Brace yourself: The era of 'citizen developers' creating apps is here, thanks to AI

1 day ago 11
BOOK THIS SPACE FOR AD
ARTICLE AD
Can citizen developers now use generative AI to build applications?
ZDNET

Generative AI (Gen AI) has eliminated much of the grunt work of building applications for professional software developers. Now, the question is: can citizen developers also benefit from this new paradigm in code creation?

Some experts certainly think so. Over the coming year, citizen developers will deliver 30% of Gen AI-infused automation apps, predicted Craig Le Clair, principal analyst with Forrester. 

Also: The best AI for coding in 2025 (and what not to use - including DeepSeek R1)

"They have the necessary domain expertise to envision and develop these solutions," he said, recommending concerted training of citizen developers to ensure the safely provisioned and controlled proliferation of AI models and copilot platforms.

One big issue is that citizen developers might not be ready to handle bare-metal Gen AI when creating applications.

Also: The five biggest mistakes people make when prompting an AI

"While Gen AI is breaking down barriers by allowing them to experiment and rapidly create no-code applications just by describing what they need in natural language, a hybrid approach remains essential," suggested Burley Kawasaki, global VP of Creatio, also co-author of the No-Code Playbook

He said one reason is that many tasks, such as designing user interfaces and workflows, are better suited to visual representation: "A good analogy is how a word processor allows users to switch between draft mode and full WYSIWYG layout depending on editing needs."

Another issue is customization. Kawasaki said citizen developers need to extend their apps easily: "While they could directly modify generated code, it's easier for humans and AI to update declarative models, which are at the heart of no-code platforms."

He also said it's important to recognize that citizen developers have little prior experience in software creation: "While the development may now be simpler, navigating the broader software development lifecycle is unchartered territory for them."

Also: I put GPT-4o through my coding tests and it aced them - except for one weird result

Within enterprise environments, citizen developers "must consider design trade-offs, best practices, and compliance with governance, security, and regulatory standards," he continued. "Lack of familiarity with structure development methodologies can slow adoption of Gen AI."  

Lessons from the professionals

Gen AI-powered coding has proven to be a preferred solution for many developers. This proliferation offers important pointers for non-professionals.

"Gen AI coding for developers has rapidly taken off because it literally speaks their language -- the language of procedural code," Kawasaki said. 

"The code may be created differently than traditional software development but, once generated, the code output fits naturally into existing development methodologies and DevOps practices. Its growth in the developer community is only accelerating."

Kawasaki said the use of Gen AI by professional developers so far has helped to highlight some important risks.

"While GenAI coding is powerful, it's the responsibility of the enterprise to ensure proper governance to mitigate risks. Without proper oversight, AI-generated code can introduce bugs, security vulnerabilities, and inconsistencies across applications."

Also: I tested DeepSeek's R1 and V3 coding skills - and we're not all doomed (yet)

He said lack of standardization also poses risks in GenAI coding environments: "If AI-generated applications are deployed without governance, organizations may face data inconsistency, variations in workflows, and uneven usability standards; or perhaps they aren't optimized to work correctly with enterprise back-end systems and data, which can impact system integrity and performance."

Kawasaki said it's also important to consider legal and ethical considerations, such as potential copyright issues or biases in AI-generated logic: "One effective strategy is to complement Gen AI coding with investments in composable architectures, instead of generating everything from scratch." 

In this development context, AI helps recommend re-using proven, secured components that are part of a curated marketplace, "whether from the platform vendor or a list of curated and validated ecosystem partners," Kawasaki said. 

Also: Can Perplexity Pro help you code? It aced my programming tests - thanks to GPT-4

However, despite the challenges, he said Gen AI is becoming a default, built-in assistant for low-code and no-code platforms

"At the design stage, AI-assisted development accelerates productivity and reduces the learning curve by generating app structures, suggesting workflows, and even creating UI elements based on natural language descriptions. It also acts as an intelligent assistant, offering recommendations and troubleshooting issues before they become problems."  

Read Entire Article