I just read that Lloyds Bank is hiring almost 300 AI specialists. That is another huge signal, and it's a bigger one than most people will register when they scroll past the headline. AI is no longer a side project. Big companies are moving from "let's test it" to "let's rebuild the entire workflow." That is the shift. It's not demos anymore. It's execution. And the winners coming out of this next phase will not only use AI. They will reorganize around it.
Let me explain why this particular headline stopped me, because the number is not the point. The number is the symptom.
Lloyds is a 261-year-old bank. That's not a startup chasing a trend. That's one of the most established, most regulated, most conservative institutions you can imagine, and they're not dipping a toe in. They're recruiting close to 300 roles tied specifically to agentic AI, with more than 1,000 AI-related roles planned this year, on top of the 700-plus people they already have working on it. When an institution that old and that careful moves that decisively, it's telling you something about where the whole field is.
Here's what it's telling you. The experimentation phase is over.
For the last two or three years, AI inside most big companies looked like a science project. A small team off in the corner, running pilots, testing tools, writing reports about what might be possible someday. It was a side project. It had a budget line that could get cut in a tough quarter and nobody would feel it. That era is ending, and Lloyds is the proof. You don't hire 300 specialists and reorganize teams for a science project. You do that when you've decided to rebuild how the company actually works.
And that's the phrase I want to sit on. Rebuild how the company works. Because there are two completely different ways a company can approach AI, and the gap between them is about to decide who wins.
The first way is to bolt AI onto the existing workflow. Keep everything the way it is, and sprinkle some AI on top. Add a chatbot here, an AI feature there, a copilot button in the corner of the screen. The org chart doesn't change. The processes don't change. The way work flows through the company doesn't change. You just add AI as a layer on top of the old machine. This is what most companies are doing, and it produces almost nothing, because the old machine was built around human limitations that AI was supposed to remove.
The second way is to rebuild the workflow around what AI makes possible. That's harder, it's slower, and it's what actually works. It means looking at a process that exists because humans used to do every step, and asking what that process should look like now that they don't have to. It means changing the org chart, not just the software. It means the kind of decision Lloyds is making, where you go out and hire the people whose entire job is to redesign how the work gets done.
This is the difference between using AI and reorganizing around it. Using AI is a feature. Reorganizing around AI is a strategy. And the companies that treat it as a feature are going to lose to the companies that treat it as a strategy, because you cannot out-execute a competitor who rebuilt their entire operation while you were adding a button.
The pattern shows up everywhere once you start looking for it. Lloyds itself isn't being quiet about the implications. The bank's own head of data and AI science said plainly that this will transform organisational structures and the nature of many jobs. That's not a vendor making a pitch. That's the institution telling its own people that the shape of the company is going to change. They're not promising AI will leave everything the same. They're promising it won't.
I respect that honesty, because it's the part most leaders are afraid to say out loud. Reorganizing around AI means the org chart you have today is not the org chart you'll have in three years. Some roles change. Some disappear. New ones get created that didn't exist before, like the 300 Lloyds is hiring for right now. Pretending otherwise is how companies sleepwalk into being disrupted by a competitor who told the truth earlier.
So when I say the winners will reorganize around AI, here's what I actually mean in practice. They will not ask "how do we add AI to what we do." They'll ask "what would we build if we were starting today, knowing what AI can do." Those are completely different questions, and they lead to completely different companies. The first question gets you a slightly faster version of what you already had. The second question gets you something that makes what you already had obsolete.
I think about this constantly from where I sit, because it's the bet the whole company is built on. We're not trying to add AI to the existing way real estate transactions get done. That would be the bolt-it-on version, and it wouldn't move anything. We're trying to rebuild the workflow itself around what AI now makes possible, so the work that used to take hours and three different tools and a laptop under a car seat becomes something else entirely. That's a harder thing to build than a feature. It's also the only version that matters.
The winners won't only use AI, because using AI is going to be table stakes the same way using the internet became table stakes. Everybody will use it. The differentiation will come from the companies brave enough to reorganize around it, to change the structure and not just the software, to ask the harder question and live with the harder answer. A 261-year-old bank just decided to be one of those companies. The signal is not that they're hiring. The signal is that the experimentation phase is over, and the rebuild has started, and the companies still treating AI as a side project are already behind.
It's not demos anymore. It's execution. Watch who's reorganizing, not who's announcing. That's where the next decade gets decided.
*Judd Hoffman is CEO and Co-Founder of Ethica AI, building AI-powered tools for real estate transaction workflows.*