Software:Chat-code
Chat-code, also called chat-code development, is the general term for software development using a conversational user interface that automatically translates natural language requirements into software applications. The term was coined by Rierino in 2023 and suggested as the next step in rapid application development.[1]
Origin
Chat-code concept emerged as a result of two main trends in recent years:
- Popularization of low-code development platforms which is forecasted to grow to $32 billion by 2024 by Gartner.[2]
- Advancements in conversational AI technologies, such as ChatGPT, improving accuracy of natural language processing.
Autocompletion, conversion of comments into code and translation of natural language into programming languages has been made possible with solutions such as GitHub Copilot. Despite their substantial public interest, such solutions are received with a level of skepticism by some, especially in terms of usability and security of generated codes in real-life use cases.[3][4][5]
The promise of chat-code development is to address these challenges, by leveraging the structural aspect of low-code development platforms as a governance layer to voice or text based conversational programming interfaces.
Approach
Chat-code development architecture consists of three main layers:
- Conversational AI layer that translates human language into machine readable format (such as JSON, XML, YAML) that is specific to the lower layer's configuration requirements
- The low-code layer that translates this configuration into a specific programming language using code generation, generic programming or similar techniques
- The execution layer, which is typically the programming language compiler or interpreter converting generated code into the machine code
The main difference between chat-code and NLP based code generation solutions is the introduction of the low-code layer, which promises the required safe-guards and standardization to increase usability and security of generated applications.
Reception
As a relatively new concept, chat-code development is still in its early days. However, with the increasing popularity of ChatGPT among startups[6] it is anticipated to become a trending topic in the rapid application development market.
References
- ↑ Ozmen, Berkin (2023-01-20). "From Low-code/No-code to “Chat-code”" (in en). https://medium.com/rierino-tech/from-low-code-no-code-to-chat-code-4c637423b8c3.
- ↑ "Gartner Forecasts Worldwide Low-Code Development Technologies Market to Grow 20% in 2023" (in en). https://www.gartner.com/en/newsroom/press-releases/2022-12-13-gartner-forecasts-worldwide-low-code-development-technologies-market-to-grow-20-percent-in-2023.
- ↑ Perry, Neil; Srivastava, Megha; Kumar, Deepak; Boneh, Dan (2022-12-16). "Do Users Write More Insecure Code with AI Assistants?". arXiv:2211.03622 [cs]. http://arxiv.org/abs/2211.03622.
- ↑ Pearce, Hammond; Ahmad, Baleegh; Tan, Benjamin; Dolan-Gavitt, Brendan; Karri, Ramesh (2021-12-16). "Asleep at the Keyboard? Assessing the Security of GitHub Copilot's Code Contributions". arXiv:2108.09293 [cs]. http://arxiv.org/abs/2108.09293.
- ↑ Morrison, Ryan (2022-12-21). "AI coding assistants leave developers "deluded" about software quality – study" (in en-US). https://techmonitor.ai/technology/ai-and-automation/ai-coding-assistants-github-copilot.
- ↑ David, Emilia. "ChatGPT is coming for classrooms, hospitals, marketing departments, and everything else as the next great startup boom emerges" (in en-US). https://www.businessinsider.com/chatgpt-startup-boom-funding-generative-ai-trend-jasper-runway-2023-1.
 |  |
