The Dangerous Illusion Of “Doing AI”

Everyone is currently obsessed with how fast AI can complete tasks.

Write blogs in seconds.
 Create marketing instantly.
 Qualify leads automatically.
 Generate content at breakneck speed.

And yes… it’s genuinely mind-blowing.

But faster tasks are not the same thing as a transformed business.

In this episode, Mark Burgess breaks down the critical difference between using AI for productivity and building an actual AI operating system inside an estate agency.

Using the analogy of putting a McLaren engine inside a 1970s car, this episode explores why most agencies are accelerating old operating models instead of redesigning how the business fundamentally works.

Topics include:

•  why AI task automation alone won’t massively improve RPE 
•  the hidden bottlenecks that still require manpower 
•  why disconnected AI tools create chaos instead of leverage 
•  the importance of centralised intelligence 
•  why the future belongs to system-led agencies, not faster manual agencies 
•  the uncomfortable reality of AI adoption inside teams 

This isn’t an episode about prompts or software tricks.

It’s about the structural divide quietly forming inside estate agency.

Transcript

SPEAKER_00 0:10

INMR'T Impact on Business. For the past few years, we've been building INR systems first agencies of transforming our own company into an INR first business. This podcast documents what actually happened during that transition. The resistance, the mistakes, and the breakthroughs that change how businesses work. Welcome to this episode of my AI operating system diaries. This week I want to talk to you about something I was going through with a client that just triggered me into thinking about this episode. So I do some mentoring with some clients one-to-one, and I was working with a client, and we were talking about the core of their company. Everything should stem off of the back of this, their valuation pitch, everything. We seemed to hit a little bit of a roadblock with that, um, whereby they really needed some tangible tasks, and I really felt like I was giving them tangible tasks. We've gone back and forward over this for a few sessions, and so I decided to sit down with them and do some practical work on creating that stuff, how I would use AI to help me create some of that stuff, and we had a pretty mind-blowing session for them where they actually watched me create all this stuff in minutes using combination of different AI tools, you know, use GPT for certain things, use Claude for certain things, and um they they at the end of the day they were saying like they really feel like they've had a massive breakthrough on this, they couldn't see it before, they couldn't see what I was talking about, and that's great. Um but I mean there was a it was a great session, it was great for them. There was no problems with it whatsoever, but it just made me think about this podcast and what else is going on out there in the world for some of you guys listening, and I wanted to talk to you about how this isn't the AI that's going to really make or break your business. Some of that stuff is so cool to see. AI speeding up tasks is genuinely mind-blowing, it is totally magical the what you can do achieve now in minutes, and but everyone who's using AI is experiencing this right now, writing brand guidelines in minutes instead of months, you know, it the whole thing is magical, but the point I want to make is that faster tasks do not equal a completely transformed business. So the analogy I think that kind of puts this into perspective is imagine if you had like a 1970s board escort, or let's go even further back than that, you know, like the the Mark II, that sort of thing, pretty cool car now, but the infrastructure of the car is all still 1960s, 1970s technology. And I want you to imagine that somehow you put the engine of a McLaren in that 1970s S Court, you don't upgrade anything else, you just somehow get that McLaren engine connected up. What would happen? Like, when you first use the car, it's gonna feel pretty incredible, like the acceleration that you could where you could press the accelerator down just a little bit, and that the car would just move at a rate that would just seem insane if you had only ever been used to driving that 1960s, 70s Ford Escort, massive acceleration, and suddenly like you'd feel like you were obsessed with this kind of speed kick that you could get out of it, it'd be crazy compared with all the other cars on the road, but the structure of the car wasn't built for that speed, like the brakes will not work, the whole chassis of the car is gonna shake and start falling apart, so the bottlenecks still remain inside the car, right? Eventually, the whole thing is gonna strain under the pressure and just collapse. And most businesses are putting McLaren engines inside their 1960s stroke 70 operating models at the moment. It's just it's not becoming apparent just yet. So using AI for you know things like writing blogs, writing property descriptions, even qualifying tenants, summarizing meetings, creating marketing, replying to emails, it's all all all useful stuff. Like I'm not saying it isn't, you it's impossible not to see the benefits of it. But remember, your humans are still orchestrating everything. The AI is just doing the task far faster. Humans are still moving the information from one place to the next place. Humans are still triggering workflows, humans are still deciding what happens next, what takes priority, what work is being done. So if they haven't got time to do the blog, then it won't get done. If they haven't got time to call these people, it won't get done. Even though AI might have started it, it doesn't finish jobs at the moment. So the tasks changed a little bit, but the operating model underneath hasn't really shifted, and this is what I see as the future of companies that just accelerate away into the distance over the course of the next couple of years. We're talking about an AI operating system, so it's not how do we do tasks faster? Why are humans doing these tasks at all? And again, I still think I still feel like it's I'm not explaining it fully to people whereby they're not fully grasping it, because it's not about like I get I get contacted by people saying, like, I need I've seen this piece of AI kit that will qualify my tenants over the phone, amazing, it's gonna save us so much time, and that there's no doubt about that. There is absolutely no doubt that it would save you so much time, but it is no different to putting the McLaren engine into the 1960s escort, that is going to blow up your company. Like you haven't changed the model. If you if you go all the way back to episode one of these podcasts, I spoke about the fact that when uh electricity first started to replace like steam engines inside factories, you would think that the productivity would just go through the roof. Everything could be done so much faster and more efficiently. But productivity did not change for another 40 years until they decided to start to look at how the factory was set up and how the operating system of that factory worked, because just having electricity produce things faster and more efficiently was great, but it still had to go through all of the same steps. The factory hadn't changed at all. Think about the blockbuster Netflix analogy, like it just it doesn't work the same. It's not that blockbuster carried on, it's that the whole thing got blown to bits. They did actually try to create online streaming blockbuster. Like most people think they didn't even try, they did, but they they didn't change the whole operating model, it was a it was a bonus to the to the stores and the videos and the DVDs, Netflix just blew it apart. So operating system thinking kind of works like this. You must have centralized intelligence. I need you to really think about that part. You must have centralized intelligence. No one has this right now. No one. No one right about your valuations that took place, not in the last couple of days, but in the last couple of months. Are you one hundred percent confident that all the notes are up to date, all communication with the people over the past few months has been logged, and it is all one hundred percent up to date, and the answer to that has to be no, it has to be because your systems are not connected. Not only are is all the internal notes not up to date because you use things like a random WhatsApp, you use a random email system, you have phone calls with people, and then you log notes into the CRM. So not possible for all that communication to be up to date and centralized. You you then do not, well, most of you possibly do, but there's a lot of people out there that do not have a centralized tracking system, tracking your customers' behavior all across your website, all across your blogs, all across your email marketing, all across your instant valuations, all across your market appraisals, all across everything. I want to know who's doing what and when they're doing it, not just when they click a link. I'm talking about tracking them across the whole thing. Once you have that, right, you start taking away the randomness, and you start being able to do all the things that you need to do to have this operating system thinking. Centralized intelligence, it knows all conversations that have taken place, it knows all the emails that have been sent, all the replies that have come in, it knows all the WhatsApp conversations that have taken place, it knows all the telephone conversations that have taken place, it knows all of that, it centralizes it. The AI, we're not we're not unfamiliar with the fact that AI transcribes that stuff and summarizes it. So that task that we once were using starts to stack up into intelligence. Take all of those conversations and put the notes in the system. Have a phone call with somebody, put the phone down, move to WhatsApp, have a WhatsApp with them, move to email, have an email with them, never need to log any notes. It's all in the centralized intelligence, and then track their behavior across your website, across your market appraisal presentations, across your blogs, across your emails, so that AI can then put all that together and prioritize the opportunities, create autonomous workflows, have system-led decision making, orchestrated nurturing. So that AI is now acting with context. That's the critical difference here. Isolated AI equals blind AI. Without connected systems, AI can't see the intent, it cannot prioritize correctly, it cannot act intelligently, it would just be performing random tasks like I did in the in the workshop, not the workshop in the mentoring session. It was magical what I did. Putting all that information into a GPT, building a mu building a brand document was in phenomenal. But not nothing will happen unless the brand document actually gets used for something. So AI without centralized intelligence is just guessing. Most agencies or businesses have got a CRM over here, email platform over there, email marketing system somewhere else, website is with another company, content marketing, their blogs and all that stuff is being put somewhere else. Social media is disconnected, lead tracking is disconnected, lead qualification is disconnected. As I said before, someone the other day was saying to me, I want to use this AI tool that's going to qualify my tenants. Just another disconnected tool. I can see the benefit of it, I can see why you want to do it for the short-term impact. But the long-term impact is putting the McLaren engine into an escort. This isn't going to get you anywhere. Everyone's going to get it. Traditional estate agency. Someone did a valuation 18 months ago. And yeah, they've, you know, they've got a marketing system that's been sending nurture emails. Maybe someone clicked something once upon a time. Right? Yeah, great. But nobody knows what they're reading right now, what they're thinking about, whether their intent is increasing, because people don't have the behavioral tracking system across their entire ecosystem. An AI operating system uh estate agency knows they had evaluation previously. It knows that they have just opened some nurture emails and that they visited the website yesterday reading how to prepare your house for sale. Now AI understands the intent, AI can score the behavior, AI can prioritize the opportunity, and AI can react. So the magic is not in the email. The magic is the system understanding what the behavior means. Yes, you can get AI to perform tasks, and yes, it absolutely saves time. But the real question is what happens after that? Where's the next bottleneck going to hit the company? Is doing it in reverse is gonna be almost impossible. It's what still depends on manpower, what still relies on humans remembering things, saving time in one department doesn't transform the entire operating model of the company. Completing these tasks fast on their own won't massively increase revenue per employee. It will look like it does, but you will keep running into walls. You must create structural leverage. The real gains will happen when the system absorbs the operational workload. That's the bit that the AI is going to be phenomenal in transforming our in our industry. We have to stay focused on that. It's a bit boring, it's a bit more boring than AI will take some of my phone calls away from me. That's quite cool, but it won't create the same structural shift as if we get this operating system working. And the key to getting the operating system is centralizing the intelligence. To centralize the intelligence, we must centralize the communication. Once we've done that, then things are gonna start to get real interesting. So people just tend to celebrate visible speed, flashy AI moments, and call demos, and I get it, you know, I'm no different to that. But people are undervaluing invisible system leverage, automation infrastructure, operational intelligence. So I just want people to stop thinking about features so much and how how the feature creates this centralized intelligence is the real thinking that you need to have around this. I've been down the other road. I've got obsessed with getting my team to use this tool, this tool to help you do this faster, and this tool to help you do that faster, this one to help you code faster, this one will help us design faster, this one will help us do this faster. And yeah, it does, and then everything will blow up in your face, and then you will get really frustrated with the team because there'll still be a bottleneck somewhere, something will still be going slow, and you won't be able to get your head around it. You're angry with people. AI makes you super intolerant with humans because you're used to getting things done in seconds, and you have to get past that. I had to get past that. So, what is the operating system that we need to be working here? A lot of people, as I as I've said, are just thinking about the feature. Can it also do this? I saw another AI tool that does X, but isolated features are easy, but connected infrastructure is the hard bit. A lot of people I've spoken to about this have just said, oh yeah, but that that would just be utopia, wouldn't it? And then, like, it's not possible, it is possible. We're building it, we built it. We've built it now. So random AI tools create fragmented chaos. Infrastructure first will equal long-term leverage. Long-term leverage. That's what we need. If you're running a business, that's what you want. Long-term success. And a five-minute thing that seems magical today, but next week everybody's got it, and we're back where we started. You know that's gonna happen. We're not building isolated tricks, we're building an intelligence layer for the business. That is what you need to keep thinking about with this new world. An intelligence layer for the business. Not it's gonna do it far, it's gonna do this faster for me, though. AI is powerful, transformational, incredible. But it's not magic. It's magical, it's just not magic, it's not infinite. AI is only as intelligent as the context it can access. If you know your data isn't up to date, then it's going to be shit. And if you know it isn't connected to everything else, it's going to be shit in the end. And then we come across to the staff resistance. We've covered it many times in this podcast. Teams resist the change. Old habits feel more comfortable. But that friction is just unavoidable. The discomfort isn't proof the system is wrong. It's proof that the model is changing. Think people were comfortable when they changed around the factories? No way. In fact, they probably just needed all new workers to get their head around it. That blockbuster thing that I spoke about. They tried to improve their stores. They tried to go online. Netflix just blew the model to pieces. Taxi firms improved how many they could have dispatching. Uber just broke the system to pieces and changed it. Retail, they fixed up their stores, they tried to add a bit of online. Amazon just blew the infrastructure to bits. The winners never improved the old model, they replace it completely. So AI tasks will become normal, and everyone will have AI tools. And the speed will just become commoditized. That is normal. It's already normal to a lot of people. The real divide will be agencies with system intelligence. Agencies still dependent on manpower. Those two things will be the real split. So the future isn't businesses that work the same way faster, the future is businesses that operate completely differently. Don't forget that.