
Over the past few decades, various movements, paradigms, or technological explosions (whatever you want to call them) have rocked the software world to either offload much of the programming tedium to the end user or automate more processes. I promised. CASE Tools, 4GL, Object-Oriented Programming, Service-Oriented Architecture, Microservices, Cloud Services, Platform as a Service, Serverless Computing, Low-Code, and No-Code all theoretically diverge from software development to cumbersome Removed the burden. And it can threaten developer employment.
Still, here I am. Software developers are busier than ever, and the demand for their skills is only increasing.
Also: 74% of developers are willing to leave for a new job.And it’s not just about money
Pluralsight author Vlad Catrinescu says: “Guess what – we’re still here and busier than ever.”
The question is how the developer’s job will eventually evolve. The application of artificial intelligence to application development and maintenance may eventually make low-level coding a thing of the past.
For example, Fixie.ai CEO and co-founder Matt Welsh predicts that “programming will be obsolete in the next decade or so.” “I believe the conventional notion of ‘writing a program’ is dying out,” he predicted in a recent article published by the Association for Computing Machinery. “In fact, except for very specific applications, most software, as we know it, will be replaced by AI systems that are trained rather than programmed.”
“If you want simple programs – after all, you don’t want models of hundreds of billions of parameters that all run on clusters of GPUs – those programs can be run by AI rather than hand-coded. generated,” adds Welsh.
So what exactly will the roles of IT professionals and developers look like? Catrinescu believes that the new generation of automated or low-code development solutions is actually “empowering IT professionals and developers to tackle more challenging applications.” enterprise. “
Also: The future of the web will require a different kind of software developer than ever before
Until very recently, “the focus of development was on getting better use of engineering or getting more reuse from a broader pool of code writers,” said chief creative technologist and co-founder at argodesign. author Jared Ficklin said. “This has created a tool that facilitates orchestration, which allows a typical application developer to use a graphical interface and code modules, called skills, created by machine learning experts. AI solutions can be orchestrated, as well as this allows subject matter experts in the business to orchestrate entire campaigns with an interface.”
Machine-learning-enabled tools like this “help us gather requirements and leverage engineering,” continues Ficklin. “If there are gaps, coders need to jump in and fill them. In all these cases, there are many points to maintain interoperability and security, so the architecture is still handled by IT. increase.”
With the advent and rapid progress of AI and machine learning, Welsh predicts that trained models could replace coding at a very basic level.
AI coding assistants like CoPilot are just the beginning of what I’m talking about. Of course, it’s clear that all future programs will eventually be created by AI, and humans will be relegated to a supervisory role at best. If there’s one thing I’ve learned in my work in AI over the last few years, it’s all too easy to underestimate the power of increasingly large AI models. I’m not just talking about things like his CoPilot on Github replacing programmers. What I’m talking about is replacing the whole concept of writing a program with a training model.
Moving away from coding entirely opens up a new way of looking at application development, allowing for a more conceptual, higher-level business role. “Exciting changes are coming from surprising directions,” he says Ficklin. “The wider world imagines low-code/no-code as a visual interface that moves nodes to piece together code. Is required.”
Also: Low-code development: how it’s being used
Fricklin demonstrates this new way of developing and updating applications. “One of our current clients, Builder AI, takes a unique approach of using his AI analysis of voice conversations to gather requirements and then build and deliver those experiences further. I do,” he explains. “They also have a voice assistant that you can add to your Zoom call. This assistant listens to someone explaining the mobile application to the project manager and automatically captures and lists the features. , and the AI pairs them up to be a pattern for the architecture of the app.If the code module exists, it’s patched, if it doesn’t, the code writer comes along and adds the module.Over time With , this process is becoming increasingly automated.”
This means more real-time computing, Ficklin continues. “This is where software latency, rendering, and assembly come into play in real time. You can imagine asking Alexa to create an app that helps you organize your kitchen. , wirelessly in real time. Deliver applications to mobile phones or wearable mobile computers.”