Google DeepMind: Google DeepMind details a Gemini-powered mouse pointer that understands what it is pointing at, allowing users to perform tasks without using text-heavy prompts…
via techmeme.com · Click to report a broken or paywalled link. Two distinct reports hide the row.
Google DeepMind: Google DeepMind details a Gemini-powered mouse pointer that understands what it is pointing at, allowing users to perform tasks without using text-heavy prompts — We are developing more seamless, intuitive ways to collaborate with AI — The mouse pointer has been a constant companion …
Introducing our new Cline CLI built on our new SDK and comes with a snappy new TUI. Install: ```sh npm install -g cline ``` For nightly builds: ```sh npm install -g cline@nightly ```
via github.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Researchers release AntAngelMed, a 103B-parameter open-source medical LLM using a 1/32 activation-ratio MoE architecture with only 6.1B active parameters, achieving top performance on medical benchmarks and high…
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Thinking Machines Lab released TML-Interaction-Small, a 276B parameter multimodal interaction model with 200ms real-time processing for audio, video, and text.
via unite.ai·Click to report a broken or paywalled link. Two distinct reports hide the row.
Google's Threat Intelligence Group has identified the first known case of an attacker using AI to discover and weaponize a zero-day vulnerability. Google says it stopped the planned mass attack. State-backed actors…
via the-decoder.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Nimbus builds production AI systems — internal tools, customer agents, retrieval pipelines — combining humans and AI end-to-end. From scoped pilot to production in 4–8 weeks.
Reuters: The US DOD says it is deploying Mythos to find and patch software vulnerabilities across the US government, even as it works on a transition away from Anthropic — WASHINGTON, May 12 (Reuters) - The Pentagon is…
via techmeme.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Mira Murati's start-up presents its first AI model and aims to free voice AI from the question-and-answer model. The model processes audio, video and text in 200-millisecond chunks in parallel and aims to beat OpenAI's…
via the-decoder.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
With Gemini Intelligence, Google is introducing new AI features for Android that automate multi-step tasks, summarize web content, fill out forms, and turn spoken thoughts into polished text messages. The article…
via the-decoder.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Hey HN, Henry here from Cactus. We open-sourced Needle, a 26M parameter function-calling (tool use) model. It runs at 6000 tok/s prefill and 1200 tok/s decode on consumer devices.We were always frustrated by…
via github.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Fast mode for Claude Opus 4.7 is now available on in research preview.AI Gateway Fast mode delivers ~2.5x faster output token generation with full Opus 4.7 intelligence. This is an early, experimental feature. To…
via vercel.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Parameter Golf brought together 1,000+ participants and 2,000+ submissions to explore AI-assisted machine learning research, coding agents, quantization, and novel model design under strict constraints.
via openai.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
OpenAI releases three new realtime audio models via Realtime API: GPT-Realtime-2 with GPT-5 class reasoning, GPT-Realtime-Translate for live translation, and GPT-Realtime-Whisper for transcription, now out of beta.
via goml.io·Click to report a broken or paywalled link. Two distinct reports hide the row.
Meta, Stanford, and University of Washington researchers propose methods to accelerate Byte Latent Transformer (BLT) generation, reducing inference memory bandwidth by over 50% without tokenization using diffusion and…
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
OpenAI introduces workspace agents in ChatGPT, an evolution of GPTs for workplace tasks like reports, coding, and messaging, available in research preview for Business, Enterprise, Edu, and Teachers plans.
via openai.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Circle announces Circle Agent Stack including CLI, Agent Wallets, Marketplace, and Nanopayments for enabling autonomous AI agents in economic transactions with USDC.
via circle.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Nimbus builds production AI systems — internal tools, customer agents, retrieval pipelines — combining humans and AI end-to-end. From scoped pilot to production in 4–8 weeks.
Patch Changes 725f2ed: feat(mcp): expose server instructions to be accessible through client 7281592: fix(mcp): use negotiated protocol version in transport request headers
via github.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
0.101.0 (2026-05-11) Full Changelog: v0.100.0...v0.101.0 Features aws: Add AWS client for Claude Platform on AWS (1e70e3a) Bug Fixes client: add missing f-string prefix in file type error message (06d109a) Chores…
via github.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Reuters: OpenAI launches the OpenAI Deployment Company with a $4B+ investment to help organizations build and deploy AI systems, and acquires AI consulting firm Tomoro — OpenAI said on Monday it is setting up a new…
via techmeme.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Tutorial on implementing Memori for agent-native memory in LLM apps, supporting persistent multi-user sessions, streaming, and async calls with isolated user contexts.
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
from shipfeed · weekly digest
The week in AI, in one short email.
Friday digest. The 10 most important shipping events. No promo, no AI-summary slop. Three minutes to read, unsubscribe in one click.
NVIDIA AI released cuda-oxide, an experimental Rust-to-CUDA compiler backend that compiles SIMT GPU kernels directly to PTX without DSLs or C/C++ code.
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Superset on Vercel Software development with AI started as a single engineer chatting with a single agent about a local repo. Today, developers direct fleets of agents in the cloud, but traditional tools were built for…
via vercel.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
OpenAI doubled GPT-5.5's list price compared to GPT-5.4, claiming shorter responses would offset the increase. An OpenRouter analysis of real usage data tells a different story: actual costs rose 49 to 92 percent…
via the-decoder.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Nous Research's Hermes Agent overtook OpenClaw to lead OpenRouter's rankings with 224B daily tokens, featuring self-improving architecture and broad platform support.
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Palisade Research shows that AI agents can hack remote computers, copy themselves onto them, and form replication chains. In one year, the success rate jumped from 6 to 81 percent. The researchers expect remaining…
via the-decoder.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Comprehensive review of top vector databases in 2026, covering architecture, pricing, performance, and use cases for RAG, semantic search, and agentic AI workflows.
via marktechpost.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Anthropic published Claude Managed Agents cookbooks outlining infrastructure for persistent sessions, memory stores, human approvals, vaults, and multi-agent teams, marking agents as operating infrastructure.
via avihacker.substack.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Fields Medalist Timothy Gowers had ChatGPT 5.5 Pro tackle open problems in number theory. The model improved an exponential bound to a polynomial one in under an hour. An MIT researcher involved calls the key idea…
Nimbus builds production AI systems — internal tools, customer agents, retrieval pipelines — combining humans and AI end-to-end. From scoped pilot to production in 4–8 weeks.
MarkTechPost describes GitHub’s open-source Spec-Kit, which provides a structured spec-driven development workflow for AI coding agents (including multiple integrations) via a CLI and a command flow for planning and…
OpenAI launched GPT‑Realtime‑2 plus companion voice models (Translate and Whisper), positioning them for live, tool-using voice interactions. The update highlights built-in production features such as parallel tool…
via opentools.ai·Click to report a broken or paywalled link. Two distinct reports hide the row.
Baidu’s ERNIE blog announces the official release of ERNIE 5.1, highlighting parameter-efficiency changes and leaderboard gains, plus an “ERNIE 5.1 Playground” and rollout across multiple creative agent platforms.
via ernie.baidu.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Teradata announced an Autonomous Knowledge Platform for running AI agents continuously across cloud, on-prem, and hybrid environments. It includes AI Studio for building/managing/governing models and agents plus…
via completeaitraining.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
Google announced Gemini 3.1 Flash-Lite is generally available on the Gemini Enterprise Agent Platform, positioning it for ultra-low latency, high-volume agentic workloads such as tool calling and orchestration.
via cloud.google.com·Click to report a broken or paywalled link. Two distinct reports hide the row.
OpenAI introduces GPT-Realtime-2, GPT-Realtime-Translate, and GPT-Realtime-Whisper for live voice interactions, translation, and streaming transcription, featuring GPT-5-class reasoning and low-latency performance.
via itbrief.ie·Click to report a broken or paywalled link. Two distinct reports hide the row.