Latest Artificial Intelligence News & Technical Dispatches

OpenAI's 'Research Acceleration: The View Inside OpenAI' scores 122 HN points. An unprecedented inside look at the infrastructure, culture, and velocity driving AGI development. Full analysis for 2026.

OKF Agent Memory implements Google's OKF v0.2 spec with sub-300µs BM25 search, embedded MCP server, and progressive persistence. Benchmark against Graphiti (Zep AI) for AI coding agent memory patterns in 2026.

UseAgent goes open source (283 GitHub stars) — AI coworkers with their own cloud computer, your tools and context, handing back finished work websites, decks, and reports. Full analysis of the open-source AI coworker platform transforming how teams work with AI agents in 2026.

OrcaReplay by Continuum AI (OrcaRouter.ai team) gives AI agents time travel capabilities — record, replay, fork, and debug any agent run with any model. A production-grade agent observability tool for debugging complex multi-step agent failures in 2026.

Easel (380+ GitHub stars from ZJU-REAL) is an open-source AI agent for social media that discovers trends, creates platform-native content, publishes across channels, and learns from engagement metrics. A full architecture deep dive for automated social media workflows in 2026.

PhiloLabs open-sourced Fable 5.1 under Apache 2.0 on September 2, scoring 158 HN points. The causal latent diffusion framework lets AI agents simulate action outcomes before executing — proven 47% reduction in incorrect autonomous decisions.

Fable 5.1 (158 HN pts, PhiloLabs), General Intuition ($6B valuation), and NVIDIA NOOA — three radically different approaches to world modeling for autonomous agents. Head-to-head on simulation fidelity, compute cost, causal reasoning, and enterprise production readiness.

Google's Gemini 3.8 Flash launched to 863 HN points on September 2 — its highest-rated model launch of 2026. The Flash Cyber variant embeds 27M security advisories into a dedicated Mixture of Security Experts layer, achieving 89.4% zero-day detection recall.

Meta's Muse Spark 1.3 launched to 429 HN points — the company's highest-rated AI launch of 2026. Generates 512x512 images in 0.8 seconds on consumer GPUs with 4.2 FID, released under CC BY-NC 4.0.

WebLLM runs 27B models in-browser at 45 tok/s — 86% of Ollama's native speed. Head-to-head comparison on latency, VRAM, privacy, deployment complexity, and which wins for your production agent pipeline.

Gemini 3.8 Flash scored 863 HN points — Google's highest-rated AI launch of 2026. Deep dive into the Flash Cyber security-first variant, its 89.4% zero-day recall on SECURE-bench, and what the 3.8 architecture means for enterprise AI.

OpenAI launches GPT-5.6 Sol API with sub-100ms time-to-first-token latency and 180 tok/s throughput at $2.50 per million input tokens. The fastest inference launch in OpenAI's history positions Sol as the premium choice for real-time agent applications requiring instant responses.

Evaluating AI agents is fundamentally different from evaluating LLMs. Agents make tool calls, follow multi-step plans, use external data, and produce outputs that are hard to score with static benchmarks. This guide covers production-grade eval harnesses for task completion, tool accuracy, latency, cost, and regression detection.

The Model Context Protocol registry surpassed 10,000 registered MCP servers in September 2026, marking a pivotal milestone for the AI agent ecosystem. From 2,000 servers in January to 10,000 in September, MCP has become the standard protocol for AI agent tool integration.

RAG, fine-tuning, and agentic retrieval are the three main approaches to injecting knowledge into LLMs. This comprehensive comparison covers accuracy, latency, cost, and maintenance trade-offs with a decision framework for enterprise use cases in 2026.