AI news for builders and product teamsUpdated Nov 18, 2026, 09:00 UTC
AI news, straight from the source.
Releases, research, and ideas for developers and product teams. Updated every hour.
Across the feeds
Newest first
LangChain published a blog post describing how healthcare AI teams Abridge and Included Health use LangSmith to convert clinical review into reusable datasets, evaluators, and release gates. It cites Abridge cutting its release cycle from one to two months to days, and Included Health reporting a 75% engagement lift and flagging over 99% of high-risk situations.

Patricia and James Poitras have given MIT $10 million to fund 50 two-year fellowships through the Poitras Center for Psychiatric Disorders Research at the McGovern Institute, with five fellowships awarded annually for a decade. Five researchers were named to the inaugural cohort.

Simon Willison released llm 0.36, adding support for new OpenAI models gpt-6-sol and gpt-6-luna, a way for model plugins to declare they do not support conversations, and collapsed reasoning traces in llm logs Markdown output.

Gary Marcus published a commentary piece dated September 22, 2026, responding to reports that Sam Altman will brief the UN Security Council on AI, and quoting criticism of that briefing from Drainpipe.io founder Dominic Romano.

Simon Willison quoted a TikTok creator, @therealcornpop, who described telltale signs of AI-written scripts for TikTok and YouTube videos, including stock phrasing patterns and a lack of distinctive voice or opinion.

NVIDIA described Confidential Computing adaptations in TensorRT LLM for private AI inference on Blackwell GPUs. On eight B200 GPUs with DeepSeek-R1, confidential compute retained 96.1-98.2% of baseline output-token throughput with 1.2-4.3% per-token latency overhead across concurrency 1-16.

NVIDIA announced Topograph, an open source toolkit that discovers cluster network topology and publishes it in formats schedulers can consume, including Kubernetes node labels, Slurm configuration, and Slinky ConfigMaps. It supports cloud providers Google Cloud, Lambda, Nebius, Nscale, and OCI, plus on-premises InfiniBand, Spectrum-X, and Multi-Node NVLink domains.

Simon Willison released llm-anthropic 0.29, which adds support for Claude Opus 5.5, invoked via the command llm -m claude-opus-5.5 "prompt goes here".

Anthropic released Claude Opus 5.5, which it says delivers Fable 5.1-level performance for most work at about 40% lower cost, with tokens priced 20% below Opus 5. The company also raised five-hour usage limits by 20% and said Claude Sonnet 5.5 and Haiku 5.5 are coming in the next few weeks.

Simon Willison released llm-typesafe 0.1a0, a new plugin for LLM that adds support for TypeSafe AI's Jev model, enabling yes/no, choice, and scoring questions from the command line.

vLLM is introducing new "HW agnostic" layers to keep the project portable across diverse hardware as its frontier optimizations move away from fullgraph torch.compile. On NVIDIA H100 GPUs, these layers achieve total token throughput within 3.4% of the native implementation (geometric mean across three recent models).

Stanford researchers described Paper2Agent, an open-source framework that turns academic papers and their code into interactive AI agents, in Nature on 16 September. Tested on computational biology, it generated 22 validated AlphaGenome tools in about 45 minutes for under US $15.

Google has open-sourced AX, an Apache 2.0-licensed orchestrator and declarative runtime for autonomous AI agent workloads, hosted at agentexecutor.io and on GitHub as google/ax. Running on Agent Substrate, it treats agents as stateful actors and offers sub-second task suspension and resumption.

A PyTorch case study describes how Shopify built a continual learning loop using PyTorch and vLLM, turning production failures into model weight updates. Shopify reports its GraphQL agent surpassed frontier-model quality while cutting serving costs by 96% and reducing latency.

Nathan Lambert interviews Epoch AI senior researcher Jean-Stanislas "JS" Denain on the state and trajectory of AI, covering predictions for recursive self-improvement, the US-China model gap, robotics, open versus closed model safety, and frontier post-training. Both hosts emphasize their uncertainty about where AI is heading.

Jev, a classification-style model built to run inside tools, is now open to everyone, and builders are showing demos including sponsor-segment skipping, comment filtering and Gmail search. Meta's personal agent Muse has a Mac app and connector developer platform, though Amazon has begun blocking it from shopping.

NVIDIA announced DLSS 5 with 3D-Guided Neural Rendering and granular developer controls, alongside NVIDIA ACE updates and RTX Kit capabilities for game developers. DLSS 5 is available now in NBA 2K27 for GeForce RTX 50 Series GPUs.

Zvi Mowshowitz's newsletter recounts how AI existential-risk concerns briefly gained political traction, then reversed after Trump called the issue a 'hoax'. It describes Trump's September 19 statement, his plan for an 'AI Force' and AI czar, and his poll on renaming AI, citing concerns from Sacks, Huang and Zuckerberg.

A sponsored CoolIT article argues that beyond roughly 250 kW per rack, hybrid liquid-and-air cooling stops working and near-total liquid heat capture is required. CoolIT says its modular coldplates, proven across six generations of fanless designs, support fanless AI servers.

Xiaomi released MiMo-V2.6-Pro, an open weights omnimodal model with 1.02T total and 42B active parameters, which Artificial Analysis says debuts as the top open-weights model on its Intelligence Index at 46. Xiaomi says the model was trained for $3M.