
Microsoft's Free 21-Lesson AI Course Quietly Hit 117k GitHub Stars
Microsoft's free 21-lesson course on building generative AI apps has crossed 117k stars, covering prompt engineering through agents with Python and TypeScript samples.
20 stories

Microsoft's free 21-lesson course on building generative AI apps has crossed 117k stars, covering prompt engineering through agents with Python and TypeScript samples.

IWE turns plain Markdown files into a structured knowledge graph that both you and AI agents can query, navigate, and refactor — no cloud, no embeddings, no lock-in.

AA-Briefcase reveals frontier agents take 10–25 minutes per task on realistic knowledge work, exposing a new speed-quality tradeoff that existing benchmarks miss entirely
Higgsfield brings Seedance 2.0 to native 4K on its platform, covering pre-production, production, and post in a single AI pipeline.

Cursor's SDK lets any product embed full coding agents in weeks, with Notion as the first major proof of concept

Perplexity's Computer agent now plugs into Docusign, NetDocuments, and Midpage, turning a general AI assistant into a full legal workflow engine

VSS 3 adds 16 agent skills, Nemotron Nano Omni integration, and production-ready 3D multi-camera tracking so coding agents can query any video library in plain English

OpenAI ships a new GPT-5.5 Instant with sharper intent understanding, fewer hallucinations, and smarter personalization for ChatGPT's most-used model

Exa Connect lets AI agents tap ZoomInfo, Crunchbase, Similarweb, and a dozen more premium data providers through a single API call, moving beyond the public web.

Zed v1.8 lands with a new agent terminal init hook, smarter Git internals, and two handy editor selection commands

Google Research reveals why reasoning models recall more facts, even for simple questions, and how hallucinated reasoning steps poison final answers

NVIDIA's open-source Earth2Studio gives Python developers a unified API to run, chain, and benchmark state-of-the-art AI weather models in minutes

agtx is an open-source terminal kanban board that runs multiple AI coding agents in parallel, from idea to merge, hands-free

Boston Dynamics is transforming a 323,000 sq ft Waltham facility into a robotics and AI hub, backed by a $25M state EDIP award and $100M company investment targeting 1,250 new jobs.

NeMo AutoModel layers Expert Parallelism and DeepEP on top of Transformers v5, delivering 3.4–3.7x faster MoE fine-tuning with a single import change

Sakana AI's Fugu Ultra is now on OpenRouter — a 7B orchestrator that routes across frontier models to match Claude Fable 5 and Mythos on benchmarks, without training a single frontier model.
xAI launches the Grok Build Plugin Marketplace in beta, shipping six partner integrations including MongoDB, Vercel, Sentry, and Cloudflare from your terminal

NVIDIA compresses its 120B flagship model to 75B using Iterative Puzzle, doubling server throughput while preserving near-identical accuracy across reasoning, coding, and long-context tasks.
Runway's new Ad Localization App turns a single image ad into market-ready versions in any language with one click, no design work required.

OpenAI's first custom silicon, Jalapeño, goes from blank slate to running GPT-5.3 workloads in nine months — and it's just the start of a multi-generation compute platform targeting gigawatt-scale deployment.