Full-text search across 61889 articles. Combine with topic and date filters; results sorted by relevance.
Researchers introduced the first application of program embeddings from LLMCompiler, an LLM pretrained on intermediate representation (IR) code, to program analysis and optimization tasks. Combining source and IR code em…
AI software engineering is shifting from manual coding to system architecture and verification, according to analysis of AI-assisted development workflows. Developers must transition from writing code to defining specifi…
A Hacker News user asks the community what they are building with AI, prompting discussion on LLMs, coding agents, and development tools used for personal, employer, customer, or startup projects. The post inquires about…
A software developer who rejoined a former team found that members now rely heavily on AI for coding, code review, and PR descriptions, diminishing craftsmanship and engagement. Another developer described a team where A…
A new study of over 100,000 GitHub developers finds that AI coding agents boost coding activity by up to 180%, but actual software releases increase by only 30%, suggesting AI helps with coding but not shipping. The pape…
A veteran software engineer with over 20 years of experience questioned why the Hacker News community is overwhelmingly critical of AI-generated code, arguing that users only care about working products, not how code is …
Claude Opus 4.5 can write fully playable game mod code — including new weather systems, physics interactions, custom UI elements and in-game menus — when paired with a coding harness such as Claude Code running on a Wind…
Anthropic removed over 80% of Claude Code's system prompt for models like Claude Opus 5 and Claude Fable 5 with no measurable loss on its coding evaluations, according to a July 24, 2026 post by Thariq Shihipar titled "T…
A software engineer argues that coding agents should be used to rapidly prototype features first, then the resulting code must be made maintainable through ruthless human review before any optimization or polish. The eng…
A manual, non-automated test of four AI coding-agent workflow tools found that OpenSpec required four actions to reach the first prompt, after a 6-second install of 80 packages and an init step that presented 40 tool opt…
Codacy has completed more than 300 customer migrations from SonarQube, with the share of new Codacy customers coming from SonarQube rising from roughly half historically to about two-thirds in 2026, according to the comp…
Growomat, a product built with AI coding tools Claude and Codex, outlines a development process that prioritizes a 'harness' of principles, user stories, and testing gates over the models themselves. Citing Stack Overflo…
Tricentis' 2026 Quality Transformation Report finds that 60% of organizations regularly ship untested code into production, driven by leadership pressure to move faster and too much code to test. The report underscores t…
Infrastructure developer Earendil Works gave its Pi coding agent a single tool called fabric_exec, which accepts TypeScript programs that call Pi's capabilities as an API, and one such program ran continuously for eight …
A new workflow using Moonshot AI's Kimi K3 coding agent and AI video tools lets a single person build cinematic scroll-driven websites for roughly $1, replacing what previously required a designer, video editor, and deve…
Brandon Gubitosa of CodeRabbit reports that 2026 is becoming the year of AI quality, as engineering teams shift from generating code to understanding and trusting AI output. Stack Overflow's 2025 Developer Survey found 8…
Developer and writer James Friend asked Claude Code to recreate the 1986 game Thrust in the browser, but the AI produced unplayable slop that failed to capture the original's physics and feel. Friend then used Claude to …
An experiment comparing two AI agents—Postman AI Engineer and Claude Code with Postman plugin—on the same API spec drift problem found they produced opposite fixes: one updated the spec to match the code, the other updat…
Supacode, a native macOS app built on The Composable Architecture and GhosttyKit, has evolved from a harness for running CLI coding agents into a full IDE with flexible panes for agents, editor, file management, and Git,…
A developer is building kUML, a new UML modeling tool that treats diagrams as code in a typed host language (Kotlin), addressing the structural flaw of model-code separation. The tool aims to enable first-class IDE refac…