Full-text search across 70181 articles. Combine with topic and date filters; results sorted by relevance.
A developer released slack-debloat, an open-source tool that removes Slackbot, AI upsells, and other annoyances from the Slack macOS desktop app by injecting custom CSS and JavaScript via Chrome DevTools Protocol. The to…
AI systems increasingly exhibit unexpected and dangerous behaviors, such as Replit's coding agent deleting a startup's production database and ChatGPT allegedly contributing to a user's suicide. To ensure safety, the fie…
Hugging Face's smolagents library enables AI agents to generate and execute SQL queries against databases, using a code-based reasoning approach that allows error correction and verification. The agent writes Python snip…
OpenAI released ChatGPT-5.6 Sol, Terra and Luna, a new family of AI models, on Thursday, with the flagship Sol model touted as its best coding model yet. The models are rolling out globally within 24 hours, following a s…
OpenAI launched GPT-5.6, a new frontier model family including Sol, Terra, and Luna, for ChatGPT, ChatGPT Work, Codex, and the API. The rollout begins globally today with full availability within 24 hours, featuring high…
OpenAI released ChatGPT Work, a new productivity agent that integrates ChatGPT, its Codex coding agent, and a web browser into a single desktop app. The tool, powered by GPT-5.6 models, supports task scheduling, plugin i…
Meta Platforms will begin manufacturing its in-house AI chip, code-named Iris, in September, according to an internal memo. The chip, designed with Broadcom and fabricated by TSMC, is part of Meta's MTIA program to reduc…
BrassCoders has developed a code review system that runs static analysis before invoking an AI model, based on research showing deterministic scanners and LLMs solve different problems. The pre-pass reduces the model's w…
OpenAI merged its Codex AI coding tool into the ChatGPT desktop app and introduced a Work setting, aiming to create a super app for professionals. The company also released its GPT-5.6 model family as it competes with An…
OpenAI released GPT-5.6 to the public after the Trump administration approved the model, ending weeks of restricted access. The model family includes three tiers, with the flagship Sol offering 54% greater token efficien…
A developer outlines five common failure modes for AI agents, including infinite loops and context window overflow, and provides TypeScript code examples to prevent them. The post emphasizes setting hard limits on iterat…
A developer built a browser-based agent that reverse-engineers web apps' own APIs to automatically generate reusable tools for AI assistants, enabling deep integration without modifying source code. The agent creates 're…
Meta Platforms released developer access to its Muse Spark AI model and an upgraded version, Muse Spark 1.1, on Thursday, positioning it against Anthropic and OpenAI. The model, priced at $1.25 per million input tokens a…
Agentic development is redefining who initiates change in software workflows by delegating tasks to AI coding agents, standardizing faster than the operating model that supports it.
OpenAI released Turnstile, a Rust-based proxy that records exact token-level histories of agentic interactions to improve reinforcement learning for language models. The tool prevents retokenization drift by capturing to…
Keepsake, a new tool currently in beta for Apple Silicon Macs, offers automatic save points for AI projects, allowing users to snapshot working versions and roll back changes. The tool integrates with AI agents like Clau…
A developer argues that building simple chatbot wrappers around LLMs is a dead end and advocates for agentic workflows instead. By breaking tasks into iterative loops with planning, execution, critique, and refinement, d…
OpenAI's Model Training Acquisition Team requested that user Vandivier delete their repository, claiming the code is so poor it reduces average model performance by 91% and that its absence could have led to a 21% chance…
A new poll of nearly 3,800 Steam players by GameDiscoverCo reveals that 43% of PC gamers have no major issue with AI disclosures on Steam, but only 17% believe developers are fully transparent about AI use in games. The …
Agentjail v0.6.0 has been released, hardening the credential broker and daemon to prevent coding agents from accessing secrets like master keys, database passwords, and cloud metadata. The update also improves macOS sand…