Artificial intelligence has so dominated headlines and conversations that it seems like every company is announcing their own AI-related feature, solution, or initiative for their business. And you wouldn’t be wrong: the latest McKinsey Global survey shows that organizations are most commonly using generative AI (gen AI). In fact, 40% of those reporting AI adoption at their organizations say their companies expect to invest more in AI, and 28% say gen AI use is already on their board’s agenda.
And that’s because generative AI’s expected disruption of – and benefit to – businesses will be significant. At VMware, for example, we’re conducting internal AI experiments with various departments such as engineering, marketing, and customer success to assess the impact of generative AI on our operations. Specifically, we’re looking at whether AI will make employees more productive or satisfied in their jobs.
But with gen AI and technologies like ChatGPT tangibly transforming how developers and people work, it has understandably prompted age-old anxieties around its potential to replace human tasks and jobs. However, most technologists believe that generative AI is designed to be an assistant, not a replacement.
It’s already optimizing employee experiences and business workflows, including software development. And when software agility increasingly correlates with business success, adopting AI to speed up app development and enable developer velocity will be table stakes for any modern business.
AI’s impact on software development
Software development takes a lot of time. Developers are writing code for apps that require different user provisions and resource configurations, pulling from the latest API specifications, and so forth. The application of AI in software development can accelerate these areas – for example, by quickly pulling info from many different sources and synthesizing documentation sources.
Similarly, developers often experience writer’s block when creating code. AI-driven tools can assist in overcoming these creative roadblocks by suggesting code patterns, offering auto-completion suggestions and even generating sections of code. AI can unlock developer productivity by allowing teams to bridge the knowledge and change context to gain new information and solve issues faster.
From a DevOps perspective, operators face some of the same issues as developers when it comes to accessing the right information. AI can not only summarize information based on existing data into an answer but can take a multi-phased approach to generate queries into a data store that contains operational information on environments and then synthesize that information. With tools like VMware Intelligent Assist, teams can ask questions about the environment they manage and translate issues by generating queries about what should be prioritized first.
AI app accelerators in practice
App accelerators primarily help developers build intelligent assistance – whether for customer support or a specific product. They serve as a comprehensive guide to help developers understand the fundamental building blocks required for these intelligent systems. App accelerators can help enable summarization services via a chatbot to help teams understand what they need to build these solutions.
Accelerators help teams understand patterns and help put pieces of the puzzle together faster.
For example, engineers can use embedding models or similarity analysis to radically simplify data models and get important details more quickly. To get started, developers need to identify the relevant documentation, often in the form of PDFs and app catalogs. Furthermore, implementing external model tracking and embedded model processing are key as they ensure that AI systems can efficiently process large volumes of documentation using LLMs (Large Language Models).
Today, DevOps teams are inundated with the complexities of their tech stacks and ultimately need ways to interact with systems in more natural terms. The true power of AI accelerators is their ability to help teams talk more naturally. For example, rather than developers knowing the internals of their Kubernetes environments and operator service endpoints, with AI accelerators, they can ask which of their applications is having problems.
AI app accelerators are ushering in a new era of AI-driven intelligent assistance. They serve as a key roadmap for developers and businesses to navigate the intricacies of building AI solutions. Although AI can simplify processes and accelerate software development cycles, anything produced by AI must continue to have human oversight to ensure it is accurate and applicable.
To learn more, visit us here.