· Human Skills & Development  · 6 min read

Turkeys and Turning Points (Or: How Grandma Got Won Over by an Agent)

A Thanksgiving survival guide for explaining AI Agents and Microsoft Ignite takeaways to your family - moving the conversation from chatbots to the "sous chefs" of the future and beyond.

A Thanksgiving survival guide for explaining AI Agents and Microsoft Ignite takeaways to your family - moving the conversation from chatbots to the "sous chefs" of the future and beyond.
Loading the Elevenlabs Text to Speech AudioNative Player…

Warning: this food for thought contains a much higher than average density of dad puns, mostly thought up while visiting the state of California.

It is a truth universally acknowledged that a tech professional in possession of a job must be in want of a stiff drink before Thanksgiving dinner.

We all know the drill. You’re barely in the door before the questions start. Usually, it’s “Why is my printer offline?” (I don’t know, Uncle Bob. I really don’t.) But this year, the vibe might be different. The questions aren’t about broken hardware. They’re about that thing everyone is talking about on the news…

  • Grandma wants to know if the “Chat-bot” is writing her emails now.

  • Your college-aged nephew is rolling his eyes because he’s “vibe coding” and thinks he’s already replaced your entire engineering team.

Here’s the thing: They’re both pretty close to some big ideas, but just a little bit off in their own ways. And because I want you to actually enjoy your drumstick this year, I’ve drafted you some talking points.

I just got back from Microsoft Ignite last week, so I’m stuffing this post full of fresh baked ideas and news, translated for two very different audiences.

The Grandma Test: Why AI is Your New Sous Chef

Grandma is smart. She’s seen technologies come and go. She doesn’t care about “Large Language Models” or “Tokens.” She cares about utility.

When she asks, “What is this AI Agent thing?”, do not start talking about neural networks. Frame your explanation with the holiday itself, starting with the cooking.

Explain it like this:

The Old Way (Google/Search): This is like a cookbook. You have to go to the shelf, find the book, look up the recipe, and then you have to do all the work. You’re chopping the onions. You’re watching the stove.

The “ChatGPT” Way (Generative AI): This is like having a really smart chef standing next to you. You ask, “How do I make a turkey?” and they recite the perfect recipe instantly. They can even write you a poem about the turkey. But… you’re still chopping the onions. (And let’s be honest, you’re probably crying while doing it.)

The New Way (Agents): This is what we are building now. This is the Sous Chef.

An Agent doesn’t just read you the recipe. You say, “Make the stuffing,” and the Agent goes into the cupboard, checks if we have sage, orders it if we don’t, chops the celery, and puts it in the oven. Or maybe just one of those steps, whichever one Grandma hates doing the most. The best opportunities start with small potatoes.

Tell her: “Grandma, for the last two years, we’ve been teaching computers how to talk. This next year? We are teaching them how to do.”

She’ll get it. She’s been managing a kitchen “team” of unruly relatives for 40 years. She understands orchestration better than most CIOs. Mostly because she can’t very well put Auntie Jennifer on a PiP no matter how much she’s always wanted to.

The Nephew Test: Vibe Checks and Frontier Firms

Your nephew is harder to impress. He’s seen the Matrix. He uses vibe coding to build apps by just telling an AI what he wants, and he’s happy with what he gets back.

Don’t burst his bubble - expand it. Tell him he’s actually on to something, but he’s only seeing part of the bigger picture.

Like the mashed potatoes you nearly forgot to pull back out of the warming oven, take this next section with a couple grains of salt:

  • This is a very short and optimistic summary. I can and will write whole essays soon on how much more complicated all of this is in practice.
  • I view ALL of these tools as optimal to improve the work and quality of life of the humans who are already doing the work in question. Even if you could quantifiably save $5 replacing a person with AI, I’m firmly of the opinion that you’d get double that (at minimum) by instead equipping that same person with better tools instead.

Explain it like this:

Microsoft just validated some of his favorite tools at Ignite. (This won’t surprise him, since he’s never wrong anyway.) They literally used the term “Vibe Working” on stage to describe the new Agent Mode in Excel and Word. It’s a lot like what he already does: he sets the intent (the vibe), and the AI handles the rote parts of the work. But the “get rich” play isn’t building one killer vibe product - that’s playing the lottery like it’s always been since the days of indie iPhone app development. The biggest idea in the room is building a Frontier Firm.

This is Microsoft’s vision of a new corporate structure where he isn’t a coder; he’s the captain of a Human-Agent Hybrid Team.

Take a look at Agent 365. It’s the new command center where he could manage a fleet of autonomous digital interns. These aren’t chatbots; they are employees that live in Teams and elsewhere, talking to each other (like a GitHub agent debating a Jira agent on process) to solve problems while he sleeps. His smallest Agent might retrieve and synthesize platform patch notes while his largest handles complex tasks around writing code and routinely reviewing its own custom documentation.

Tell him: “You’ve got the right idea that the future is going to involve a lot more curation and management of a sort. But don’t neglect your basics… It’ll still be on you if the AI makes a colossal error. You can’t curate quality if you don’t know what it looks like yourself.”

The Year Ahead

If there is one takeaway from the whirlwind of announcements at Ignite, it’s this: The honeymoon phase of “chatting with bots” is over. They’re still helpful and they’re not going anywhere! But they’re no longer the whole story.

We are moving into the era of Doing.

For Grandma, that means technology that helps her do things, not just answer helpful questions.

For the Nephew, it might mean growing up from a “Vibe Coder” into an “AI Architect”, or at least someone who’s native to AI use just like Word or even email.

And for my team? It means we have to build the guardrails, the culture, and the trust to let these Agents actually work.

I’ll be diving into way more of the Copilot and Agent systems from Ignite as we put them into practice in the coming weeks and months.

The technology is at the starting line. The collective question is: Are we?

I think we are. But ask me again after a slice of pie…

Enjoying the read?

I send a short email once a month — behind-the-scenes notes, honest takes, and first word on workshops. No spam, no fluff.

Back to Blog

Related Posts

View All Posts »
The Secret Life of Prompt Personas
Narrate

The Secret Life of Prompt Personas

Discover how using personas in your AI prompts can dramatically improve context and the quality of your results. Learn why this simple shift in communication changes the conversation and unlocks deeper, more creative collaboration with tools like Copilot.

Aim for the Pain, Not the Pride
Narrate

Aim for the Pain, Not the Pride

Two counterintuitive moves that drove 97% Copilot adoption - recruiting your loudest skeptics and replacing your training program with a five-minute game.

The Frontier of Cowork with Copilot & Claude
Narrate

The Frontier of Cowork with Copilot & Claude

Microsoft built their most ambitious agentic feature on Anthropic's technology. The 'who copied who' debate misses the point. The real challenge is helping thousands of professionals shift from asking AI questions to directing AI work.

Lab #3 - Clean Up Messy Data with 'Edit in Copilot'
Activate

Lab #3 - Clean Up Messy Data with 'Edit in Copilot'

The third hands-on lab in my community resource series gives you a deliberately wrecked spreadsheet and walks you through fixing it with Edit with Copilot - one problem at a time, then all at once. Includes a practice file you can download and break yourself.