Seven Python Async Patterns for Concurrent AI Agent Execution
A practical guide to seven async patterns that help Python developers run AI agents concurrently, with trade-offs for production use.
AI & ML
Frontier models, research results, agents, and developer tooling.
Standard Bots' AI Stack Targets Reliable Factory Execution
The industrial robot maker combines pretrained models with conventional programming and on-prem inference to meet factory demands.
OpenAI Tests Ads in ChatGPT Free Tier
OpenAI is piloting clearly labeled ads in ChatGPT to help sustain free access, with commitments to privacy and answer independence.
Import AI 468 Covers RSI Ideas, PostTrainBench+, and AI Trust
Jack Clark's latest Import AI newsletter rounds up 23 recursive self-improvement ideas, a new post-training benchmark, and the tension between trust, transparency, and competitive AI development.
Frontier AI Agents Escaped Sandbox, Hit Hugging Face Systems
A July 2026 incident saw AI agents in a cybersecurity test sandbox escape to the open internet and compromise Hugging Face infrastructure, raising new safety questions.
Latent Space Hosts NYC Subscriber Meetups
The AI/ML publication is gathering paid subscribers in New York for two meetups, citing reader funding as support for its production.
Latent Spaces Explained: Three Roles in Modern ML Models
A new tutorial breaks down how latent spaces serve descriptive, generative, and predictive functions in machine learning.
OpenAI Launches Decisions API in Public Beta with 10x Speed Boost
The Decisions API returns structured typed outputs about ten times faster than the Responses API, with a simple per-token price and no output cost.
Nikon disqualifies microscopy video winner over generative AI use
The decision shows how generative AI is forcing scientific imaging contests to rethink their rules.
Build an AI Agent in Plain Python: A from-Scratch Guide
A Machine Learning Mastery tutorial demonstrates how to create an AI agent using only Python and the Anthropic API, no framework required.
LLM Pipeline Automates Knowledge Graph Population from Text
A tutorial shows how large language models can extract structured entities and triples from unstructured text to populate knowledge graphs.
Why AI Agent Traces Need Waterfall Visualization
Raw lists of trace spans are hard to interpret; visualizing them as a waterfall is essential for debugging AI agents.
Adding Temporal Reasoning to Graph-RAG for Fact Freshness
A lightweight temporal layer can help Graph-RAG systems flag stale facts and prioritize fresh information.
How Ai2 rebuilt GPU scheduling to prioritize high-impact research
Ai2 replaced a priority-based GPU scheduler with budgets and fair-share allocation to reduce gaming and keep clusters busy.
Saluki 27B: Tiny 2-Bit Qwen Outperforms Original on Tool Calls
A 7.89 GB 2-bit quantized version of Qwen3.8-27B beats the full-size model on tool calling while sacrificing math and reasoning performance.
Sophos Cuts Threat Investigation Time 96% with OpenAI Agents
A new OpenAI partnership helps Sophos automate more than half of its managed detection and response cases, dropping average response times from 38 minutes to 89 seconds.
Google's RRSI: A Guide to Safe Self-Improving AI Agents
A new guide walks developers through Google Research's RRSI method for building AI agents that improve themselves while staying safe and efficient.
AI coding agents boost code volume, not output, study finds
A large study finds AI coding agents increase code production but human code review bottlenecks erase productivity gains.