Are you tired of AI yet? It seems like there’s no escape from AI – it couldn’t be any buzzier than it already is. From fashion to politics, healthcare to entertainment, big tech, and more, people love talking about AI. And for good reason. While forms of AI have been around for decades (yes, we’re all still trying to get over 2001: A Space Odyssey), this new wave feels different. Some even compare it to a new industrial or (more frighteningly) nuclear age.
But is all this hype real? At Flowfactory, our mission is to unleash unique value by helping customers create ultra-specialized business applications, so naturally, we’ve been following all the latest developments in AI closely. Particularly from where I sit as Chief Product Officer, I’ve had many great conversations internally and with customers about AI, low-code, and what’s next for development. The good news? While AI is powerful and seems poised to disrupt many industries, the need for humans to collaborate on and lead development projects isn’t going away any time soon.
Let’s take a closer look at how AI is used in software development today, and then explore how AI and low-code can work together to create more powerful applications more quickly. Finally, we’ll look at what Flowfactory in particular is doing to integrate AI into its products and customer experiences. Let’s dive in!
The reality of AI in development today
If you’ve been following the news, the abundance of AI-focused headlines can feel overwhelming. When it comes to practical use, it can be difficult to decipher what’s real and what’s perhaps a bit sensational. In software development, for example, technologies like ChatGPT have the potential to develop code quicker than ever, bringing a new level of efficiency and speed previously unseen. But how reliable can this be, and how will it impact the workforce?
At Flowfactory, we believe development teams should approach AI with eyes-wide-open optimism. It’s a capability that isn’t going anywhere (and will only continue to grow in sophistication) and should be used as a way to spark conversations and collaboration. Some headlines have noted that technologies such as ChatGPT can produce code faster than humans, even if that code isn’t entirely accurate, as one analysis showed that 52% of ChatGPT’s answers to software programming questions were wrong.
It’s for these reasons that we believe we still need developers – individuals who oversee the process and can identify potential biases, issues, or gaps. Developers as a result should view AI-enabled tools as a partner and an extension of their own human capabilities and a way to make them more efficient and productive.
In addition to writing code, software testing and debugging are other popular use cases for AI in development. Because AI is so good at rapidly analyzing large data sets, it is especially good at pointing out errors in code and automating many common QA tasks. But there are limitations here too – especially if there are issues in the dataset provided.
One developer tested ChatGPT’s debugging capabilities around reformatting an array using real-world code, and in their words, “the result was a total failure.” While excelling at many development tasks, bias is a common concern with AI. You still need excellent software and QA engineers – people – to control the development process. The mind has the ability to sense if something isn’t right or doesn’t make sense – there is always something to be said for gut instinct and experience!
Because AI aces some development tasks but generates errors in others, I believe that neither extreme – fully outsourcing all development to AI or not using it at all – is the right move. The answer lies somewhere in between.
AI is just one more productivity tool in a long line of technological innovations. Ultimately, AI makes development more accessible to more people (something we feel strongly about here at Flowfactory). For example, Business Insider reports that Oded Netzer, a Columbia Business School professor, thinks that AI will help coders rather than replace them.
I agree with Netzer. If you look at the history of coding, there’s been a steady evolution, of which AI represents the next logical step. We started with inserting punch cards into computers that were the size of whole rooms, evolved to C and C++ and beyond, and now we’re using low-code to build many applications. AI is not at war with low-code; AI can make building applications using low-code faster and more efficient.
How to get started with AI in development, aided by low-code
My recommendation for organizations looking to harness the power of AI in their development is to identify a specific use case – a bottleneck, a recurring pain point – and pair it with low-code to realize value quickly. It’s tempting to want to go to extremes like optimizing all the code in your organization with AI, but at this stage in its evolution, it’s much wiser to start small and grow from there. The best question to ask right now is: how can I use AI and low-code to make incremental (but valuable) improvements that solve pressing business challenges?
To illustrate this more clearly, let’s look at one real-world example from an industry-leading company: Siemens. As you can imagine, their sales processes, especially quotation management, are highly sophisticated and must be responsive to market dynamics and regional pricing differences worldwide. AI is ideally situated to ingest a massive quantity of both internal and external data, analyze it, and then propose the optimal pricing for a given quotation. And this is all accomplished from within a low-code application built on Flowfactory’s platform.
Siemens started with a specific but significant question that represented both a process bottleneck and an opportunity to increase margins while improving customer satisfaction: what is the right discount, if any, we should offer on any given sales quote? And they didn’t just tell their sales team to fire up a new browser tab pointed at ChatGPT to do this. They used API integrations to connect AI capabilities into an existing low-code application, ensuring a seamless and secure experience with results they could trust to be accurate.
Flowfactory and the future of AI and low-code
This is just one of an infinite number of use cases for combining the power of AI with enterprise-grade low-code applications. In the near-term, Flowfactory’s low-code development platform has two primary ways of using AI to make development faster, easier and more powerful.
First, as mentioned in the Siemens example, Flowfactory’s platform offers pre-built connectors with existing leaders in the AI space. Using modern APIs, it’s easy to supercharge any application with AI capabilities without compromising on either security or the user experience. In practice, the magic of AI should be invisible, adding value without distracting the user or calling too much attention to itself. Low-code integrations make that possible without rewriting entire applications from scratch – a benefit that Flowfactory’s platform has always enjoyed.
Next, what if you could always have a senior developer sitting next to you, guiding you through the application development process? Flowfactory’s platform will soon offer an AI-powered helper to make low-code development even faster and easier. Not sure of the next steps to take? Just ask Flowfactory’s AI helper, and you’ll get customized advice. We’ll have more details on this in the months to come.
Will AI ultimately replace low-code?
It’s too early to say, but if it does, that’s fine with us. When it comes to ultra-specialized business applications, the tools used to build them will continue to evolve, with Flowfactory at the forefront. What won’t change is the need for people to come together to build applications that solve very “human” challenges. That’s what Flowfactory really excels at, and I’m excited to watch how the space evolves in the coming months and years.
