This was the week AI stopped being a separate app you open and became a worker that lives inside the tools you already use — Microsoft Office, your email, your file storage — and got noticeably cheaper to run. For a business doing between $1M and $20M, that combination is exactly when “interesting technology” turns into “this could actually save me payroll.” Here are the five questions worth asking this week, with a few aimed squarely at specific industries.
- Now that AI agents live inside Microsoft Office and my email, what can they actually do for my front office?
- Why should HVAC, plumbing, and other home-services owners care about the new “connector” agents?
- AI just got cheaper this week — does that change whether I should adopt it now?
- Could using AI to write my marketing put my business at legal risk?
- What does AI video generation mean for restaurants, retail, and real estate listings?
1. Now that AI agents live inside Microsoft Office and my email, what can they actually do for my front office?
The biggest shift this week is location. On May 29, Perplexity rolled its “Computer” agent directly into Microsoft 365 — Word, Excel, PowerPoint, Outlook, and Teams — and OpenAI’s Codex gained “Computer Use” on Windows, meaning it can see, click, and type inside Windows apps the way a person would. In plain terms: the AI no longer just answers questions in a chat window. It can sit inside the software your office already runs and do the clicking.
For a $1M–$20M business, the practical wins are unglamorous but real: drafting and cleaning up Outlook replies, building a quarterly Excel summary from messy data, turning a one-page note into a client-ready PowerPoint, or reconciling two spreadsheets that never quite match. These are the tasks that quietly eat your office manager’s afternoon. Because the agent works inside Microsoft 365 rather than as a separate subscription, the friction of “getting your team to use it” drops a lot — it’s in the menu they already stare at all day.
The honest caveat: these agents are powerful but not flawless, so treat them like a sharp new hire on week one. Give them low-stakes, reviewable work first — a draft, not a sent email; a calculated sheet you still eyeball before it goes to your accountant.
2. Why should HVAC, plumbing, and other home-services owners care about the new “connector” agents?
This one is specific, and it matters. On May 29 xAI shipped Grok Build, its fastest coding model, on the back of two quieter releases that are arguably more useful to a service business: Custom Skills (May 26) and Connectors that plug Grok into SharePoint, Outlook, OneDrive, Google Workspace, Notion, GitHub, and Linear. Anthropic’s Claude Code added “dynamic workflows” that orchestrate large numbers of background agents, and Google previewed Gemini Spark, a 24/7 personal agent.
Here’s why home services has the most to gain. An HVAC, plumbing, roofing, or landscaping company runs on a loop of inbound calls, scheduling, dispatch, quotes, and follow-up — and most of that data already lives in email, a shared drive, and a calendar. A connector agent can read across those tools at once: pull the new request out of Outlook, check the calendar, draft the appointment confirmation, and flag the customer who got a quote two weeks ago and never booked. That “never booked” follow-up alone is often the single biggest pile of lost revenue in a service business, and it’s exactly the repetitive, after-hours work a 24/7 agent is built for.
If you run a home-services company, the move this quarter isn’t to rip out your field-service software. It’s to pick one leaky process — most owners already know which one — and test whether a connected agent can cover it. Aim small: missed-call follow-ups, or turning completed jobs into review requests.
3. AI just got cheaper this week — does that change whether I should adopt it now?
Yes, and this is the underrated story. DeepSeek’s new V4 models slashed the compute and memory needed to run them — V4-Pro uses roughly a quarter of the resources of the prior generation at long context — and effective June 1, its pricing drops to about a quarter of list price. Google launched Gemini 3.5 Flash as its strongest coding-and-agent model yet and cut its top Ultra subscription from $250 to $200 a month. Across the board, the same capability that cost a dollar last quarter is trending toward pennies.
For a business owner, this kills one of the last good reasons to wait. A year ago, running AI on every customer email or every invoice could get expensive at volume. The economics are now tilting so that “use it on everything” is becoming affordable rather than a luxury reserved for your highest-value tasks. The competitive math changes too: if AI is cheap enough that your larger competitors are applying it to every lead and every follow-up, the cost of sitting out grows.
The practical takeaway is to stop pricing AI off last year’s numbers. If you ran the math months ago and shelved an idea because the per-use cost didn’t pencil out, this is the week to re-run it.
4. Could using AI to write my marketing put my business at legal risk?
This week added a real cautionary note. On May 28, CNN sued Perplexity over the alleged unlicensed copying of roughly 17,000 works — joining existing suits from The New York Times, Reddit, and Dow Jones. The legal fight is about how AI tools ingest and reuse published content, and it’s becoming a genuine business risk for the companies building these tools, not just an abstract debate.
For a small business, you are almost certainly not the target of these lawsuits. But there’s a sensible lesson in how you use AI for content. If you run a real estate brokerage, a retail brand, or any business that publishes a lot of marketing, be cautious about having AI generate copy that leans heavily on a specific outside source — a competitor’s listing description, a publication’s article, a photographer’s images — and treat anything client-facing as something to review and make your own. The safest pattern is using AI to draft from your information: your listings, your product details, your customer FAQs.
When you choose a vendor, it’s also fair to ask about indemnification — whether the AI company will stand behind you if its tool produces something that creates a copyright problem. Larger providers increasingly offer this; it’s a reasonable question to raise before you standardize on any tool.
5. What does AI video generation mean for restaurants, retail, and real estate listings?
Alongside Gemini 3.5, Google previewed Gemini Omni, a system for multimodal video generation, while Meta launched its first Superintelligence Labs model, Muse Spark. The headline for visual businesses: producing decent marketing video is rapidly getting cheaper and faster, moving from a hire-a-crew project to something closer to a same-day task.
If you run a restaurant, retail shop, or real estate office, video has always been the highest-engagement content and the hardest to produce consistently. A restaurant could turn a few dish photos into a short social clip for the week’s special. A retailer could generate product spotlights without booking a studio. A real estate agent could produce a quick walkthrough-style video to pair with listing photos. None of this replaces a real videographer for your flagship work, but it makes the weekly drumbeat of content — the part most small businesses skip because it’s a hassle — genuinely doable.
The thing to watch is brand consistency and authenticity. AI video is a volume tool for everyday content; your customers can usually tell the difference between a real plate of food and a generated one, so use it where speed matters and keep the human touch where trust does.
What This All Means
This week marked a turn from AI as a clever toy to AI as a coworker embedded in the apps you already pay for — and one that’s getting cheaper by the month. The biggest opportunities right now are unglamorous: front-office cleanup inside Microsoft 365, connected agents that chase the follow-ups your service business keeps dropping, and faster marketing content for visual businesses. The biggest cautions are equally practical: review AI-written content, be mindful of where it sourced from, and ask vendors hard questions about risk. The owners who win won’t be the ones with the fanciest tools — they’ll be the ones who picked one leaky process and let AI plug it.
Not sure how any of this applies to your specific business? That’s exactly what our free audit is designed to answer. In 30 minutes, we’ll map out where AI can realistically save you time, generate leads, or give you visibility you don’t currently have — no jargon, no pressure. Book your free audit →