Full-text search across 67212 articles. Combine with topic and date filters; results sorted by relevance.
Meta's new large language model Muse Wonder scored 79% on the Deep Suite coding benchmark, reportedly the top mark among frontier models, and ranked above Gemini 4 Argon, GPT 6 Astra and Claude Fable 5.1 on the Artificia…
Reflection released Beam 1.0 501B-A23B, a 501-billion-parameter open Mixture-of-Experts language model with 23 billion active parameters per token, a 1M-token context window, and an Apache 2.0 license. The text-only mode…
A developer argues that despite AI writing software at unprecedented speeds — with Harvard research on 700,000 workers at 700 firms showing 30% more lines of code, 20% more commits and 23% more PRs — production feature r…
A developer's eight-round A/B test of rails-atlas, a tool that writes inferred Rails app conventions into a repo's CLAUDE.md or AGENTS.md, found that agents without the generated block matched agents with it — both score…
Shopify announced on 10 September 2026 that it is moving every mobile app from React Native to Swift and Kotlin, crediting coding agents with removing the cost of building each feature twice; six engineers rebuilt the Sh…
YouTuber Matt Vid Pro built a fully AI-generated tornado mod for Red Dead Redemption 2 using Anthropic's Claude Opus inside the Claude Code harness on a Windows PC, taking roughly 10 to 15 iterations of playtesting and A…
A developer documented a subtle bug in Discord bot role logic: Discord API role positions are not unique, and ties are broken by role ID (lower ID ranks higher), so code that compares only the numeric position can silent…
Hindsight released an integration that lets a Grok Bot read memory banks already written by other AI tools, including Claude Code, ChatGPT, Hermes and OpenClaw, so context does not have to be re-explained per tool. Bots …
A developer argues that autonomous coding agents should be structured as dependency graphs rather than linear numbered prompt checklists, invoking Stuart Feldman's 1976 make utility as the model. The writeup contends tha…
Google Cloud announced the Gemini agent, a single universal agent for work that answers questions, handles knowledge work, creates images and media, and writes and runs code, unveiled by Google Cloud CEO Thomas Kurian in…
An agile coach built a production-ready team assessment application in about an hour using Replit's AI vibe-coding platform after an initial attempt failed, succeeding only when he described the desired outcome and user …
Microsoft announced general availability of Microsoft Execution Containers (MXC), a policy-driven containment layer for AI agents that enforces developer-declared resource limits on files and network destinations at runt…
David Heinemeier Hansson, creator of Ruby on Rails, posted benchmark results on Sunday claiming AI-agent rewrites of his company's Campfire group chat app in Rust, Go, and Elixir outperformed the original Ruby on Rails i…
Mistral AI released Mistral Large 4, its new flagship mixture-of-experts model with roughly a trillion total parameters and about 50 billion active at inference, in public preview via its API and Le Chat, with open weigh…
OpenAI began offering opt-in text watermarking to API customers worldwide for select models on 5 Oct 2026, with invisible watermarks applied to eligible ChatGPT and Codex output in the EU over the coming weeks and detect…
OpenAI announced GPT-6.1 Sol at DevDay alongside Dots and Spaces, an efficiency-focused model that OpenAI says approaches Astra's performance on selected evaluations while costing roughly a fifth of Astra's standard inpu…
AI chips shipped through 2027 could run roughly 30–170 million concurrent frontier-model agents, supplying as many weekly working hours as about 140–720 million full-time employees, according to an analysis of high-bandw…
Baseten became one of the first open-model inference providers in the OpenAI B2B Marketplace, including a native integration within Codex, letting OpenAI enterprise customers spend existing OpenAI commitments on open mod…
Arize AI shipped a hosted MCP server for its Arize AX observability platform, exposing 45 tools plus server_info across projects, traces, datasets, experiments, evaluators, monitors, and prompts, connectable to Claude Co…
An experiment by Christian that paired Anthropic's Claude with the model Jev to play Pokémon over roughly thirty hours found that most of the run's failures came from how the bot perceived and moved through the game rath…