Tag
This paper revises the estimated proportion of newly written code that is generated or reviewed by AI, analyzing its impact on software development.
This article reports on a study estimating the percentage of newly written code that is now generated or assisted by AI, highlighting the growing role of AI tools in software development.
A developer benchmarks Claude's C++ code, finding that prompting for maximum speed often introduces memory-safety violations without actual performance gains, debunking the assumed trade-off between speed and safety.
A comparison experiment shows that Kimi K2.7 Code generates landing pages at about 94% lower cost than Claude Fable 5 with similar quality, especially when given design context via an MCP server.
This article introduces Environments AI, a tool that generates and runs code for physics simulations, enabling easier creation of simulation environments.
Miguel Grinberg explains his refusal to accept unsolicited pull requests generated by LLMs, insisting on human involvement to avoid becoming a 'reverse centaur'.
The article explores the phenomenon of 'rockstar developers' who write clever but unmaintainable code, and draws parallels to the challenges introduced by AI-generated code, emphasizing the need for maintainability and team cohesion.
A discussion on Twitter about the impact of AI writing 90% of code on tech jobs, suggesting that if 10x more code is produced, everyone might be affected.
Anthropic reports that over 80% of its new production code is now authored by Claude, leading to an 8x increase in code shipped per engineer. The article outlines a roadmap for enterprises to adopt similar AI-driven development workflows.
A reflective essay on the value of not building features, arguing that unbuilt code is a hidden asset and warning that AI-driven development may lead to unnecessary code accumulation.
A developer reflects on a messy core data structure in their app, acknowledging technical debt but accepting they will never prioritize fixing it unless it hinders AI code generation.
A reflective blog post on how agentic code generation can hinder skill retention, and strategies to add friction back into development for deliberate learning.
The post argues that AI code generation for Go is more effective when the tool understands the organization's internal codebase and conventions, leading to higher acceptance rates and fewer edits needed.
A cautionary note from Addy Osmani that relying too heavily on AI to write code can hinder learning and degrade one's mental model of software development.
An analysis of Bun's controversial rewrite from Zig to Rust using AI-generated code, raising concerns about 6,755 AI-written commits merged without human review and the risks of AI-translated code in production.
Pietro Schirano announces that MagicPath can now run as a native canvas inside Codex, enabling design and building of functional apps.
A guide on setting up a new application when using AI to write the code, focusing on workflow and integration.
The author discusses the frustration of reading Markdown in the terminal and describes using Claude to quickly build a custom macOS Markdown viewer (MDV.app), illustrating how AI enables rapid creation of personal software tools.
The article draws parallels between the outsourcing era of the early 2000s and the current trend of AI-generated code, arguing that the real cost of cheap code is the loss of human comprehension and context.