Put into terms of productivity:

• Microsoft has said it expects 500 million apps will be created over the next five years, and 450 million of them will be designed on LCNC platforms.  
• Gartner estimates that 70% of new applications developed by enterprises will use LCNC technologies by 2025 (up from less than 25% in 2020).  
• LCNC will make it to the mainstream, with 75% of development shops adopting this platform.  
• 41% of developers want more than half of their organizations’ app development to be LCNC.  

Will Citizen Developers Go the Way of Classic Developers?

In the beginning, there was machine language, and only an elite few could create applications with it. Then came the interpreters and compilers. If you believe the name of the earliest interpreter, application development in those long-ago days was very BASIC (Beginners All-purpose Symbolic Instructional Code). The introduction of compilers like Fortran, COBOL, PL/1, Pascal, and others began the steep growth curve of programming languages that led to the current age of Java, Python, C, Go, Ruby on Rails, and others. Each of these occasioned the need for more and more sophistication on the part of code-cutters.

Given that the LCNC platforms will inevitably travel a similar curve, will the skills of citizen developers need to similarly increase and diversify? There are several components to the answer to this key question:

• The consultants at net solutions estimate that more than 442 LCNC platforms are currently available, of which 227 are low-code, and 176 are no-code platforms. Can any or all of these accommodate the increasing demand for LCNC apps? Can they improve fast enough as the demand for more sophisticated applications increases?
• How will the skills requirements for citizen developers grow and change as these platforms become more and more sophisticated?
• As artificial intelligence is inevitably integrated more and more deeply into LCNC platforms, how will the role, the skills, the tasks, and the knowledge of citizen developers need to evolve?

The Growing Role of AI

Many have asked if LCNC will be replaced by AI. Anything is possible, but consider this: a product manager for an LCNC platform developer recently told Forbes, "Although there will likely be classes of application that can entirely be created by AI at some point without any human intervention, or low-code platforms for that matter, this does not mean that all future software development (and developers) will lose the need of low-code development platforms. In fact, there may be a new class of low-code platforms infused with AI that will become the dominant method to design, build, debug, and maintain sophisticated, custom, and high-stakes applications."

We have already witnessed examples of that “new class” of LCNC platform emerge in the marketplace, most notably, and expectedly, Microsoft PowerApps. AI Magazine listed the "Top 10 no-code tools revolutionizing the world of AI" in May 2023, which included PyCaret, DataRobot, Obviously AI, RunwayML, Google Teachable Machine, Microsoft Lobe, Apple CreateML, Akkio, Google AutoML, and Amazon Sagemaker. Two things are notable about this list. The first is that it suggests that while AI is improving LCNC, LCNC products are also emerging to improve AI.

The other is that the names of the producers of these top 10 are, for the main part, well-known names, as opposed to obscure new entrants. This indicates that another era of LCNC is already upon us.

The Citizen Developer of the Future: Code is Dead; Long Live Code

Clearly, developers are seizing upon the widespread distaste for software coding. Like the intrepid pioneers who wrote the first compilers and interpreters to eliminate the need-to-know machine languages, today’s pioneers are building AI-based platforms to eliminate the need ever to code software again.

The operation of AI comprises two major components. One is the engine, the software that actually does the work. The other is the large language model (LLM), the enormous bases of information made available to the engine so it can work its magic. Already, it has become clear that selecting the right LLM will be key to accomplishing a given task with AI. Extending this to AI-Assisted LCNC, our citizen developers will need to become adept at LLM selection. Hopefully, the industry will reduce this to a catalog search, as they have so much computer hardware and application software.

The Citizen Developer of the future will also need to become conversant with conversant computers. Interaction with popular consumer AI devices such as Alexa, Google Home Assistant, and others has shown us that the AI trains us to speak in specific ways just as we try to train it to do what we want when we want it. No doubt more sophisticated AI engines will require users to adopt a more structured way of speaking, which often brings with it a more structured way of thinking. That really can’t be a bad thing, can it?