Extreme estimates of GenAI mean the destruction of the software industry.
In theory, a powerful GenAI model could run fully designed and analyzed software autonomously from natural language or other instructions, or by its own wisdom, without human review, into a fully optimized machine. Can be generated and deployed in code. This eliminates the need for all business applications (adios, SAP), all types of development platforms (peace out, Pegasystems), major software components (au revoir, Oracle database), and most of the existing tools, processes, and roles. Become. The world of software, including developers (thank you for all the fish for a long time).
We do not support this faith-based apocalyptic scenario. That's unreasonable. But the opposite view, that all individual products, practices, and roles in the software industry will continue to exist, with AI pixie dust gently sprinkled on top, is naive.
Between these two extremes, the rapid advances in TuringBot (an AI tool that assists with various tasks in the software development lifecycle) and low-code platforms point to a more realistic future for much of software. Application generation platform.
Application generation platform
Application Generation Platforms (or “AppGen” for short) are not magic and will never be. This category represents an evolution of practical platform engineering to get the most out of AI (especially generative AI) while mitigating its drawbacks. AppGen unifies the steps of software analysis, development, security, testing, and delivery by providing Turingbot for both low-code and high-code development across all steps while incorporating Agile and DevOps principles. To enable the generation of larger chunks of functionality (or entire applications), the core authoring experience becomes a cycle of natural language prompts followed by iterations with efficient, visual mediums (drawings, graphical user interfaces (GUIs), visual low-code models and domain-specific languages). Lower-level code generation for custom components, extensions, and visibility also becomes central. And, crucially, business and industry “domain knowledge” and “best practices” are built into the AI models that support this generation process, eliminating the distinction between “software development” and “off-the-shelf applications” where business excellence is required. In theory, it is predefined.
In the short term, there are some hurdles to overcome. Most natural language-based AppGen features are only suitable for “simple” generation, and the typical security and privacy concerns associated with using public LLMs apply. But AppGen is not a theory. The building blocks already exist, as do modest examples of patterns, such as the company that told us of its experience creating a logistics app to help manage maritime containers.
Low-code platform vendors got a head start with AppGen and are currently the standard-bearers in this category, but global hyperscalers, other end-to-end development platforms, and startups will also be important players. Sho. As AppGen matures, it compresses and blurs SDLC steps and roles, further democratizing development faster, enabling real-time collaboration for application design and delivery, and enabling dynamic end-user experiences. and redefine not just the software development industry, but the software development industry as well. It also includes a wide range of software in general. We estimate that the AppGen platform will mature over the next three years.
This post was written by John Bratincevic, Principal Analyst, and Diego Lo Giudice, Vice President and Principal Analyst, and originally appeared. here.
