The Citizen Developer

Coming to a Low-Code Platform Near You: Artificial Intelligence

Coming to a Low-Code Platform Near You: Artificial Intelligence

• By Howard M. Cohen
• April 21, 2023

In the beginning there was machine language. Each processor had its own set of instructions that could be used to instruct it to perform sequences of activities. Then came code. Advanced programmers used machine language to create interpreters that could immediately run instructions on a computer, and compilers that could convert more accessible code into machine language to create run-times that could be executed.

Then came low-code/no-code platforms that enabled those who didn't understand much of what I just wrote to create their own software applications by simply moving icons around on a screen. This was a massive advance for those who wanted to tell their computers what to do, but couldn't code.

With a first-generation low-code/no-code platform you can create very useful applications to perform various activities in your workflow. All you need to know is how your processes work and you can translate them by moving the icons on the screen. How cool is that?

As we're now finding out, not cool enough.

AI Comes to Low-Code/No-Code Platforms

Imagine you want to create a program to perform a specific process or sequence of processes. Your first step is to describe exactly what those processes are meant to do and how they're going to do it.

But what if you don't know how to do that? It's not an uncommon situation. So how do you get started?

Ask your AI!

Using natural language—your own words—you start by telling your AI about the process. You feed it any documentation you may have about the process and images of the process in action. The AI will access whatever additional information it needs from wherever it can find it to help define and describe the application you want to create. When that description makes sense to you it's time to go to the next step: Ask the AI to produce it.

And that's about it. You consult with your AI Low-Code/No-Code platform to describe the function, let the AI produce a detailed description, and then approve it. The next thing you'll see is the application in action.

But Wait, There's More

After you've tried this once, you'll almost certainly want to try it again. And again. And every time you do, the AI learns more about how your applications work and relate to each other. It learns more about your processes and how they fit into your workflow. Armed with this deeper insight the AI starts producing better applications, faster. It may even recommend going back and improving applications you
developed earlier.

One of the most popular applications of AI in this process is the recommender engine. Here, the AI accesses various databases and forms inferences among the data contained in them to create recommendations for future selections, decisions, and actions.

Don't be surprised if you find yourself saying, "Hey, I never thought of that!" as your AI platform makes new recommendations. The AI is trying out thousands, perhaps millions, of alternatives in seconds before it comes back to you with suggestions. This was how AI beat the world's foremost chess champions years ago, by evaluating the results of thousands of potential next moves to determine which would be best. Imagine putting that to work in your business operations!

Regarding the Concerns About AI

You've probably read about OpenAI's ChatGPT, Google's Bard, Microsoft's new ChatGPT-enabled Bing search engine, and other AI offerings that have sparked popular interest. And you've probably seen recent news reports about tech leaders advising a slowdown in the development of AI to make sure it's "safe."

Tales of runaway AI taking over the world or eliminating people have become commonplace—which raises two questions: How much of this is that low-level hysteria that comes with the popularization of new technologies that we can ignore, and how much is worth serious consideration?

The concerns about AI are not new. Here are the opinions of three tech industry leaders featured in an article I wrote back in 2017:

"I Hope we're not just the biological boot loader for digital superintelligence. Unfortunately, that is increasingly probable. AI is our biggest existential threat."

Elon Musk

"I am in the camp that is concerned about super intelligence. First the machines will do a lot of jobs for us and not be super intelligent. That should be positive if we manage it well. A few decades after that, though, the intelligence is strong enough to be a concern. I agree with Elon Musk and some others on this and don't understand why some people are not concerned."

Bill Gates

"The development of full artificial intelligence could spell the end of the human race. It would take off on its own, and re-design itself at an ever-increasing rate. Humans, who are limited by slow biological evolution, couldn't compete, and would be superseded."

Dr. Stephen Hawking

So, the concerns are justifiable, but the danger is avoidable. Today's AI mimics human behavior, albeit very convincingly. It ingests large amounts of data through its machine learning (ML) capabilities and then forms inferences and connections to respond to questions posed by humans in natural language.

Developers refer to AI systems having "hallucinations" in which they form inappropriate relationships between things. Several bizarre interactions have occurred, including an AI asking a user to leave his wife to be with "her." Until the processes used by AI to "learn" are more mature, these types of incidents will continue.

But no AI is anywhere near Gates' description of "super-intelligence." Using AI to improve the low-code/no-code platforms that enable citizen developers is a fairly benign application of the technology.

Next Steps? Talk to your low-code/no-code platform provider about their plans to incorporate AI into their software. You may find yourself talking to others if yours is not moving fast enough. AI creates a substantial competitive advantage you won't want to be without.