I help organizations get the most of out of their collaboration tools and applications. I post about AI, Copilot, technology, and when in season, Da Bears!
Fun fact: $6,000,000 in 1973 (when The Six Million Dollar Man originally premiered as a pilot movie on March 7, 1973) has the equivalent purchasing power of $45,532,432.43 today.
Today marks a new chapter for Windows, as we bring unmetered intelligence to every desk and every home, and make every PC a place where agents can work securely on your behalf.
Some highlights of what we announced:
โข MAI-Code-1.1 Flash: 137B parameter coding model w/ 256K context window, which is now optimized to run on your PC!
โข GitHub Copilot now hands off work to local models like MAI-Code-1.1 Flash, helping projects cost a lot less without sacrificing quality.
โข With Hybrid Intelligence, Copilot can now take action directly on the PC and keep sensitive work on your device.
โข And with Code in Copilot, you can essentially build any software you need on your desktop, without any cloud token spend, and itโs just super at it. Youโre no longer limited to whatโs in an app store! Your PC becomes an infinite software factory.
โข Security is foundational to all this, which is why we are also bringing together Windows and Agent 365 so agents can work within secure boundaries on-device, including MXC a local sandbox for agent execution. Windows becomes your secure agent box!
โข All this comes to life on a new generation of devices, like Surface Laptop Ultra, powered by NVIDIA RTX Spark.
Canโt wait to see what you build with all this.
@metraUPW you guys need to figure out freight train interference, cause your schedule mean almost absolutely Nothing because every train is 10 to 15 minutes late on a good day.
Agent Builder in Copilot turns one prompt into a working agent. The person who knows the process best builds it, no code required. See Agent Builder in action. https://t.co/zo2QVETfHb Build a specialist AI agent without writing a line of code. Describe what you want in one prompt, and Copilot Agent Builder generates the name, instructions, skills, and knowledge setup for you โ grounded only in the SharePoint sites, files, and Copilot Connectors you choose. #CopilotAgentBuilder #aiagents #copilot #microsoft365 #microsoft
Playwright MCP or Playwright CLI: pick either, your agent gets all of this ๐ญ
๐งฐ WebMCP: pages hand the agent their own tools
๐ Find: grep the page instead of reading all of it
๐ชถ Leaner snapshots, fewer tokens on every step
๐ด Do it by hand once, get Playwright code back
๐ฃ๏ธ Codegen in TypeScript, Python, Java and C#
๐ Your real Chrome, logins included
๐ฌ 60 fps video with an animated cursor and chapters
๐ฑ Mobile and device emulation
๐ Light/dark flipped mid-session
๐ Mock any request, or pull the plug and go offline
๐ธ Hi-res screenshots
๐ฉน Crashed tab or dropped connection? It'll recover
npx @playwright/mcp@latest
https://t.co/52TE8sSErw
npm i -g @playwright/cli@latest
https://t.co/21UHzfLJN0
AI Guardrails vs AI Governance
๐๐ ๐ด๐๐ฎ๐ฟ๐ฑ๐ฟ๐ฎ๐ถ๐น๐ are built to control what an AI system can do. They can filter risky prompts, block sensitive data, restrict what information can be retrieved, or limit which tools and APIs an agent can use. That makes them useful for enforcing specific technical boundaries inside the system.
But guardrails only solve part of the problem.
With ๐๐ ๐ด๐ผ๐๐ฒ๐ฟ๐ป๐ฎ๐ป๐ฐ๐ฒ, the focus is broader. It defines how AI should be managed across the organization: what risks are acceptable, who owns them, what standards teams should follow, and how systems are reviewed over time.
So governance defines what should be allowed and who is accountable. Guardrails help enforce what the AI can actually do.
But for coding agents, that creates another challenge: how do you make those engineering standards available while the agent is actually writing code?
๐๐ณ ๐๐ต๐ฒ๐ ๐ผ๐ป๐น๐ ๐๐๐ฟ๐ณ๐ฎ๐ฐ๐ฒ ๐ฎ๐ ๐ฟ๐ฒ๐๐ถ๐ฒ๐ ๐๐ถ๐บ๐ฒ, the code has already been written.
Thatโs where Qodoโs Agentic Toolbox comes in. It lets coding agents pull in team rules and codebase context, then bring in independent review while they work, ๐ฏ๐ฒ๐ณ๐ผ๐ฟ๐ฒ ๐๐ต๐ฒ ๐ฃ๐ฅ becomes the first real checkpoint.
So the same agent writing the code isn't the only agent checking it.
Install the Toolbox โ https://t.co/8qH3tNroGK
What else would you add?
โโ
โป๏ธ Repost to help others learn AI.
๐ Thanks to @QodoAI for sponsoring this post.
โ Follow me ( Nikki Siapno ) to improve at AI and system design.
I don't get how secondary market ticket price gouging is legal. There should be a limit over the price of the ticket, and they should have to be sold by individuals, not companies who use bots to gobble them up...
@mandyarthur@KnoxieLuv@grok Does it matter? They're going to get away with it anyway and there will be no consequences. Why aren't Democrats following on his office claims. Surely someone could find out if he's working or not??? Image one of us AWOL from work, nobody knows if you're coming back. Fired day 2