This is a big step for anyone who wants to turn a useful AI workflow into something other people can actually discover and use. Anthropic has opened a developer portal for submitting Claude plugins, giving builders a clearer path from MCP-powered automation to directory distribution, review, and analytics. • Plugins bundling MCP connectors and Agent Skills • Single remote MCP server submissions • GitHub-hosted plugin bundles • Auto-validation and safety scans before review • Review feedback and publishing controls • Install analytics by product surface and version • Search and listing-view insights for discovery • MCP Apps enabling interactive in-chat interfaces • Enterprise Managed Auth supporting zero-touch OAuth • One discovery experience rolling out across Claude and Claude Code To me, this matters because the winning automations will not just be clever, they will be easy to install, secure, measurable, and discoverable. What plugin would you guys build first if you could package one workflow you use every week? Read the full article here: https://claude.com/blog/build-plugins-for-claude

Anthropic is merging Cowork and Chat into one Claude, so you hand over the task and it figures out what the job needs. - Rolling out to Pro and Max first over the next few weeks. Team and Free plans soon after - Nothing to turn on. Your Cowork chats, projects, connectors, and skills all carry over - New Claude Docs and Claude Slides launched same day, plus Design now works inside conversations - Slides export to PowerPoint or PDF. All three are beta on paid plans - Keeps working after you close your laptop. Check progress from your phone - Schedule recurring work, like a Monday report Claude starts without being asked - Report and a 5-slide deck from one conversation, so they actually match - Default is Claude asks before acting. You can loosen that if you trust it - Enterprise admins get 30+ days notice before changes hit their org The real shift here for us builders: the line between "chat tool" and "automation tool" just got really blurry. Scheduled, context-aware work from a normal conversation is basically a lightweight agent workflow with zero setup. https://claude.com/blog/cowork-is-now-claude

Google just dropped Gemini 3.5 Transcribe, and it's their most accurate speech-to-text model yet, built to handle real conversations instead of just clean studio audio. - Word error rate down to 4.0% for streaming and 2.6% for pre-recorded audio, measured by Artificial Analysis - That's a big jump from their old model, Chirp 3. Time to final transcription is 70% faster - Handles up to 85 languages and picks up on accents and dialects automatically - Cleans up your speech as you talk. Cuts filler words like "um" and "ah" and fixes self-corrections like "let's meet Tuesday, no wait, Wednesday" - Can identify up to 3 speakers in recorded audio with timestamps, which is huge for meeting notes and call analytics - It can also call other Gemini models to do stuff like generate images or analyze files, all triggered by voice - Already live in things like Rambler on Android and the Gemini app on macOS, and rolling into Chrome soon - Developers can plug it into their own apps now through the Gemini API and Google AI Studio If you're building any kind of voice agent, meeting notes tool, or automation that starts with "listen to this audio," this is worth testing against whatever you're using now. I still use Deepgram Nova 3 or Flux for my voice agents. But I might start using Gemini 3.5 Transcribe if I like the results. Will ultimately come down to latency. https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-5-transcribe

Google just quietly turned Sheets into an app builder. They rolled out Sheets canvas, a Gemini-powered layer that turns your spreadsheet rows into interactive mini-apps from a plain English prompt. - No formulas, no code, you just describe what you want and Gemini builds it - Sits on top of your data as a read-write layer, so edits sync both directions in real time - Lives as a tab inside Sheets, so sharing works exactly like a normal sheet - You can keep prompting to tweak layout, design, and functionality after it's built - Google's own examples are study trackers, fantasy football dashboards, and drag-and-drop wedding seating charts - Available globally in English right now, but only on Google AI Pro and Ultra, Workspace Business or Enterprise Standard and Plus, and the AI Pro for Education add-on This is the same pattern we keep seeing across the board. The spreadsheet stops being the destination and becomes the database, and the interface gets generated on demand. That's a big deal for anyone who has ever built a client-facing dashboard on top of a Sheet with Glide, Softr, or AppSheet. What would you guys build with this first? I'm curious whether it actually replaces a no-code app layer or just makes your sheets prettier. https://blog.google/products-and-platforms/products/workspace/sheets-canvas-for-google-sheets-spreadsheets/

I love seeing these type of use cases for AI. Imagine getting an extra day to prepare before a major cyclone hits. Google DeepMind's WeatherNext model is delivering more accurate cyclone forecasts while being open sourced for researchers and forecasters to build on. • More than 24 hours of added forecast lead time • Three-day predictions matching older two-day forecasts • Track, intensity, and wind structure in one model • 1,000 possible storm scenarios per cyclone • 15-day forecasts in under one minute on a TPU • Training across nearly 5,000 historical storms • Open-source weights and code for wider experimentation • Potential uses in disaster response, renewable energy, and infrastructure planning What would you guys build if you could connect this kind of forecasting AI to automated alerts, logistics, or local response workflows? Read the full article here: https://deepmind.google/blog/weathernext-ai-model-achieves-breakthrough-in-forecasting-cyclones