Earendil released Pi 1.0 and Pi Durable, both of which reached the front page of Hacker News, with Pi Durable porting Pi to TypeScript and externalizing all stateful components so agents resume from t…
Google DeepMind released Gemini 4 Argon, a model it says takes first place on 13 of 19 published benchmarks against GPT-6 Astra and Claude Opus 5.5, with access initially limited to government users a…
OpenAI's API team has shipped a Decisions API that is currently a Luna wrapper, part of a new developer stack that includes async tool calling, mid-turn steering, WebSockets, UltraFast inference, prom…
OpenAI announced Dots, GPT-6.1 Sol, an ultrafast mode, the Decisions API, ChatGPT Spaces, and a Marketplace at its DevDay 2026 event, with the company reporting 1.2 billion weekly active ChatGPT users…
Anthropic's Claude Opus 5.5, which shipped this week, leads SimpleBench at 88.4% and ranks as Anthropic's best vision model to date, according to AI Battle and @skalskip92, performing better than Fabl…
Anthropic released Claude Sonnet 5.5, the latest addition to its Claude model line, announced alongside a podcast episode featuring Anthropic's Thariq Shihipar discussing Claude Code's evolving interf…
Runway introduced GWM Worlds 2, a research preview of its "autoregressive diffusion" world model that generates high-fidelity video and audio as real-time interactive simulation, featuring a new input…
Meta used its Connect 2026 keynote to present Muse, its personal agent, as the center of a hardware-plus-agent strategy, shipping voice and real-time video features and new glasses while teasing but n…
Radical Numerics co-founder Eric Nguyen said defense is currently losing the bio-security arms race, arguing that the same genomic language models that increase biological capability can also keep def…
Anthropic released Claude Opus 5.5, the first model in its new Claude 5.5 family, which the company says performs at the level of Claude Fable 5.1 for most tasks while costing 40% less to run than Opu…
Google researcher John Platt and his team developed Empirical Research Assistance (ERA), an AI system that uses Gemini and a Monte Carlo Tree Search-style algorithm to automate scientific discovery by…
TypeSafe AI CEO Diogo Almeida introduced Jev, a new model family the company positions for production systems rather than frontier "God" models, in a launch video now viewed roughly 40 million times. …
Six independent reimplementations of Jev, a non-generative "System 1" decision model launched on Wednesday, appeared within two days of its release, according to AINews' 9/17/2026-9/18/2026 roundup. T…
Steve Yegge shut down Gas Town, admitting that despite spending many thousands of dollars a month on coding agent subscriptions, Gas Town was the only thing he ever built with them. Separately, Databr…
AIUC announced a $40M Series A round today, led by cofounder Rune Kvist, to build standards, testing, and insurance infrastructure for AI agents through its AIUC-1 standard. AIUC-1 is backed by real i…
TypeSafe launched Jev, a "System One Model" that only decides, classifies, routes, and scores, claiming it is over 100x faster and over 200x cheaper than small frontier LLMs, and the announcement sat …
Good Start Labs, spun out of Every last October with $3.6 million in funding from General Catalyst, Inovia, Every, and angel investors, is training AI models on games to teach transferable skills, co-…
The AI Evaluator Forum published AEF-1, a proposed baseline standard for independent third-party AI evaluations covering access, conflicts of interest, funding relationships, recusal, and transparency…
Richard Socher, founder of You.com, AIX Ventures and now Recursive, said Recursive has raised a $4.65B seed round and assembled open-endedness and self-improving-agent researchers to build an AI syste…
Vinoo Ganesh, CEO of Kepler, writes that the forward deployed engineer (FDE) role has become one of the hottest jobs in AI, with labs, startups and private equity firms hiring engineers to sit inside …