Each week we find a new topic for our readers to learn about in our AI Education column.
What if creating computer programs was so easy, anyone could do it?
Welcome to another week of AI Education, where, this week, we’re going to talk about low-code and no-code software development and its collision with artificial intelligence—a collision that has software developers pondering the future of their industry.
To be clear, we are not a computer programmer, though we have them in our family. We were introduced to coding in elementary school, programming lines of Basic on Apple II computers, but it’s not a skill that we honed or retained beyond school. But even our rudimentary education in coding taught us that more often than not, everything done by a computer is accomplished by writing line after line of command that tell the machine, in a language that it can understand, what to do. A five-second interaction between man and machine might be supported by one line or thousands upon thousands of lines of code.
While we wrote most of our simple code in a computer’s notepad application, different languages, interfaces or environments are used for different tasks and different machines. While computer code mostly has a consistent syntax and logic, as in spoken languages, proficiency in one type of coding does not necessarily carry over to other languages or coding environments. So being a competent coder with broad knowledge may or may not be as valuable as being a specialist with in-depth knowledge of one particular type of coding, depending on the need.
But what if none of that actually mattered?
What Do Low Code and No Code Mean?
We’re going to discuss low code and no code together, but since low-code development is a precursor to no-code development, we will define it first. Low-code is software development with minimal use of actual coding, allowing people with lower levels of computer literacy to create their own programs and applications. No code development, on the other hand, is application and software development without coding at all, permitting users to create their own apps with no programming knowledge whatsoever.
Low-code and no-code are kind of a nice marketing gimmick, as there’s still plenty of code involved under the hood. What the terms actually define is a software development environment and interface that is entirely graphical, where users can use a mouse and pointer and literally drag-and-drop to create new programs. To use a low-code or no-code interface, we wouldn’t even need to know that there’s code under the hood supporting our instructions, and that the system automatically fills in the boilerplate code that makes our programs work.
The “how we got here” this week is pretty simple—these have become ubiquitous technology terms that are often used without being explained.
Why Low-Code and No-Code Caught On
Low-code and no-code empower individuals to create the software solutions they need when and where they need it. So rather than impose technological tools from above, enterprises can adopt a no-code system and allow their employees to custom-build their own tools specific to their own individual needs and preferences.
Low- and no-code systems can also be deployed relatively quickly, since the idea is not to create a finished product out of the box but to create an environment where workers make their own refinements. No-code systems not only democratize and accelerate development, they also accelerate automation when deployed across an enterprise.
There are some drawbacks to no-code systems as we’ve known them so far. The platforms can only be customized to a certain extent, so they may in some ways create obstacles to innovation—and they come with their own sets of integration issues, potentially adding another dimension of difficulty to building a technology stack. These systems rely on APIs to interact with other third-party software and may or may not support the other tools businesses and institutions are using.
Where Does AI Come In?
In some ways, artificial intelligence challenges low- and no-code development, as some AI models are pretty competent at coding themselves. If we have AI to write our code, the value of a no-code environment would seem to be greatly diminished. AI can offer all of the efficiency and democratization benefits of low- and no-code environments without any of the inflexibility issues (though integration will remain an issue).
AI also offers us a new way to interact with development environments. Instead of dragging and dropping in a no-code system, we can just tell an AI chatbot with text or voice commands exactly what we want, in natural language, and it will deliver what we’re asking for—no fancy mouse-work necessary. Companies are already using no-code AI to create analytical models on the fly, automate workflows and develop customized tools.
What Is the Future of AI and Low/No Code?
AI and no-code platforms shall converge. Whereas developing AI and machine learning models still usually takes a huge collective wealth of knowledge from IT professionals and data scientists, the low/no-code concept will allow individuals to do this work without the professional help—AI development will be cheaper, faster and easier, from a human resources perspective.
As we’ve noted, the no-code concept is already being applied to AI to offer users environments where they can create their own AI agents, chatbots—even train their own customized AI models—with little-to-no technical knowledge of AI. And applying AI within no-code environments may alleviate some of the technology’s customization issues and enable low-knowledge users to build sophisticated applications.
We’ve already discussed AI-driven development and some of the concerns for what it might do to the software industry. We’re approaching a moment where anybody can be a proficient software developer and solve their own problems with technology—and, thus, any business or institution will be able to do the same. The combination of AI with no-code environments may make human computer programmers as relevant as fax machines and pay phones… not to mention all those technology consultants out there. We imagine that IT departments will also start to look a lot leaner as no-code AI becomes the norm.
