It hasn’t always been easy to get employees excited about how artificial intelligence will one day transform their day-to-day work tasks. Despite all the well-deserved excitement amongst technology evangelists like me, the technology is still in the early phases of the diffusion curve as many are still wrapping their heads around its potential.

For more people, that may be finally changing – and we have low-code application development tools to thank. As low-code makes it easier for “citizen developers,” or nontechnical workers, to spin up custom applications, they’re also empowering employees to leverage artificial intelligence to make those applications even smarter and more effective.

Indeed, IDC expects the number of industry-specific digital apps and services to explode to 500 million by 2023. And by 2025, roughly 10 percent of these apps will be led by artificial intelligence.

Low-code will be at the core of both the explosion in the number of apps and number of citizen developers who are creating them – allowing organizations to develop more apps while enabling further sophistication of those creations through artificial intelligence. This will lead to both greater business efficiencies and better outcomes. Here are three of the most exciting developments we can expect:

Improving the end-user experience

One of the key benefits of low-code development is that it puts the tools of creation in the hands of people closest to the problem, such as sales professionals who interface with customers or marketing leads who are driving internal and external branding. In fact, according to 451 Research, nearly 60% of all custom apps are now built outside the information technology department. Of those, 30% are built by employees with either limited or no technical development skills.

This is an exciting change for organizations within an enterprise. When citizen developers have an intimate understanding of the end user’s pain points, they’re able to create apps that address those pain points more effectively as the subject matter experts.

For example, one area where low-code has seen a lot of momentum is in the development of virtual agent tools for customer service. Typically, when developers design customer-facing applications, they include a series of dropdown menus and text fields so that customers can use them to describe any problem they’re experiencing. Those responses are then used to direct the customer to the right resource or human agent.

Artificial intelligence is transforming this entire process. Rather than asking users to select those categories manually, developers can simply allow users to describe in plain English the problem they’re having. Conversational AI works in the background, analyzing the description of the problem to automatically identify and categorize the issues.

This approach results in two big improvements to the issue resolution process. For one, this improves the diagnosis of issues and consequently increases productivity and resolution time dramatically. It also allows citizen developers to create more streamlined user experiences via low-code, since a human no longer needs to design and develop interfaces to accommodate a series of dropdown menus.

Enabling hyperautomation

There is an ever-increasing number of terms used to describe technology that allows enterprises to automate their manual processes, from robotic process automation and digital process automation to AI and machine learning.

Although every one of those technologies alone is exciting, what I’m particularly thrilled about is the result when all these components work together. This is termed hyperautomation, or the idea that anything that can be automated in an organization should be automated.

A key to hyperautomation is when one automated process flows directly into the next. Let’s consider the AI-powered virtual agent low-code example I cited above. In this context, paired with low-code, a “hyperautomated” virtual agent system would automate the process of identifying customer issues while also seeking out ways to remove bottlenecks from the entire end-to-end issue resolution process. The result is a more productive, streamlined process that will become more efficient over time.

Of course, considering that many organizations are still in the early days of automating even some of their processes, we’ve still got a way to go before we can see a fully hyperautomated enterprise in action. Still, I remain excited about how hyperautomation can improve operations for all kinds of firms, even those that have already automated their most repetitive tasks.

The next frontier for low-code and AI

So far, I’ve focused only on how AI is elevating the end-user experience and ensuring that low-code apps deliver smarter outcomes. The future, however, is even more exciting.

For example, we’re already seeing some exciting glimpses of the future of conversational low-code app development. New language learning models such as GPT-3 from OpenAI can take simple commands and use them to generate code as we see with applications such as CoPilot on GitHub.

It can even be used to help people develop websites as we see with Neuroflash focusing on marketing websites, or Enzyme being used to create landing pages. This allows people simply to use descriptions and content to create what they need.

It’s an exciting possibility that will no doubt cause some developers to wonder how automation will continue to affect their jobs. I remain optimistic. Just as automation in manufacturing allowed workers to focus on more complex problems, greater abstraction for coding platforms and artificial intelligence will allow developers to spend their time working on harder, more interesting application challenges. Artificial intelligence and low-code technology will support developers, not replace them.

Dave Wright is ServiceNow’s chief innovation officer and acts as an evangelist for how to improve workplace productivity. He has worked with thousands of organizations to implement technologies that create efficiencies, streamline business processes and reduce costs. He wrote this article for SiliconANGLE.

Image: mohammed_hassan/Pixabay