{"id":1038,"date":"2026-07-10T11:14:49","date_gmt":"2026-07-10T10:14:49","guid":{"rendered":"https:\/\/metrics.blogg.gu.se\/?p=1038"},"modified":"2026-06-05T11:18:03","modified_gmt":"2026-06-05T10:18:03","slug":"levels-of-automated-code-development","status":"publish","type":"post","link":"https:\/\/metrics.blogg.gu.se\/?p=1038","title":{"rendered":"Levels of automated code development&#8230;"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-scaled.jpg\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"559\" src=\"https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-1024x559.jpg\" alt=\"\" class=\"wp-image-1039\" srcset=\"https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-1024x559.jpg 1024w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-300x164.jpg 300w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-768x419.jpg 768w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-1536x838.jpg 1536w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-2048x1117.jpg 2048w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-1200x655.jpg 1200w, https:\/\/metrics.blogg.gu.se\/files\/2026\/06\/framework-1320x720.jpg 1320w\" sizes=\"(max-width: 709px) 85vw, (max-width: 909px) 67vw, (max-width: 1362px) 62vw, 840px\" \/><\/a><\/figure>\n\n\n\n<p>Image generated by Gemini based on this blow post<\/p>\n\n\n\n<p><a href=\"https:\/\/www.mdpi.com\/2076-3417\/16\/10\/4788\">https:\/\/www.mdpi.com\/2076-3417\/16\/10\/4788<\/a><\/p>\n\n\n\n<p class=\"has-drop-cap\">The practical meaning of automated code generation is shifting rapidly. What was recently categorized as simple &#8220;autocomplete&#8221; has expanded into complex workflows involving multi-file modifications, test execution, and repository navigation. However, as Zhenhan Chen et al. argue in a recently published article in <em>Applied Sciences<\/em>, the software engineering community still lacks a shared, operational language to describe exactly how much work is being delegated to these AI systems.<\/p>\n\n\n\n<p>To fill this critical gap, the researchers proposed &#8220;Levels of Automated Code Generation&#8221; (LACG), a six-level taxonomy (L0 to L5) designed to classify the degree of automation in AI-augmented software construction.<\/p>\n\n\n\n<p> The proposed levels include:<\/p>\n\n\n\n<ul>\n<li><strong>L1 (Assisted Generation):<\/strong> Localized, token-level assistance (e.g., inline completion) with full human fallback.<\/li>\n\n\n\n<li><strong>L2 (Partial Generation):<\/strong> Generation of complete code units (e.g., functions, classes) from prompts, still requiring human integration and verification.<\/li>\n\n\n\n<li><strong>L3 (Conditional Automation):<\/strong> The system executes multi-step tasks (e.g., bug fixes, feature implementation) within a constrained OCD. The human is the final fallback.<\/li>\n\n\n\n<li><strong>L4 (High Automation):<\/strong> The system autonomously manages end-to-end development of subsystems within a broader OCD and owns the operational fallback.<\/li>\n\n\n\n<li><strong>L5 (Full Automation):<\/strong> Unrestricted software engineering across any domain, owning all fallback and recovery duties.<\/li>\n<\/ul>\n\n\n\n<h4 class=\"wp-block-heading\">Conclusion<\/h4>\n\n\n\n<p>The LACG taxonomy provides a disciplined, operational vocabulary necessary for future empirical work, benchmark design, and reasoning about responsibility allocation in AI-augmented coding. While the study demonstrates the framework&#8217;s applicability, the authors clarify that it does not serve as a prediction of tool performance, security, or productivity outcomes<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Image generated by Gemini based on this blow post https:\/\/www.mdpi.com\/2076-3417\/16\/10\/4788 The practical meaning of automated code generation is shifting rapidly. What was recently categorized as simple &#8220;autocomplete&#8221; has expanded into complex workflows involving multi-file modifications, test execution, and repository navigation. However, as Zhenhan Chen et al. argue in a recently published article in Applied Sciences, &hellip; <a href=\"https:\/\/metrics.blogg.gu.se\/?p=1038\" class=\"more-link\">Continue reading<span class=\"screen-reader-text\"> &#8220;Levels of automated code development&#8230;&#8221;<\/span><\/a><\/p>\n","protected":false},"author":68,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[6,4,5],"tags":[],"_links":{"self":[{"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/posts\/1038"}],"collection":[{"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/users\/68"}],"replies":[{"embeddable":true,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=1038"}],"version-history":[{"count":1,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/posts\/1038\/revisions"}],"predecessor-version":[{"id":1040,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=\/wp\/v2\/posts\/1038\/revisions\/1040"}],"wp:attachment":[{"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=1038"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=1038"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/metrics.blogg.gu.se\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=1038"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}