
Microsoft Breaks Free From OpenAI With Seven In-House MAI Models
Microsoft AI's first full-stack lineup spans reasoning, code, image, voice and transcription, trained from scratch on Maia 200 silicon with no distillation.
20 stories

Microsoft AI's first full-stack lineup spans reasoning, code, image, voice and transcription, trained from scratch on Maia 200 silicon with no distillation.

GitHub's new desktop app turns AI agent chaos into a single control center, with canvases, sandboxes, and enterprise deployment via Rayfin
An internal OpenAI reasoning model autonomously cracked the planar unit distance problem, an open question Erdős posed in 1946, using algebraic number theory.
Google's PHRM system uses your phone's selfie camera to passively measure heart rate after face unlock, matching wearables across all skin tones.

Factory's new Router automatically picks the best LLM per task and fails over across providers, cutting token spend 20-25% while keeping sessions alive.
Prime Intellect Lab adds day-0 post-training for NVIDIA's 550B Nemotron 3 Ultra, opening the closed-lab agent loop to every team on Blackwell.
OpenAI folds content safety signals directly into Responses and Completions, removing the need for a separate moderation call before or after generation.

Cursor 3.7 adds point-and-click canvas editing and a live token breakdown report so you can see and fix context bloat in real time

Microsoft's new text-to-speech model jumps from English-only to 15 languages, beats its predecessor 72% of the time, and fools listeners on speaker identity.
Higgsfield AI's new Claude skill packages a full pipeline that turns brand assets into scroll-driven animated marketing sites from chat.

OpenAI's Build iOS Apps plugin brings a live simulator, SwiftUI previews, and hot reload directly into Codex — closing the iOS dev feedback loop without leaving the agent.

NVIDIA's Nemotron 3 Ultra is the most capable open-weight model from a US lab, delivering 5x faster inference than rivals at 30% lower cost for agentic workloads
Cognition will fund Devin usage up to $10M if its agent fails to deliver enough engineering value, backed by a new hours-equivalent estimator.
A new regret metric called RP-Regret enables agents to reason about adaptive opponents in repeated games, unlocking cooperative equilibria that standard external regret cannot reach

Replit's Shopify integration lets anyone describe a store idea and go live in under 10 minutes — no code, no theme picking, no developer needed.

Microsoft Research's CLSA architecture computes sparse attention routing once and shares it across all decoder layers, hitting 17.1x throughput gains at 128K context with near-zero quality loss.
Pika's new Group Chat feature lets a single AI agent join shared conversations with friends, family, or teammates — generating media and completing tasks in real time
LM Studio launches an official iPhone and iPad app that tunnels into the heavy local models running on your Mac or workstation.
Google's NotebookLM team published a Sherlock Holmes themed notebook that reframes studying as detective work, showing how persona prompting can transform passive review into an interactive deduction game.
Unsloth releases GGUF-quantized versions of NVIDIA's 550B Nemotron 3 Ultra, making the fastest US open-weight frontier model runnable locally for the first time.