The article "Automatic Programming: Large Language Models and Beyond" by Michael R. Lyu, Baishakhi Ray, Abhik Roychoudhury, Shin Hwei Tan, and Patanamon Thongtanunam explores the growing popularity of automatic programming, particularly driven by Large Language Models (LLMs) like GitHub Copilot. The authors discuss the challenges and concerns surrounding the quality, security, and trustworthiness of automatically generated code. They highlight the need for advances in software engineering, such as program repair and analysis, to enhance the reliability of automatically generated code. The article also examines the broader implications of automatic programming, including the integration of low-code and no-code application development. It reviews historical milestones and recent literature in code generation, program repair, and LLM-based intelligent programming. The authors emphasize the importance of trust boundaries and the role of human-LLM collaboration in future programming environments. They conclude by discussing the potential of LLMs in automating non-code artifacts and processes, such as test generation, code review, and code summarization, and the need for systematic approaches to ensure the quality and security of automatically generated code.The article "Automatic Programming: Large Language Models and Beyond" by Michael R. Lyu, Baishakhi Ray, Abhik Roychoudhury, Shin Hwei Tan, and Patanamon Thongtanunam explores the growing popularity of automatic programming, particularly driven by Large Language Models (LLMs) like GitHub Copilot. The authors discuss the challenges and concerns surrounding the quality, security, and trustworthiness of automatically generated code. They highlight the need for advances in software engineering, such as program repair and analysis, to enhance the reliability of automatically generated code. The article also examines the broader implications of automatic programming, including the integration of low-code and no-code application development. It reviews historical milestones and recent literature in code generation, program repair, and LLM-based intelligent programming. The authors emphasize the importance of trust boundaries and the role of human-LLM collaboration in future programming environments. They conclude by discussing the potential of LLMs in automating non-code artifacts and processes, such as test generation, code review, and code summarization, and the need for systematic approaches to ensure the quality and security of automatically generated code.