The SlopCodeBench study found that code generated by AI agents has approximately twice the duplication and complexity of human code, with even the latest models achieving a 0% pass rate on strict tests.
The AI SDK harness layer now supports GitHub Copilot via an official adapter, enabling coding agent switching without code changes.
The OpenAI Agents API integrates with Vercel Sandbox to provide isolated code execution environments and session persistence.
OpenAI has launched the public beta of its Agents API, built on the Codex harness.
ZeroModels, a collection of pretrained models built with Keras 3, has been released and is runnable on JAX, PyTorch, and TensorFlow.
Q2D-Web has released a large-scale web search retriever evaluation benchmark featuring 190 million documents and 10 languages.
Anthropic's evaluation of AI models' tactical intelligence targeting and drone weapons development capabilities revealed that top-tier models perform at or above the level of human experts.
OpenAI has launched GPT-Live-1 on its API, supporting latency-free full-duplex voice conversations and backend model integration.
The MultiMatte model, which fine-tunes SAM 3 with LoRA to isolate specific objects via text prompts, has been released, showing significant performance improvements on major benchmarks.
Vercel AI Gateway has extended the free access period for Tako Search, a model-agnostic search tool, until September 30.
Vercel reduced CDN metadata lookup P99 latency by 91% and improved build speeds by introducing Shard-based Group Fetching.
Shopify announced its transition to Swift/Kotlin native development, citing that improved LLM agent performance has reduced the maintenance advantages of React Native.
Vercel promotes FastAPI frontends and static files to the CDN at build time, serving them without function invocations.
Gradio has released 'Workflow1111', which replicates the key features of AUTOMATIC1111 in a single workflow canvas.
Vercel Sandbox has expanded its support from 4 to all 20 computing regions.
DeepSeek has released V4.1-Flash, which offers superior performance and efficiency compared to V4-Pro, along with reduced API prices.
The Apple A20 Pro offers approximately 115GB/s of memory bandwidth through a 2nm process, 32-core Neural Engine, and 96-bit LPDDR5X.
AWS has released a feature that allows users to instantly copy Amazon EBS volumes to different accounts and re-encrypt them.
Hugging Face has introduced a hybrid search system for Papers with Code that combines PostgreSQL full-text search and pgvector vector search using RRF.
Andrew Tulloch, Meta's highest-paid AI researcher, left the company after the launch of its open-source models.
Apple is launching Siri AI in beta alongside OS 27, signaling usage limits and future monetization.
Apple has unveiled the Apple Watch Series 12, featuring the S11 chip and a new health sensing system.
Apple has unveiled its first foldable smartphone, the iPhone Duo, with a launch date set for October 23.
HP has started taking orders for the ZGX Fury, an enterprise AI inference workstation featuring the NVIDIA GB300 Grace Blackwell Ultra superchip and 748GB of unified memory.
AI Briefing