Since the recent rise of generative AI tools, such as ChatGPT which can generate text and code, there have been countless discussions exploring the future of the software development industry and the potential displacement of developers in years to come.
Kinetic IT’s ServiceNow Developer MVP and Certified Technical Architect, Paul Morris, was one of many in his field who believed ServiceNow developers would not be displaced any time soon as generative AI couldn’t be utilised to create low-code configurations. That is, until he attended the ServiceNow Knowledge 2023 conference in Las Vegas, where his perspective shifted. He shares his thoughts and key insights from the conference in this insight.
Revolutionary technology at Knowledge 2023
I personally believed that ServiceNow developer jobs would not be replaced by robots any time soon. Or at least, not as long as generative AI cannot directly configure existing platforms like ServiceNow, which are predominantly configuration-based using low-code tools. However, my recent attendance at ServiceNow’s Knowledge23 conference in Las Vegas shifted my perspective. I now realise I was incorrect in both aspects of my thinking. Let me explain why I was mistaken…
RELATED CONTENT: 3 ServiceNow AI tools revolutionising IT service delivery
The power of AI and coding
Generative AI, particularly in the realm of coding, garnered significant attention at ServiceNow Knowledge 2023. During a conference keynote session, I watched a demonstration of “text-to-code” where AI was used to write working code to retrieve records from the system using ServiceNow-specific code, based on a simple text prompt. While this one example worked well, a key flaw with current generative AI models is their reliance on generic JavaScript, rather than ServiceNow-specific code.
It’s evident that further training on high-quality domain-specific code is necessary for this functionality to become adopted by ServiceNow developers. This may take some time before it becomes truly useful, and even then, there is more to becoming a ServiceNow developer than just coding.
RELATED CONTENT: 6 powerful ways AI is transforming business
Developers low-code too
ServiceNow developers don’t solely rely on traditional coding; they leverage a range of low-code tools, which expedite development. The platform’s standout tool, Flow Designer, enables users to create business workflows without writing a single line of code. Through its user-friendly interface, sequential actions can be specified without code, incorporating coding concepts such as Boolean logic and control loops, achievable through drag-and-drop functionality. This is the key tool used by ServiceNow developers to build automated workflows, which can’t be created from a chatbot in a browser.
RELATED CONTENT: Kinetic IT identified as a ServiceNow ‘Leader’ in 2023 ISG Provider Lens™
Introducing Now Assist: The dev’s co-pilot
A ground-breaking unveiling at the conference was Now Assist, a virtual agent chatbot designed to assist ServiceNow developers. Reminiscent of the old Microsoft Clippy (but without the personality), Now Assist enhances developer productivity by providing summaries and guidance while modifying configuration. The assistant was shown to aid developers in many areas of the tool, from pro-code to low-code.
RELATED CONTENT: Case study: ServiceNow Discovery integration uplift
Text to flow: AI-driven workflow generation
Most impressively of all, Now Assist demonstrated the capability to generate flows line by line based on prompts. That’s right – it could create an entire workflow end to end in a low-code tool such as Flow Designer, directly into the tool. While it is early days, it’s my understanding this proof-of-concept uses OpenAI to analyse the vendor’s own flow designer documentation to summarise how to create a flow based on specific inputs.
Behind the scenes, this information must have been transformed into flow actions and inputs using natural language understanding (NLU) techniques to then create the sequential flow, one action at a time. Impressively, Now Assist could also summarise flows in plain English, facilitating better communication and collaboration.
RELATED CONTENT: 5 tips to unlock value in your ServiceNow Configuration Management Database (CMDB)
Why I was wrong and embracing the future
My previous belief that generative AI couldn’t be utilised to create low-code configurations was proven incorrect by ServiceNow’s proof-of-concept demonstration. However, this does not spell complete doom for ServiceNow developers. While generative AI can indeed be used to directly configure the platform, it won’t be an overnight shift. This is because Now Assist won’t be fully rolled out in its infancy until later this year and there’s a large lead time where developers will need to learn and embrace this technology. It will take some time for productivity gains to materialise and ultimately, Now Assist’s primary function is as a productivity-enhancing co-pilot, not as a replacement for developers.
ServiceNow’s Knowledge 2023 conference was very insightful, revealing further use cases of generative AI in ServiceNow development I did not previously consider possible. Generative AI, paired with low-code tools like Flow Designer, is set to revolutionise how developers work. Rather than replacing developers, AI will serve as a powerful tool to enhance productivity and streamline workflows. Whether you are a low-code or pro-code ServiceNow developer, it’s crucial to adapt, embrace new technologies, and harness their capabilities to create a brighter future for ServiceNow development.
RELATED CONTENT: 11 things to consider before implementing AI in your business
Get in touch with our ServiceNow experts and find out how we partner with you to deliver seamless ServiceNow solutions that connect your business.