Something shifted this year and you can see it in the reactions. Not to the technology. To people talking about it. Rob shared a screenshot on LinkedIn. CFO. Friday night. Using CoWork in real time. The kind of moment where you have to stop yourself because you won't sleep otherwise. And that's what set someone off. Not hype. Not a prediction. Just… "this is happening." Apparently that's enough now.

Rob calls it the knowledge cliff. AI knows three things. What's in the training. What it can pull from the web. And everything that only exists in your world. The first two feel almost the same. The third is where things break. That's where most of the frustration lives. If you haven't crossed that line yet, AI feels inconsistent. Impressive one minute, useless the next. If you have, it starts to look a lot more like real work getting done.

You can see it in companies already changing how they plan and operate. You can see it in schools trying to figure out how to respond. And you can definitely see it in the comments, where people react to the exact same example like they're living in two different worlds. You can't really be smug about it. But the people who've crossed the cliff aren't waiting for consensus. They weren't a year ago either.

This episode won't tell you what to think about AI but it will make it a lot harder to ignore what's already happening.

The job hunt is a numbers game. The problem is, the numbers are brutal.

Hundreds of applicants per role. Ghosted applications. "Entry level" jobs asking for experience no one at 22 could possibly have. In this episode, Rob brings on his daughter Ella, a college senior in the middle of it, and hands her something different. Not advice. Not a better resume template. A coworker that doesn't get tired, doesn't lose track, and doesn't stop digging.

Within 48 hours, she's using Claude Cowork to search across sources, filter for real roles, verify listings, organize everything into a system, and adjust the criteria on the fly when the market doesn't cooperate. It's messy. It's imperfect. And it's wildly more effective than doing it alone. Watching it happen in real time makes one thing pretty obvious. This isn't about AI helping you think. It's about AI helping you work.

One person scrolling and hoping. One person running a system that never stops. Listen to this episode to decide which side of that you want to be on.

The work feels different now.

You can hear it in this one. Something that used to feel like overhead suddenly starts pulling its weight. Not a demo. Not something you have to babysit. It's actually doing useful work while you're still figuring out what you want. That's a weird moment the first time you see it. And then it stops being weird and just becomes the new normal.

It shows up in a few places here. Cowork starts earning its keep. The "data gene" gets reworked into something that fits where things are going. And there's a moment that might make you a little uncomfortable if you've spent years leaning on tools like WordPress to get things out the door. Because the gap those tools were filling is getting smaller. Fast. The people who like to build and adjust as they go feel that immediately. They don't want to wait around for results. Now they don't have to.

And then there's the other camp. The folks who checked this out once, decided it wasn't that impressive, and moved on. Still pretty confident the whole thing is overblown. You can feel that tension in this episode. And it matters. Because a year ago this would've sounded like a stretch. It doesn't anymore.

Most AI still lives in the "that's pretty cool" category. It answers questions, writes a decent paragraph, maybe even points you in the right direction. And then you still have to go do the work.

That line is starting to move. Not in theory. In real, hands on, open the file and keep going kind of ways. We're talking about outputs that don't fall apart the second you touch them. Work that shows up structured, editable, and worth building on. That's a very different experience than what most people think of when they hear "AI."

Some of this stuff still feels like a demo. You try it, you nod, and then you go back to doing things the old way. Other parts are starting to feel different. You give it something real and it gives you something back you can use without starting over. That's the shift. And once you see it, it's hard to unsee.

Listen to the episode and decide where AI in your business is still a demo and where it's finally ready to pull its weight.

Most people think they've already experienced AI. They've asked a chatbot a question, had it summarize something, maybe even draft an email. That version is useful, but it isn't the one that actually changes how work gets done. The real shift starts when AI stops talking about work and starts participating in it. That's the moment Rob ran into while experimenting with Cowork tools, and it was convincing enough to push him into changes he hasn't made since the DOS era.

Microsoft just announced Copilot Cowork, and Rob thinks it could turn out to be the most significant AI product Microsoft has shipped so far. Not because of a flashy feature list, but because of where it lives. When something like this can operate across the Microsoft 365 environment where work already happens, it suddenly has real context. Files in OneDrive. Documents in SharePoint. Conversations in Teams. Meetings in Outlook. At that point the tool isn't sitting off to the side anymore. It's working inside the same ecosystem your team already runs on.

Most of the working world is still standing on the quiet side of an inflection point they don't fully see yet. Once tools like this start showing up inside the systems companies already use every day, things will move quickly. In this episode Rob and Justin unpack why this moment matters, why Copilot Cowork could change how people experience AI at work, and what it means for the people and organizations paying attention right now. If that includes you, this is the one to listen to.

Every once in a while a new tool shows up that bends the career curve for a certain kind of person. Not everyone. Just the people with that itch to poke at systems until they finally give up their secrets. The same instinct that used to turn someone into the unofficial Excel wizard in the office is now colliding with AI development tools that can help you build real software. If you have the data gene, this moment feels a little like someone just handed you a much bigger toolbox.

It has a lot in common with what happened when Power BI first showed up. For years the people who understood the business problems best were stuck with tools that could only go so far. Power BI suddenly bridged that gap. Now AI assisted development is doing something similar across the rest of the tech stack. The distance between I know what the answer should be and I can build the thing that proves it is shrinking fast.

Of course, building something is not the same thing as building a company. Rob and Justin get into that too. AI can help you spin up software faster than ever, but the hard parts of business still live somewhere else. Vision. Distribution. Understanding the real problem well enough to solve it in a way that people care about. The tools are getting easier. The thinking still matters.

Also in this Episode:
The Lion King Lyrics Revealed

Rob and Justin had a plan. Scale Justin's brain across the entire P3 consulting team. Build an AI agent that bottled up his frameworks, his instincts, the way he navigates AI conversations with clients. In theory, everyone gets smarter overnight. It was a solid idea. The tech worked. The knowledge base was deep. The guardrails were tight. And almost nobody used it. Not because it was broken. Because the team wasn't waking up thinking, "Man, if only I could channel Justin right now." That wasn't the fire in front of them. So instead of feeling like leverage, the agent felt like homework.

And that's the punchline. You can build something powerful and still miss the mark. No one was losing sleep over not having this tool. No one's bonus depended on it. So it drifted. Not rejected. Just... optional. That's a brutal place for a "strategic initiative" to land. The fix isn't a better tool. It's sequencing. Define the services, train the team, build the human infrastructure that makes the tool land on a surface that's ready for it. Every AI project that has worked traces back to the builder being a direct stakeholder. Not adjacent to the problem. In it. Proximity to the pain is doing a lot of work that no amount of clever architecture can replace.

When leaders are the ones excited about AI and employees are the ones expected to use it, you've got a stakeholder mismatch. And that mismatch is quietly killing more AI initiatives than any technical failure ever will. If you're planning a rollout, or already wondering why yours isn't sticking, this episode is for you. Be sure to subscribe on your favorite podcast platform for new content delivered directly to your inbox.

There's an easy button for hard conversations now, and it's dangerously good. You've got something complicated to say. It needs nuance. It needs empathy. It probably needs a little courage. The AI will draft the whole thing in seconds. It sounds smart. It sounds reasonable. You skim it. You send it. And most of the time, nothing bad happens. The problem is that the time it does go bad is the exact situation where you thought you were being thoughtful. This week's Raw Data walks straight through one of those moments, from both sides of the exchange, and it's a reminder that outsourcing the structure of your thinking is not the same thing as being clear.

Then there's the part that's almost more interesting. Thirteen years ago, the first real client engagement couldn't get traction around dashboards. The connection between "this is my business" and "data should change how I run it" just didn't stick. Same people, same company, different conversation recently around AI. Immediate traction. Leaning forward. Connecting dots in real time. That difference isn't about better slides or better storytelling. Dashboards improved a slice of the business. AI shows up in the messy motion of the whole thing. In workflows. In manual processes. In strategic questions leaders don't have time to chase down. That shift in surface area changes everything.

AI isn't a toy and it isn't a ghostwriter. It's leverage. Real leverage. The kind that can remove friction across an organization faster than dashboards ever could. But leverage only works if you're still the one steering. That's really what this episode comes down to. Listen in, then decide where AI belongs in your workflow and where it needs to stay out of your head.