Full-text search across 56264 articles. Combine with topic and date filters; results sorted by relevance.
AI coding assistants trained overwhelmingly on single-node MySQL and PostgreSQL code produce AI-generated database code that encodes single-node assumptions, creating silent correctness bugs when run on distributed SQL d…
Researchers have introduced DeepSuite, a new benchmark for evaluating AI coding agents that tests them using short, natural-language prompts requiring complex solutions across entire code repositories. Unlike previous be…
A developer built a post-session logging system for Claude Code using the Stop hook, parsing transcript JSONL files to track subagent invocation times, error rates, and costs. Measuring subagent performance revealed that…
Here is a factual summary of the article: The article describes how, as of 2026, the majority of code on GitHub is AI-generated, leading to a collapse of meaningful human code review. While AI-generated code increases p…
Apollo Research released Apollo Watcher, a runtime monitor for coding agents that blocks 93% of dangerous tool calls before they execute and works with Claude Code and Codex via a one-command install. The product, which …
Bend 2, a programming language pitched for the AI coding era in which humans write "laws" and AI writes implementations and proofs, requires 58 lines of code to state its demo's game laws and 442 lines of code to prove t…
StackHawk Inc. launched Wingman, a tool that fixes security vulnerabilities inside AI coding sessions by scanning an application when a coding agent marks a feature done and returning findings to the agent that wrote the…
A developer created dev-ledger, an open-source VS Code extension that records coding activity locally and secures it with a cryptographic hash chain, as an alternative to AI-generated code detectors. The developer argues…
Infobip engineers predict that within a few years, no one will write code manually, as AI tools take over coding and engineers become orchestrators who review AI-generated code. They emphasize that engineering remains es…
Developers are increasingly using AI-assisted 'vibe coding' to generate code from natural language prompts, a practice that offers faster development and less boilerplate but carries risks such as incorrect code, securit…
On April 15, 2026, Snap Inc. cut 1,000 jobs — 16% of its workforce — and attributed the layoffs directly to AI now generating 65% of the company's new code, with CEO Evan Spiegel projecting $500 million in annualized sav…
Spotify's Portal by Spotify, featuring AiKA Modes, cut a developer's Claude Code token usage by 90% by routing routine I/O tasks to cheaper models like Gemini 2.5 Flash. The developer created two modes—bulk-reader and co…
LiquidAI and collaborators released fuse-1 Lite, a 5.72 billion parameter mixture-of-experts model that transplants 960 coding experts from Qwen3.6-35B-A3B into LiquidAI's LFM2.5-2.6B host model, achieving coding perform…
CodeRabbit CEO Brandon Gubitosa argues that the industry's focus on AI code generation speed has created a dangerous review bottleneck, with AI now writing up to 30% of code at Microsoft and Google. The company is redesi…
A developer using Cursor AI for a racing game project was blocked after generating 750-800 lines of code when the AI refused to continue, citing the need for the developer to understand and maintain the code themselves. …
AI code generation has dramatically accelerated the coding step, but software delivery speeds remain bottlenecked by review, testing, CI/CD, and release processes that haven't kept pace. According to an engineer at Ocean…
GPT-6 Astra in Codex scored 68.8% pass@1 and Claude Fable 5 in Claude Code scored 67.3% pass@1 on 333 GameDevBench Godot 4 tasks, according to the GameDevBench leaderboard from Carnegie Mellon and Princeton researchers p…
NEXUS AI Team outlined a governance framework for enterprise vibe coding, citing Gartner's forecast that 40% of new enterprise production software will use vibe coding by 2028 and a 2026 scan finding 65% of live vibe-cod…
Meta's internal data shows significant lines of code per human-landed diff increased by 106% and diffs per developer per month rose 51% over the past year, with over 80% of that growth from agentic AI, while the percenta…
Anthropic's Claude Code, an AI coding assistant, has been rewritten in Bun, a JavaScript runtime, resulting in significant performance improvements including faster startup times and reduced latency for code analysis and…