Don’t bite the hand that feeds you, said the Greek poet Sappho around 600 B.C.
But the age-old adage, originally about horses, might need a 21st century update as generative artificial intelligence (AI) capabilities continue advancing.
Something along the lines of, “Feel free to augment the hand that created your code.”
Infusing generative AI solutions into enterprise developers’ toolkits is increasingly helping to streamline software development workflows while making it easier for junior developers and those of all skill levels to write their own code and develop innovative products using text-based prompts.
And as payment acceptance and embedded payments more broadly become a key experiential differentiator, along with personalization and purpose-designed offers both at and after checkout, streamlining processes and creating new experiences with the help of AI can help firms augment, enhance and revolutionize their coding workflows.
Importantly for incumbent enterprise players, generative AI can help speed up the process of modernizing legacy code, as well as enhance the translation of code from one programming language to another — helping accelerate digital transformation initiatives and smooth out endemic software onboarding frictions.
With many companies undertaking system enhancements across areas like accounts payable (AP) and accounts receivable (AR), the benefits of having AI-generated code on hand can’t be overstated.
But for businesses to truly get the most out of leveraging AI, they need to understand how it works and be clear about what their desired outcome is — and this holds true across all areas where AI is applied, not just for coding assistance.
That’s because understanding how AI works in one domain can inform how organizations apply the innovative technology in others.
“Generative AI can essentially retrain the old AI and machine learning models, [giving it a] really interesting trajectory when it comes to payments,” Tom Randklev, global head of product at Cellpoint Digital, told PYMNTS in May.
Echoing that sentiment, Jason Verlen, the former SVP, product management, at CCC Intelligent Solutions, told PYMNTS in a past interview that, “AI will greatly increase the speed and convenience of payment processes — and it’s not just the payment workflow [where AI will have an impact], but across everything.”
But while integrating AI into payment operations presents benefits, and leveraging the technology to develop code can no doubt open new doors and accelerate internal productivity, the question remains: How does it work?
Typically, when generating code using AI, programmers and internal software developers tell an AI system using plain text prompts what they want to do, and then the generative AI solution spits out code bits, or even full functions. This helps reduce repetitive manual tasks and vastly accelerates the coding process.
He explains that “all the state of the art” generative AI models trained to produce software code typically auto-generate test cases based on the prompt provided by the programmer, because deep learning neural networks are built on the concept that answers are what most people agree tend to be most correct.
“Basically, the model generates thousands of versions of the code and then looks for the solutions that get all of the test cases correct,” Lintilhac said.
He added that even the most advanced models can’t predict re-output of their own code, and that AI-generated code is “nowhere near industry grade at this point, and no one should use it for safety critical or financially critical or legally critical applications [without reviewing, editing, and refining it first],” explaining that, “in order to achieve [wholly] trustworthy program synthesis, we need to develop neuro-symbolic AI systems.”
At the end of the day, AI can help firms generate code quickly and at scale while removing many repetitive and time-consuming steps along the way — much as the technology can across a variety of enterprise workflows and processes.
And as in all of them, the technology still needs to have a human in the loop to validate its outputs and course-correct as needed.
“No matter the ways and means in which AI is being harnessed, it’s incumbent on firms to mull how they can enhance value rather than just chase a trend,” Shaunt Sarkissian, founder and CEO of AI-ID, told PYMNTS in May.
Within the payments space, the opportunities to enhance value while streamlining legacy cost centers is immense. Areas like transaction routing optimization, checkout personalization, fraud protection and more can all benefit from right-now applications of AI.