Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
The $15 Trillion Question: Why Construction Projects Keep Getting Bigger
0 Playsin 4 hours
Join Jason and Carlos as they explore three major developments in construction and technology:
🏗️ Mega Projects Surge: Analysis of Mace's new report showing 11,000+ live megaprojects worth $15 trillion globally - a 4x increase since 2010. Why are projects getting bigger, and what's driving this trend?
🌐 AI Browser Revolution: The emerging battle between Perplexity, OpenAI, and Google for AI-enabled browsers that could automate construction workflows from form-filling to document control.
💰 UK Infrastructure Wave: Major project approvals including Sizewell C nuclear, A66 highway expansion, and the Midlands Rail Hub connecting 50 stations.
Key Timestamps:
00:00 - Intro
02:32 - Mace’s Mega Project Report
19:00 - AI Browser Revolution
32:23 - UK Infrastructure Wave
Check out the Off Site newsletter here
Follow Carlos on Linkedin | Follow Jason on Linkedin | Check out Aphex
Recommended
Transcript
Introduction and Podcast Milestone
00:00:00
Speaker
we just made it so hard to get projects approved that that the projects get bigger so that they only have to get approved once. And then by doing that, we fall into the trap of the iron law of mega projects by making things more and more complex, which that drives the productivity problem.
00:00:15
Speaker
And all the crap around how run you run your six week look ahead is just rubbish.
00:00:28
Speaker
Welcome back to the Offsite Podcast, where we talk all things construction and technology, episode 97, three away from the big 100.
Guest Introductions
00:00:40
Speaker
I'm Jason Lanzini calling in and dialing in from a station building in Germany. I'm Carlos Cavallo. How are you doing, mate?
00:00:50
Speaker
Not too bad, thanks. You are you mentioned 100 episodes. Do you remember we had a conversation ah probably two years ago? i'm already regretting this conversation. The theoretical one that I've never had.
00:01:03
Speaker
The vast majority of podcasts never get past, I think it was 30. So to get to the century, quite a thing. Yeah, I've not updated the stats on how many podcasts make it to 100. Yeah, our five listeners are absolutely buzzing. Yeah.
00:01:18
Speaker
ah So you're in London today. I'm in Germany. I'm doing my best impersonation of the Fat Controller sat literally inside a station a station control building with a track less than five meters from me.
00:01:33
Speaker
Did you look out the window when you chose the office? i I came in the back door and I didn't i didn't use Google Maps very well. So yeah. There was this low rumbling noise. I couldn't work out what it was until I opened the window and saw a train flying past every three and a half minutes.
Overview of Mega Projects
00:01:51
Speaker
Today, we are covering a ah quite an interesting deep dive into Mega and Giga projects ah released by Mace, the consulting and contracting ah company out of the UK. a Super interesting dive into that.
00:02:05
Speaker
The second thing is we're going to look at this cutting edge release of new AI first browsers that probably feels like not the topic that we'd normally cover, but I think it's charting a course for things that you'll see on a construction project near you soon.
00:02:21
Speaker
And then finally, we're going to unpack very briefly at the end, the latest infrastructure spending announcements out of the UK, which includes some major nuclear investments.
00:02:33
Speaker
Should dive off into topic number one? So um the consulting division of MACE recently released a a really well-structured and comprehensive report called the Future of Major Program Delivery, which dives into globally the state of and the breakdown of mega and giga projects around the world.
00:02:56
Speaker
So for some context, a mega project is a project that has a capital value of greater than $1 billion, dollars while a giga project is a ah project that exceeds $10 billion dollars in value. So these are very, very large projects and schemes.
00:03:12
Speaker
um And the depth of report is actually really remarkable. So Mase analyzed over 5,000 projects globally, shows that we now have over 11,000 live projects.
00:03:26
Speaker
and 250 gigaprojects in delivery worldwide. And together that is representing more than 15 trillion in capital expenditure.
00:03:37
Speaker
This is- That's a zeros. It's a lot of zeros. There are a lot of places to dive in here. um In particular of note to me was that the, if you compare the number of these projects since 2010, the number alone has increased approximately three yeah Sorry, it's increased by three. So it's actually fourfold up on the number from 2010.
00:03:59
Speaker
Lots and lots of mega and giga projects. A trend towards more and more
Economic and Political Influences on Projects
00:04:04
Speaker
of them. Is that because we're spending more? ah Is it because we're doing more of these projects in mega, mega projects? Or are things just costing more?
00:04:13
Speaker
Do you think... um When was thinking through this, I wonder how many projects started as mega projects and finished as giga projects. Is that what's skewing the number?
00:04:25
Speaker
Yeah, the client, the you can imagine the client being really happy. Now that we've got all these very great news, all these variations, we're now a mega project. We have a big giga party. Yeah. Yeah. The, the, the numbers are actually quite shocking.
00:04:37
Speaker
Like if we think about, um, the volume of projects here, you've got, uh, what is that? 500, nearly 500 mega projects in the Oz and UK, which is relatively small population.
00:04:51
Speaker
so the, the volume was actually quite shocked by, especially when you look at the 20 plus giga projects in each of those two. Um, it's massive. One thing I was trying to think about is perception of a smaller, and when I say smaller, I'm still talking hundreds of millions or a billion pound project. The perception of a small project and the perception of a giga project, is it, does it incentivize ah organizations or companies to bundle up these massive schemes because it's almost more acceptable for cost overruns and inflating contract values and change and everything out the backend? And is that why we're seeing much more mega schemes than before?
00:05:31
Speaker
Because it's a bit of a license to not just print money, but spend money. i think that there's ah an element of truth to this argument, right? Which is, so yeah, playing back what you're saying is,
00:05:42
Speaker
You can almost make a project too big to cancel yeah and too important to cancel. But I don't know that that is the, I don't know if that's a result or a cause of the trend.
00:05:57
Speaker
It's like having a product and like add more features, the value is getting stronger. that, oh my God, this thing does everything. We need it. Like we can't see that. Well, you think, yeah, it's like, it's like a human nature politics thing that probably is driving this. And so there's probably an element of, I think one element is there's definitely some inflation element. If you look at the growth of these projects from 2010 to now, I think if you do a rough calculation, uh, global, uh,
00:06:23
Speaker
Global inflation over the 15 year probably accounts for something like 50% increase in prices. increase in prices Now, that's ah they're citing a 3 to 4x increase in the number of these projects, so you can't really compare apples with apples there. But basically, in other words, a project that in 2010 would have been about 700-ish million is now going to be in the mega project category.
00:06:49
Speaker
That definitely doesn't account for a majority of this you know three-, four-fold increase. So there's something else at play. The other theory that I would put out there is that, yeah, this is a politics driven approach. And I think that if we think over 15 years of infrastructure procurement has progressed and the number of like legislative, regulatory and other hoops to jump through to get a project greenlit with the number of reports, stakeholder consultation, all of the steps you've got to go through before you can get a ah ah project greenlit.
00:07:21
Speaker
As the bar increases in the number of steps there that you have to go through, you can see how it encourages like joining together what would have been many smaller projects into bigger projects. Yeah, we're going run this process five times, let's do it once.
00:07:34
Speaker
Exactly. So you know we talked last week about the concept of frameworks and how a lot of frameworks these days are starting to include bigger and bigger. ah projects in there. so I think that there's an element of the front end loading of a project becoming more complicated, more stakeholders involved that causes projects to be grouped together yeah and become mega projects that way.
Mace's Report on Mega Projects
00:07:56
Speaker
And then there's probably like a, you know, it's probably some parallels to what we see in the US with like the, that big, beautiful bill, right? Yeah. If we can make this project big enough that gives a lot of a benefit to a number of people along the route of a project, like ah a rail line, suddenly you can get more support for the project.
00:08:16
Speaker
Like a HS2 that went from London to Birmingham is probably at its initial concept way less popular than one that goes all the way to Manchester and Peru. Yeah, yeah, yeah. It's like populist politics.
00:08:28
Speaker
Yeah, so you can see... yeah yeah Yeah, so you can see how the politics probably drives the outcome. Yeah, there's definitely that as a factor. The inflation that you mentioned, I think the UK from 2010 is 64%. So that's a massive increase in cost.
00:08:45
Speaker
And that just won't be um like new contracts going off of value. Obviously, we've got a bunch of magazines that been running for 10 years. And you can imagine that bill just spiraling and spiraling each year. Yeah, on their way to gigaproject status.
00:08:57
Speaker
Yeah, exactly. Trying to hit that milestone. um If we look at where the money is being spent, so there's an interesting chart that sort of maps it in a way that you can see which division which sectors the the spend is in.
Regional Variations in Project Types
00:09:11
Speaker
Yeah, we'll put the I think we should put the link to the report in the show notes because yeah Mace have actually done a really good, big shout out to Mace here. They've done a really solid report and not many people are pulling together this data. So a really useful resource for folks in in the industry.
00:09:29
Speaker
Yeah, for sure. and There's obviously a lot of spending in RESI, which we would expect. But I think the two that we sort of touched on the theme before is obviously energy and defense seem to be quite substantial in terms of that spend alongside infrastructure, which is like, obviously, we know that but that's encouraged um because it's good for the economy and everything else.
00:09:50
Speaker
Yeah, I'd imagine there's a big spike in energy because of if we think about Ukraine crisis in Europe, everyone's trying. It's kind of an accelerator to green energy. So we're less reliant on other nations and importing fossil fuels. So that makes a lot of sense. And a nice win from that is an accelerated clean energy sort of win.
00:10:09
Speaker
um And then defense, like it's all over the news. Trump's trying to get every nation to commit to 5% and they've got to spend that money somewhere um and increase capabilities. So there's a huge amount of money being pumped into defense projects at the moment, which we're seeing ah particularly across the UK and ours, but ah yeah, across Europe. So yeah.
00:10:30
Speaker
Yeah, a lot of cash being spent on these schemes. the What's interesting when you look at the distribution of the same chart that you're talking about on page 12, which is this you know this distribution of the projects by type, ah because this is a count of projects. Yeah, so it's a number of unique projects by each of the categories. Yeah, it's quite striking in certain ah regions, especially say the UK where um infrastructure is like the second smallest or third smallest of each of of the categories. Yeah, but it was by value. By value, that's right. Yeah. Going back to the point of like this politics of of infrastructure projects.
00:11:08
Speaker
Yeah, yeah. High Speed 2 on its own would knock out like the whole residential industry you would expect. Yeah, yeah. Yeah, for sure. um And so it's it's interesting to see the distribution of that ah vary by country. So you you don't get that, for example, in the US ah here, where it's kind of roughly equal by six categories. So resi building, commercial, energy, institutional, it's roughly equal.
00:11:32
Speaker
So there's like some interesting... It kind of gives you a little flavor of ah maybe the like political landscape or how project development happens in in each country. Because as Australia is, yeah it's like flipped. Australia infrastructure is like the second largest of the categories, as an example.
00:11:51
Speaker
Yeah. Seventh on the list, the Philippines. Quite surprised to see that above France, above Germany, above Japan. Like that was quite surprised to see that. Yeah. And the large majority of it is energy and utilities. Yeah.
00:12:04
Speaker
It does say the top select countries. So yeah, I think they've ah taken a few as an example. Oh yeah, yeah they's excluded yeah. It's a selection of countries, Saudi, UAE, Philippines, Hong Kong, India, Australia, UK, Ireland, US, s Canada, Peru, and Colombia, where they can get reliable data going back to 2010.
00:12:22
Speaker
um And also just to give a sense of Giga Projects. So like everyone knows about the HS2, the high speed rail in California, the the dam that they're building in China. But there's there's Giga Projects in Turkey, in Qatar, in Indonesia, in Russia, in Panama, in UAE, in Turkey. So it's not just a top five country spends doing the Giga Projects. It's pretty broad across all continents.
00:12:45
Speaker
so Yeah.
Challenges in Project Approval and Execution
00:12:46
Speaker
So I guess the then the question to ask from from that data is like this increasing ah prevalence of of mega and giga projects.
00:12:58
Speaker
And let's say hypothetically, we're right in saying that it's being driven by the increasing complexity of the front end development of a project. We also know at the same time that the this iron law of mega projects is true, that like 90% of them go over budget or schedule.
00:13:16
Speaker
Is the politics or is this like front end development that's being driven by all these other factors like how we consult and environmental factors, et cetera, are those things effectively almost dooming ah countries to overspend and spend more on infrastructure?
00:13:33
Speaker
And is that actually the ultimate cause of productivity and construction and all of the other stuff that software companies claim to fix is ah flitting around the edges?
00:13:45
Speaker
Have we just made it so hard to get projects approved that ah that the projects get bigger so that they only have to get approved once? And then by doing that, we fall into the trap of the iron law mega projects by making things more and more complex, which that drives the productivity problem.
00:14:00
Speaker
And all the crap around how run you run your six week look ahead is just rubbish.
00:14:07
Speaker
Yeah, there's a... Asking for a friend. If you think about who's driving that, right? So yes, you've got this exponential level of complexity with Giga projects, and it's going to get harder and harder to make Giga projects more productive than than smaller projects. Because everything from just oversight to like how you actually sort of manage and govern a project to...
00:14:30
Speaker
the the the management team that are close enough to the set of detail that they're responsible for. But you would bet that like with any tender, there's a bunch of companies that go, that's not big enough on us. that's That's too small for us, or this is perfect.
00:14:43
Speaker
Are we seeing this pushed by the big consultants who run these on behalf of the clients? And they they're probably behind the scenes putting together more and more and more and incentivizing or at least pushing clients to bundle these into a ah way that those guys are probably paid on some sort of i contract size or ah percentage of of the scheme that's being delivered. So you think they're the ones actually saying, oh, no, this is the way to go?
00:15:08
Speaker
knowing fully well that it's a like a cash cow for them, rather than the actual clients thinking, oh, let's bundle these into ones. Because I'm not sure they're incentivized to, other than the approval phase with government, where they say, we'd like to spend X, let's do it once. I think you might be overstating the power of the consultant ah in that regard. i think I think the cold, hard facts to the ground are, is if the if the process you have to go through to get a project approved,
00:15:33
Speaker
And on most very large infrastructure energy or whatever project it is, it is a very large and long process that can take 10 years or or sometimes much, much, much more.
00:15:46
Speaker
I think that the consultants and whoever are just playing within the rules of the game. You know, and then it's also expensive. It's not only time, but it's expensive to run that process ah from a client, you know, to go through the approval process.
00:15:58
Speaker
I can't remember the stats that I read the other day, but it was like shockingly expensive to get like an infrastructure scheme approved or an energy scheme. And so just the the facts on the ground are probably that you try to do it once and you try to make sure it covers all the scope that you might need to do, you know, and you'd rather go through the, you'd run that race once.
00:16:17
Speaker
You're also horse trading through the process. So i don't think it's I don't think it's the consultants. I think it's the i think it's the the rules of the game. That's not like you to have that view on consultants.
00:16:28
Speaker
Yeah, that's a fair point. um I think another part of this, which is, I think, was apparent reading through the report, this volume that's increasing at like a factor of three or four.
00:16:40
Speaker
Everyone's talking about labor shortages, material shortages, procurement issues. And we do think the rate of growth against supposedly a ah contracting body of individuals who are actually work in construction projects.
00:16:52
Speaker
This sort of report, you wonder whether the holistic view will actually go anywhere for people to really think about, is this even possible? Because if everyone's we're struggling at a scheme point of view, every government's coming out and saying, we're going to do even more infrastructure yeah stimulate the economy. and You wonder how the capacity issue will play out to...
00:17:09
Speaker
um over and above budget issues and everything else. The tough thing about very large projects is the thing that doesn't scale at ah in project delivery is, is the and the bottleneck is people, like communication between people.
00:17:24
Speaker
And so as you kind of nudge that before, you know, the moment that a single project manager can't be across the whole scope of something and you need two of them or four of them or 50 of them, there is just this inherent inefficiency in coordinating across these people and managing these people. And, you know, the yeah the bottleneck and the the bandwidth limiting factor is people.
00:17:49
Speaker
ah There could be this argument that said, if you actually just did these projects smaller at 100 million project at a time, ah you could do it with small teams, deliver the project, move on to the next. So it does go it go back to my point, which I hadn't really thought of until we started having this conversation that like it feels like but the giant problem in the whole thing is the complexity of the front end loading of projects driving bigger bigger projects.
00:18:12
Speaker
Yeah. Because, you know, the way that we build software products is the smaller the project, the better. The quicker we can turn it around, get it done with a small team, the better. Yeah, yeah, yeah. And every bit of evidence.
00:18:24
Speaker
Totally. um And the coordinating of people and everything. And yeah, and it feels like that everyone knows that. But there's this one thing that we stick at the front of the project that's like, here's the 17,000 hoops you have to jump through.
00:18:37
Speaker
And then people look at that and go, well, if I'm going to go through that, I'll go through that once, please. Yeah, yeah, for sure. And then a delay on a normal size project is like X amount. The exponential growth of that delay on a mega project is, yeah, crazy. so Correct, yeah. Because you've got all these people that ah that are not like, there's not clear, there's not the same information. They don't all have the same context. Yeah, so yeah.
00:18:59
Speaker
Everything takes time. Yeah. I'm going to play referee here because we've got 12 minutes left and we've got
Introduction to AI Browsers
00:19:04
Speaker
two more topics. So we should probably do with my job. Cool. Well, let's move on because the second one is like a super quick topic that we'll cover in two seconds. But last shout out to Mace. Great report. I think we'll dive into more details that report in coming episodes.
00:19:19
Speaker
Jumping topic, hard to switch. We're back into AI. AI browsers, what is it? What the hell are they? is it going to have any impact on how we deliver construction projects in the future? So to set this concept up, over the recent weeks, ah there have been two ah releases of AI browsers. And so what is the promise of an AI browser?
00:19:42
Speaker
You know, we we can play this video or put the the link in the show notes. There's a demo from Comet of of how it will work. But the effectively, if you imagine you have your browser and alongside your browser, you have this chat interface and you can ask the browser to do things, simple things like work out a route from how to get from A to B. And the chat interface will kind of interact with your browser.
00:20:07
Speaker
They use fancy words like agent, et cetera. but it's able to ah do the things that you might do right now in ChatGPT, like answer questions, have conversations, do research, but where it needs to, it can also interact with your browser or take context from your browser. So if you're watching this podcast on YouTube, it'll be able to get the whole transcript and summarize it for you ah to bring into the conversation.
00:20:31
Speaker
So what's the what's the promise of this? Well, if you can um provide this AI interface inside of a browser where the AI can interact with the browser as if it was the user of the browser. So Carlos is logged into Chrome or you're you're probably an Internet Explorer guy, are you?
00:20:54
Speaker
Oh, come on. but that that that The Chrome is now the old Internet Explorer. i guy yeah say I'm in that box. The Internet Explorer crew migrated to Chrome.
00:21:05
Speaker
um So yeah, you're you're logged into your browser. when you're lock When you're in your browser, you're logged into all the apps you use. You're logged into your Gmail and other um tools, which means that when the AI commandeers the browser to do certain things, it has all of the context that you would have on those ah sites.
00:21:25
Speaker
So if it needed to go and find an email, read an email, send an email, it can do it as if it is you, which gets around a common problem in trying to build these agents, which is how do we permission Carlos to see the right information? how can he take certain actions?
00:21:41
Speaker
So, you know, I've been in monologue mode for a good amount of time. don't know if you had any thoughts as someone that maybe who is less in the weeds and and and might be able to translate some of the more salient points for for for a listener. That's loosely translated to Carlos isn't very good at technology. So he can talk to construction people. I know I'm on some sort of spectrum there that that might not be where the the broad populace is. Yeah, yeah.
00:22:12
Speaker
um It this would also be interesting, obviously, side note, I think Google are being forced to sell the browser because of the anti-competition thing. I wonder if that's also incentivized to be able to go through it with browsers. But I think, yeah, to put it into construction language, um and let's even ignore like software and browsers at the moment.
00:22:30
Speaker
If we just think about a day-to-day task, which could be you've got like a Google sheet or an Excel document, you're in ChatGPT now going, okay, how do I run a formula that gives me this? How do I format this table in this way? Blah, blah, blah. You ask a bunch of questions, it'll give you the answers and you'll spend time implementing those answers into your spreadsheet.
00:22:48
Speaker
The way this will go, we'll say you'll go to the browser, right, I need to take these three sheets. I need to combine them to this bit of a comparison, I need to add a column that shows the percentage change between these numbers, I need to do a bunch of things, and I also need to send it to my boss.
00:23:01
Speaker
Hit go, and then instead the instead of spending the next 30 to 40 minutes actioning, the bit of information that you have, the browser will go and do your job. And while it's doing that, you could be doing other things. And it's going to do it like 50 times quicker than you could.
00:23:14
Speaker
So the productivity unlock is massive. When you combine that with like, ah let's think of day-to-day things you might do in software construction, like form filling, document control, communication to, ah to, to raise certain formal communications to the client.
00:23:28
Speaker
It will i be able to go in and do the clicking, the typing and the thing that really takes, takes time. Yeah. That spreadsheet and that setup is a really good ah example of the two different ways to solve that problem. And so let's say you've got your you've got a spreadsheet or you've got some form you need to fill out.
00:23:44
Speaker
And at the moment, you'll go and ask questions about how I do something in a spreadsheet, maybe on Google, maybe on ChatGPT, then you go do it. If AI can go and do that for you, there's two ways it could kind of do it. One is it could be embedded inside of the the spreadsheet application in the web or on the desktop, like a Microsoft Copilot is.
00:24:04
Speaker
The problem then is it doesn't have the context of any other browser tab that you have open. And many times when you're doing a spreadsheet or you're filling some form or raising an ah RFI, you have context that's in some other system that's usually in another tab of your browser that you kind of need to translate across.
00:24:22
Speaker
Yeah. And the reality is if you're in ChatTBT, you're giving it information. You're then here's an ah RFI form and here's the spreadsheet with some data. Here's all the stuff that you need to do your job. That's gone. And so imagine you could go into the ah into your browser and say, ah open a new RFI to basically ask the question the email from Bill.
AI Integration in Workplaces
00:24:46
Speaker
And you've got got the it can go and find your email, navigate the browser, find the email, translate that into the form, open the form, ah etc. So it can... ah by By being embedded in the browser, it can not only be embedded that's in the tool that you're trying to do something in, but it can ah navigate that like you without that tool even needing to be AI native or compatible.
00:25:12
Speaker
The tool doesn't need to change at all. The browser is doing all the work. Plus it can steal context from other tabs in your browser. Yeah. Just to make a bit of a bet, do you think the tier two, tier three will be the first to adopt? Because it's way easier to adopt the tool probably in terms of like rules, regulations and internal governance.
00:25:29
Speaker
But they're probably the least connected, so the gain is bigger. What tier threes have a bunch of APIs between tools when they can just use this and keep their old methods but increase their productivity? You mean like tier two, tier three contractors?
00:25:41
Speaker
yeah Yeah, if we just think about the context of construction and who would be first use. I think the consumer is going to lead the enterprise. yeah I think like people are just going to bring this into work. You already see it right now, right? People go home and do all their work on ChatGPT and then come back in ah and... and plug that in. Well, it's what's happening, right? people If you know the company blocks ChatGPT or the company blocks AI tool, people are like not dumb. they Why spend four hours? like out These early warnings are getting heavy. It's like the equivalent it's the equivalent of like
00:26:14
Speaker
i it's the equivalent of like someone has invented a bloody like car or a tractor, and they like park it up in the field and be like, we can't use that. You've got to do it with ah with a horse and a mule or whatever.
00:26:29
Speaker
ah People are not dumb. they They go home, do the work on AI, and then come, you know, or they you know they have their own laptop that they pop out of. Students are doing it, adults are doing it. Yeah, yeah, 100%. I think it'll be the consumer leading the enterprise here.
00:26:44
Speaker
Yeah, just I just remember back to my ah days on construction science where you couldn't even go onto YouTube. and They blocked in the browser. So like, because it is browser based, it's harder to go home and and run the search to get the answer.
00:26:57
Speaker
Like the it it needs to be in your day to day. So that's probably the limiting factor of people just sort of starting with it. Yeah. You've got to install a piece of software, which that's even harder than going to some website. And you need systems are cloud based. So like planners love it.
00:27:11
Speaker
yeah and Yeah, that probably leads to like, what does that mean for the tools that you that might exist in the construction work for the moment? Well, if this browser AI enabled browser becomes the way, then you theoretically don't need anything done to the tool that you might be using in order to leverage that capability. So you could be using ah like a terribly old P6 or something.
00:27:39
Speaker
PCX being desktop, lot yeah but like a web-based tool. yeah um It's a real lifeline for shit applications because it's going to in the mix longer. Things that are embedded. Yeah, totally.
00:27:51
Speaker
I think that there's so there's a very strong argument that says that this AI-enabled browser ah basically just removes the need for any fancy SaaS and you could just orchestrate everything through this like browser level and it doesn't matter what the user experience of the end application is or the system of record is at all.
00:28:17
Speaker
The flip side is, ah don't know whether you'll take on that, whether you think, whether you would agree with that as the the premise yeah i i yeah i think i do agree um i think it's i don't think it's going to be much uh too early so i think we get excited about this and the potential but yeah if we just look at the pace of change in some of these like some projects just take a year to approve like simple applications that are standalone isolated and secure so yeah there's going to be a bit of time here but i definitely think
00:28:50
Speaker
That'll be a combination because it can it can feed information into applications which aren't great and have poor UI. um It can do that well, but it can also, if there are tools that are integrated with APIs and that's pulling information, then you could have a combination of both and the the project's still getting what they want, which is the connected set of data behind. So I think it'll be kind of a...
00:29:15
Speaker
a bit of a messy middle for the short term. Yeah, on the basis that they can even get these browsers approved. It's so early from what we're seeing. We don't really know how contractors will even react to the idea of it and the security and everything comes out of the back end.
00:29:28
Speaker
Yeah, that's true. the the the The counter argument to this is that um the idea of navigating the UI is inherently like lot lossy or like flaky.
00:29:41
Speaker
You know, the browser it can get stuck, not know where to go next. um The UI is update. you know, it'll it'll navigate as though it's a person to a degree.
00:29:52
Speaker
And so... And it will make mistakes like a person too. Yeah. So that it's like it's like flaky um and it's ah probably also quite slow. Yeah. So the other way to tackle this problem is through APIs where you know computers talk to computers and there is this ah this framework called the the Model Context Protocol MCP, which is a framework created by one of the AI model providers, which is a framework in which ah an AI model can um understand and interact with an API effectively.
00:30:28
Speaker
which is much more direct, much more controlled in the action that it's taking. So basically instead of going to your browser and say create an RFI, find the window, put the info in, create the RFI, it just calls this API and says, calls the create ah RFI endpoint, passes in the information it needs and it does it.
00:30:47
Speaker
It's much quicker, much more stable and reliable. And it could ask other questions if there's anything missing and things like that. Yeah. And so that's that's the like the the alternative approach. But um it does mean that the tools that you want to take the actions in need to be this MCP API enabled connected platform, which it means it's the it's the counter outcome for your legacy tool. So, you know, the argument is if the the AI browser thing is like the magic bullet, then legacy tools have a lifeline.
00:31:23
Speaker
But if if actually they just kind of plateau out as slow, flaky, um especially when you're taking complicated actions in complicated interfaces.
00:31:35
Speaker
Yeah, it could be that actually that doesn't get you very far and you you need to be using tools that have clear APIs, MCP enabled, etc. So takeaway.
00:31:47
Speaker
This AR browser is a thing that you'll hear more about. ah You're probably not using it today. You probably, I guess, most people are doing it.
00:31:58
Speaker
I'll predict in a year most people will be doing some version of it Most people? At my work or at home? Both, because I reckon Chrome will ship it. Yeah, okay. It's a big threat to Google.
00:32:09
Speaker
Google have to ship it. Yeah, yeah, yeah. Internet Explorer won't because they've stopped up there. It's been deprecated. They've got K-Pilot, which is awesome, right? ah No, yeah, yeah. Anyway, we forget about Edge.
00:32:23
Speaker
I know we're like mega short on time. Yeah, let me, shall I rattle to topic three quickly? Yeah, do it. Yeah, look.
Conclusion and Future Discussions
00:32:31
Speaker
So the final note is um if anyone has been tracking online ah the news around projects over the last couple of weeks or four, follow recruiters on LinkedIn.
00:32:42
Speaker
ah You'll see there's a whole bunch of schemes that have been greenlit in terms of funding from the government. um So we've got Sizewell C, everyone knows about that one. um That's all approved going ahead. Something like 20 to 40 billion worth of work, 10,000 jobs.
00:32:55
Speaker
We've got the A66, massive contract up in the north, big highway scheme. It's split it to three sections, which is run by Kier, Aureos, which is the old Calbrae and Balfour. We've got the M54, M6 projects, which is BAM.
00:33:09
Speaker
They've been waiting for a long time to get that approved. couple hundred million. The Midlands Rail Hub is a couple of billion or between one and two billion. um That is a, it's about connecting the Midland region.
00:33:22
Speaker
It's a hub in Birmingham with, I think it connects 50 stations across the Midlands. um That actually has been awarded, but it's in the ah calling period at the moment. So it's not public information. um And then there's other schemes um across the the country in terms of highways and rail projects. So yeah, Billion's being released. So we'll see them all coming online over the next sort six to 12 months.
00:33:44
Speaker
Great. Good. Now that, did you say Aureos is the new Calbrae? The actual delivery side of Calbrae broke away and that's called Aureos. So it got bought out by a PE firm or a big conglomerate.
00:33:57
Speaker
So Aureos is the new Calbrae. It's a lesser known Netflix series. Right. Thank you everyone for tuning it into today's show. If you did enjoy the episode, please do think about liking the video or following us on your chosen podcast platform.
00:34:10
Speaker
We really appreciate your support and we'll see you all next week. by Bye bye.