AI in Financial Modeling: Better Than Ever, Still Not There, and the Fatigue Factor
In this episode of The Mod Squad, Paul Barnhurst, Ian Schnoor, and Giles Male share their real experiences using AI in financial modeling. They discuss what works, what doesn’t, and how finance professionals can best use AI today. From testing different AI tools to running a “bake-off” on a real financial model, the conversation highlights the practical challenges, hallucinations, and creative approaches to working with AI in FP&A.
Expect to Learn
Where AI actually adds value in modeling and finance workflows
Why AI outputs still need careful checking and guidance
How different AI tools compare when building real models
Why instructions, structure, and human oversight improve AI results
Here are a few quotes from the episode:
“Most companies don’t have AI governance yet, and that’s where real risk comes in.” - Ian Schnoor
“Automation is great, but you still need to audit and understand every step.” - Giles Male
AI is powerful, but it is not a shortcut to quality work. It still needs guidance, structure, and strong fundamentals. The people who benefit most are those who understand both the tools and the work behind them. For now, the best approach is simple: test it, guide it, and never trust it blindly.
Follow Ian:
LinkedIn - https://www.linkedin.com/in/ianschnoor/
Follow Giles:
LinkedIn - https://www.linkedin.com/in/giles-male-30643b15/
In today’s episode:
[00:05] – Trailer
[03:06] – AI Progress Update
[05:00] – Mental Load & AI Pace
[08:03] – Busting AI Hype
[12:30] – Training & Business Demands
[17:57] – AI “Eye of the Storm”
[24:13] – Testing AI Models
[30:52] – Fixing vs Building AI Models
[35:15] – Challenging Misleading AI Claims
[38:41] – Improving AI with Instructions
[42:59] – Final Thoughts & Takeaways
Full Show Transcript
Host: Paul Barnhurst (00:00:31):
Welcome to another episode of Financial Modellers Corner. This week we have the mod squad with us, so I'm thrilled to be joined once again by Giles, Malel and Ian Giles, how you been?
Co-Host 1: Giles Male (00:00:43):
Good. I feel like I'm in a beard competition with you now. I've got a little bit of a way to go, but be scared.
Host: Paul Barnhurst (00:00:49):
Is your goal to how long you can get it? Is that the plan?
Co-Host 1: Giles Male (00:00:52):
I've hit this phase where it doesn't actually feel like it's growing. I just feel like I look like I'm homeless, but it's uncontrollable at times.
Host: Paul Barnhurst (00:01:01):
Yeah. And how have you been?
Co-Host 2: Ian Schnoor (00:01:02):
Great. Great. Missing you guys. But I did get a chance to see and catch up with Giles recently at the Global Excel summit in London exactly a month ago, and that was a lot of fun and a lot of great learning and nice to catch up with the global Excel and modelling communities. Life is busy. How could it not be? I mean, everybody that I know who's working in business accounting and finance right now is still kind of inundated, overwhelmed, tired, and excited with this world of air. I mean it's all of the above. We're trying to do our jobs, run our businesses, run our organisations, manage our lives, and learn and keep up with, right? It'd be one thing if we just had to learn ai, but what's not about learning ai, we have to keep up with it. Changing so fast is like whatever skill you learn in Giles, you're probably the best to talk with whatever skill you learned last month. Well now there's a whole set of new things and so now you're trying to keep up and figure out what you should invest your time in and learn. It's exhausting. No Js.
Co-Host 1: Giles Male (00:02:02):
Yeah, I agree. Well, we've talked about it before. I'm kind of almost a fully fledged co-pilot trainer now, and I think it'll extend to Claude. I know Paul, you're doing stuff as well and just trying to keep up with what changed in the last seven days. I mean, there were huge announcements from Microsoft two days ago that changed a whole bunch of stuff. So yeah, it is a nonstop battle just to stay broadly up to date. It's crazy and
Co-Host 2: Ian Schnoor (00:02:28):
It's a different level of exhaustion, isn't it? Sort of. It's a,
Co-Host 1: Giles Male (00:02:31):
Yeah, I read, what was the term you used? Maybe it was exhaustion, like AI fatigue or
Co-Host 2: Ian Schnoor (00:02:36):
Something. Fatigue. Yeah, fatigue. That's right. It's like a fatigue you feel, right?
Co-Host 1: Giles Male (00:02:40):
Yeah. And again, I know we are repeating old ground. I wasn't over the Excel fatigue. RegX has been on my list for years and now I've just got to park all of that and become an AI guru.
Host: Paul Barnhurst (00:02:54):
I think there's a difference in the fatigue in Excel, right, when they're coming with all the things versus ai, because we feel like with ai, if we don't get on the train, we're missing out with something that could fundamentally change everything. Did you ever feel like if you didn't learn RegX, it was really going to fundamentally change Excel and your modelling and what you're doing? Almost no new formula or change felt like, right? This is really what we're concerned about is this seismic shift. If I don't figure this out, is my business going to be dead in five years versus Oh, I didn't figure it out. Oh well
Co-Host 2: Ian Schnoor (00:03:26):
No, you're right. And what I would say to people is I think a lot of people, I mean I've been running a lot of webinars all over the world. The webinars are attracting a lot of people in them. There's a stress level, I would say most people are still not on the train yet or they're just getting onto the train. They're just starting to go and they feel stressed. They feel like they should be further. And my message is don't be stressed by that. Literally, I tell people that in the world of ai, the world of finance and modelling, it's really only February of this past year. Now that seems like an eternity ago, but it's still only about four months ago when it actually became truly usable and powerful at a mainstream level. So if you haven't gotten on yet, you still have time, right? You haven't missed the boat entirely. There's time to get going.
Host: Paul Barnhurst (00:04:07):
Agreed. So we'll continue discussion around AI a little bit more in some different areas, but in the meantime, what we want to do is we're having a bake off today, so we're really excited about it. What we're going to do, and then I'll turn it over in to explain the case, we'll kick it off is we've selected a case from FMI Giles in and I are all going to run the case will HCR what tool we're going to use in a minute and then we're going to come back at the end and then we're going to look at how they did and we're going to score 'em on some things like how good was the structure, how did they look aesthetically, were there any critical errors where it's like this just fell because the balance sheet doesn't balance and all kinds of things. So that's the basic idea. So what I'm going to do is I'm going to turn it over to N but before we do that, we're going to have each person say what tool they're going to be testing today. So Giles, what are you testing with?
Co-Host 1: Giles Male (00:04:56):
I'll be leaning into Anthropic. So I've been a big clawed fan for a long time. I'll be using 4.8. I really wanted to be able to join you today and use Fable, but for well-known reasons. Thank you. US government, I can't. You're welcome. Yeah, I can't
Host: Paul Barnhurst (00:05:13):
Because I'm the only US citizen here, so I'll take the blame. The other thing I'll say is you train people on copilot, but you're doing Claude. Okay, we'll leave that for later. Good. How full are you testing?
Co-Host 2: Ian Schnoor (00:05:25):
I think we talked about it. I'm going to use copilot, but I'm going to use the, did we say the, oh no, Giles, you said that you want me to use the cloud four
Host: Paul Barnhurst (00:05:34):
Eight right?
Co-Host 2: Ian Schnoor (00:05:34):
Quad four eight Within copilot though, because we expect that it's going to generate a different model, we're going to look at the Henderson model again to keep it consistent with what we've been testing over the last six months, but we're going to try three current modern tools. So Giles, you are going to use which Anthropic tool are you going to use?
Co-Host 1: Giles Male (00:05:51):
So I'll be in the Claude app in Excel and I'll use Opus 4.8 as the model. That'll be that one. And if you can do the same through the app, I think that was what we said, wasn't it? So you're going to use copilot in Excel two to 4.8. Yeah. Perfect.
Host: Paul Barnhurst (00:06:06):
Yes,
Co-Host 2: Ian Schnoor (00:06:07):
And I'll do it that way
Host: Paul Barnhurst (00:06:07):
And I'll be doing GPT. And they don't give you the model in the Excel, they just allow you to select, do you want it to do fast standard or heavy reasoning? So I'll be doing heavy things.
Co-Host 2: Ian Schnoor (00:06:20):
Isn't that interesting? Because of course you know that if you use, maybe you want to also go to copilot and then select chat GPT 5.5 from the dropdown. That's an option as well. If you
Host: Paul Barnhurst (00:06:31):
Correct, it is in copilot, but in chat GPT, they don't give you a model option they
Co-Host 2: Ian Schnoor (00:06:37):
Give you, right? No, that's right. That's right. But if you want to specifically know that you're using you 5.5,
Host: Paul Barnhurst (00:06:42):
You could go to copilot and use it. Yeah,
Co-Host 1: Giles Male (00:06:44):
It's kind of bizarre.
Host: Paul Barnhurst (00:06:46):
Figure all this out. Please let me know.
Co-Host 2: Ian Schnoor (00:06:48):
I'm glad that this conversation is kind of mind numbing, isn't it? This I think is contributing with the fatigue and I hear this all the time, which model do I use today? Just yesterday I ran a webinar and I said, how many of you here are finding yourself running one model and then using a second one to check the first one and the third one to kind of validate the first and second one. And you're doing that, but that's, I'm putting
Host: Paul Barnhurst (00:07:08):
My hands up. Yeah, I've done it.
Co-Host 2: Ian Schnoor (00:07:09):
In your lifetime, have you ever done anything in your life in any topic or any task where you use one tool and then you get a second one to validate the first one and the third one to double check the first and the second one and then you yourself check the first all. It's a level of fatigue that it's new, right? That
Co-Host 1: Giles Male (00:07:27):
I'm glad this has come up early on. A lot of the training that I focus on now, it's not building stuff enterprise level for me, the benefit is, so this is why I think copilot has got a huge chance of doing very well. You are in the Microsoft ecosystem, you've got that, the data encryption layer, it's all linked to the access that you've got with SharePoint and everything already. So I'm focusing on as a team, how do you leverage copilot to get rid of all the boring admin stuff? So I don't know that, I just haven't mentally looked at can you build me a financial model for months and months and months? It almost feels less relevant to me at the moment. So I think for a very niche group, obviously like pure financial modellers and I'm guessing a lot of people in the investment, the corporate world, I get that that would be the focus and I suspect those are a lot of the companies that will be looking at the shortcuts and whatever, tabis and others of the world. But for broader finance and Excel and anyone, I think it is, how do I deal with my emails and my data and remember what happened in that team's meeting three days ago? It's that kind of stuff.
Co-Host 2: Ian Schnoor (00:08:37):
Got it. So more of the automations and manage your life as opposed to building genetically building tools
Co-Host 1: Giles Male (00:08:43):
A little bit. So I think it's not always the automation stuff, but it is just, it's getting the agent to do some work for you.
Co-Host 2: Ian Schnoor (00:08:51):
Okay, well then why don't we see what happens in our,
Host: Paul Barnhurst (00:08:55):
Well, why don't you take us through the case here.
Co-Host 2: Ian Schnoor (00:08:57):
Do you want to share the screen then
Host: Paul Barnhurst (00:08:59):
And we'll all kick 'em up. So I'll bring it up on the screen and you can walk us through this thing.
Co-Host 2: Ian Schnoor (00:09:03):
So again, we're keeping it consistent. Anyone who's watched it before, we're going to use the Henderson manufacturing company model again. And this is what we all have. We all have the same file. We all have this exact same file and we're all going to give it the exact same prompt. So in our Excel we have a sheet called model, and on this sheet we simply have the last three years of historical financial statements. We have the income statement, we have the cashflow statement, and then we have the balance sheet. That's it. That's all that we have on this sheet. And then the rest is blank. And then what I did is there's a case study or is the case study, here's the case study. The case study is what an exam candidate would've been given, and it's a two page PDF case study that talks about the company.
(00:09:42):
It talks about their history, it provides information about the company's sales, about their operating costs around their fixed assets and depreciation, their working capital, their income tax, their debt, their equity. And what I did is I took this case study and I copied it into a separate sheet. I just find it works a little easier. Typically I copied the case study historic into one sheet as text. It's not nice formatted, it's not numbers, it's just gobbly strings of text in cells. And then there are instructions at the bottom just like this. I just copied the full case. And then what we're all going to do, and I'll show you and I think I was told I am going to be testing right here. This is where you want me to go, right Giles into Opus 4.8. So I am going to run opus 4.8 you can see, and literally I'm going to put in this prompt and we all have the same prompt.
(00:10:36):
It says on the case tab, you have been provided with information about a company called Henderson. You have also been provided with three years of historicals on the model tab, build a five-year forecast model with all the required schedules, revenues, cost, depreciation, tax, working capital, debt equity, and there are instructions in sales B 60 to B 66 here. And so there are some, so we'll see how well it incorporates these additional instructions over the last month or two. It's done a pretty good job building scenarios, adding an assumption page, et cetera. That's what we're all going to do. So I'll stop my share now, but we're going to do that and then at the end of the episode we will see how each of these tools performed, right Paul?
Host: Paul Barnhurst (00:11:15):
Yep. So what's going to happen is we're going to pause here for a minute. We're all going to get it kicked off, answer any questions our AI asks us, let it start running. We'll come back and have the conversation. So we'll be back here in a minute, but we're going to go ahead and pause it for a second while we do that. All right, we're back. We've all kicked off our cases they're all working in has let us know his computer is on LSD at the moment. We'll find out what that means when it finishes processing, but he said something about some crazy formats, I believe, right?
Co-Host 2: Ian Schnoor (00:11:44):
All three horses are off to the races. Apparently they're working away. It's pretty still wild for me to watch things unfold on the screen as it builds, but oh yeah, it's trippy. I don't know who taught it. I don't know where it learned in the, we always sometimes forget that AI is learning from what it can find out there in the world. I don't know where it found someone who has used this trippy formatting that I am going to show you at the end of the episode. This is, I hope it keeps it because this is interesting.
Host: Paul Barnhurst (00:12:13):
Well, fascinating. Mine just moved in, just got past the analysing the data phase and is now starting the creating and building. But when we talk about ai, we all know we're going to get different answers and if we did this bake off a week from now, we get something different. We've all heard of, we all familiar with hallucination at this far. I think everybody knows what probabilistic means now, if you didn't before and you were sharing an example of just how far or how crazy it can be sometimes with the hallucinations you were sharing with the two of us, can you show that to our audience? Maybe we bring
Co-Host 2: Ian Schnoor (00:12:45):
That up. Well I will. I've been so one, sure, I will share my screen and show you something that happened and I'll use it as a chance to pat ourselves on the back a little bit. But I mean I've been running a lot of webinars all over the world on what we're seeing in AI and modelling and been fun working with a lot of people and every time I do one, I get my views get stronger and stronger that it is going to be very important to invest in AI skills. I am excited about AI in the future, but I also equally believe that you have to continue investing in your traditional fundamental skills. No finance, understand accounting, understand excel because it is still making things up and they do make things up and hallucinate. So very recently at FMI, we won an award through an organisation called the the Electronic Accreditation Association and we won an award for the top e credential of the year online credential and I was building a LinkedIn post, the director of our exam operations, her name is Elena Alexandro using, I was working with one of the tools, I won't say well, I was working with Claude to build a LinkedIn post because I go back and forth, I'll show you the screen because it's pretty wild.
(00:14:01):
Nobody could believe this and I promise you I did not fabricate any of this. Oh,
Host: Paul Barnhurst (00:14:06):
I believe it. I've
Co-Host 2: Ian Schnoor (00:14:07):
Seen some
Host: Paul Barnhurst (00:14:08):
Similar stuff.
Co-Host 2: Ian Schnoor (00:14:08):
So I shared this with my colleagues because what happened was I had an interesting experience this morning. I was working on a LinkedIn post and I was going back. I always go back and forth. I never let it just do everything for me. And as part of the post, Claude mentioned Elena Mello during our dialogue and I had no idea who that was and I thought he was probably talking he about, I thought Claude was probably talking about the Elena on our team, Elena Alexandro. So I said, who's this Elena Mello? Do you mean Elena Alexandro or are you suggesting that I replace Elena with Elena? I was kind of tongue in cheek joking. This I kid you not literally word for word was Claude's response. He responded with the following word for word. I did not edit any of this and the messages, you must be careful. It said, that's my mistake. I fabricated the last name. I knew Elena is fmis director of exam operations, but I didn't have her surname and instead of leaving it as Elena, I invented Marlo out of Finn Air. That's not a small error. Making up a person's name is exactly the kind of thing that erodes trust. Sorry about that, Elena Alexandro got it. Won't happen again. But I shared this because
Host: Paul Barnhurst (00:15:18):
I'm just glad he used M dashes
Co-Host 1: Giles Male (00:15:20):
And apologised.
Co-Host 2: Ian Schnoor (00:15:20):
Yeah, exactly. So many of them actually there's only two, but the point is if you don't understand what deferred taxes is or minority interest is, if you don't understand what working capital is and it fabricates and invents a way to forecast something because it thinks that's the right thing to do, you need to be able to check and catch it and understand it. And we're seeing this kind of thing happen everywhere. Anyway, I thought
Host: Paul Barnhurst (00:15:43):
A kind of a similar example to build on that in and Giles, you'll remember this. We were testing your favourite tool and if you remember the balance sheet was out of balance and it worked. It got a little closer, then it finally came back and said, well, 0.3% is an acceptable variance. It's okay for the balance sheet to be out of balance if you don't know what you're doing, you might buy that and think, oh okay, that's okay. You don't know that it's 3 million out on 3 billion on one side and 2.97 on the other and it's netting to a tiny number.
Co-Host 1: Giles Male (00:16:13):
I remember that. But also Excel's been solved so I don't even know why we're bothering. Isn't that right?
Co-Host 2: Ian Schnoor (00:16:22):
It's done. Little
Host: Paul Barnhurst (00:16:24):
Tongue in cheek there for those.
Co-Host 2: Ian Schnoor (00:16:25):
Drop your microphone, just drop it, it's done. Excel's done.
Host: Paul Barnhurst (00:16:31):
So I think while we're letting this run, min now has some pages and it's building pretty good. It's done step three of six. So
Co-Host 1: Giles Male (00:16:39):
I'm on one of eight.
Host: Paul Barnhurst (00:16:41):
Mine almost built out the assumptions page looks pretty good, but we'll let it keep going. Let's talk a little bit about Fable five because I know both Giles and in keep giving me crap about the US reigning on everybody's parade. So Giles, what happened with Fable five? You want to tell us?
Co-Host 1: Giles Male (00:16:57):
It came out and I was using it a lot. The original plan was up until the 22nd of this month I think. So I'm on a max plan. I think the idea is that you were
Host: Paul Barnhurst (00:17:09):
Now I upgraded last week.
Co-Host 1: Giles Male (00:17:10):
Yeah, and it's great. So you've got two types of max plan one where you get five times the kind of tokens token on this 20. So I was like, right, I need to use every token I've got on Fable because the outputs I was getting for the first time within Excel world were outrageous,
Co-Host 2: Ian Schnoor (00:17:28):
So noticeably better and I didn't get a chance to use it before they shut it down. Noticeably different,
Co-Host 1: Giles Male (00:17:33):
Noticeably better. So wanted, one of the things I wanted to do for this episode was get it to build an XLE sports case about the mod squad and the three of us. I fed Fable four prior XL ees sports cases and basically said this is kind of what we do in XLE sports. Have a look at the models, come up with something creative and okay, it wasn't perfect. There was a lack of genuine creativity in some areas, but the way it looked, the way it operated, the answers that it came up, it did a whole map case that it was steps above anything that I've ever seen before. But then the US government decided that it was not up to scratch on a security mitigation.
Host: Paul Barnhurst (00:18:16):
They just said it can't be shared with certain foreign countries. They didn't say philanthropic has to shut it down.
Co-Host 2: Ian Schnoor (00:18:20):
Didn't they say all foreigners? Didn't they say all foreigners or was there a list?
Host: Paul Barnhurst (00:18:25):
I think it's the one they have certain export relationship they don't have relationships with or something. But the reality is what that meant is philanthropic didn't have a way to restrict it to the level of making sure. So it meant they had to shut it down
Co-Host 1: Giles Male (00:18:37):
And it's still down. But from my perspective, it was very, very impressive. It could take longer to get to answers, but it was coming out with things that I'd certainly never seen before.
Co-Host 2: Ian Schnoor (00:18:50):
So steps above even 4.8,
Host: Paul Barnhurst (00:18:53):
My experience, I didn't get to test it on enough, especially Excel stuff. So it seemed good. There were some interesting things that did, but I didn't see enough to say, oh yeah, this is markedly better. So that's interesting that you saw that and what you're doing because I've heard mixed and we see that with everything, right? Because different models are tuned different ways and it's so weird to have that conversation a couple years ago. I don't care just of it now. It seems like it's a regular conversation. I think the second thing that's interesting, did you get a chance to play with it at all in did you?
Co-Host 2: Ian Schnoor (00:19:22):
You know what, I didn't. I was biting my time. I knew I had plenty of time and then bam, they just shut it down. Were you playing with it Paul? I got to play with
Host: Paul Barnhurst (00:19:32):
It a little bit. I'd run a few things with it. I helped to build some website stuff and it seemed similar in that area that I had used it. I hadn't got to test it on any Excel stuff. I had planned on doing some of that this week, running it through a case or two and seeing, especially after seeing what you said Giles, but at last my government disappointed me again. That's pretty normal.
Co-Host 2: Ian Schnoor (00:19:54):
It's okay. I'm not upset or offended by it. I think it will come back and if they need a little more time, if they need a little more time to get comfortable, this is, as we all know, this is pretty potentially wild and dangerous stuff. So I don't have a problem with them. I will share with you, I will share with you very quickly. So the two of you are part of the Global Leaders Council and I'll give you a preview. You both contributed to our first research report. There are 63 members of the council all over the world and an extensive survey and we're turning that into a beautiful, beautiful, almost soon to be released research report. Most people commented that their organisations have absolutely no formal policies. I don't want to spill too much yet around AI and the usage of ai. It really feels like the wild west out there. Some companies do, but it's sort of random. Some companies have allowed certain tools and randomly restricted other tools. It feels like nobody, whether it's the government or IT teams, nobody can keep up with the support that needs to happen around this type of tool. And I don't know if you guys are feeling or seeing that you must with the companies you talk to.
Host: Paul Barnhurst (00:21:03):
So as you both know, I run a podcast called Future Finance with Glenn Hopper. And Glenn Hopper is one of the experts out there globally. He speaks all over the world on AI and one of the biggest things they're finding is the companies that are falling behind don't have governance and AI governance is not IT Governance and governance is not sticking your head in the sand and say don't use it, then you're going to have shadow use. It's going to come from the bottom up and it's just a matter of time until somebody does something that cost the company a lot of money. So governance is a huge area where people are falling behind and we've had several guests on that show specifically talking about governance and the things you need to be thinking about and measuring ROI and what's the role of the CFO and this whole governance thing. So I think it's if you don't have policies at your company, ask why not because the reality is it's a recipe for bad things in the long run. Just like if you don't have good iced tea security, who's ever dealt with a hack, a breach, any of that stuff not fun, right? It's a huge cost and a huge risk and AI is no different.
Co-Host 1: Giles Male (00:22:11):
Yeah, I agree. I mean that's certainly what we're seeing from the companies that are maybe a few steps ahead. They have an internal AI governance policy and you'll see the word champions pop up all over the place. They'll be called other things as well, but having a team of AI champions and actually the company that we're working with now, an AI centre of excellence, I think you're going to see all of these terms and assets popping up it. It's funny though to me it's not that different from anything else. If you go through what you two have probably seen before in other areas like the introduction of a new tool or the integration of SAP or Power bi, you should go through those phases like okay, what do we need it for? How are we going to use it? Who are going to be the experts? How are we going to spread that down across the team? What are they going to be doing and how are they going to use it? It's very similar but I guess the risks are even more apparent now.
Host: Paul Barnhurst (00:23:08):
It is similar and I'll add one thing. Studies have shown compared to I guess the expectations, the ROI you set out to get your problem statement, whatever, most digital transformations, most big software projects fall short.
Co-Host 2: Ian Schnoor (00:23:23):
Most software projects fall short. Meaning what?
Host: Paul Barnhurst (00:23:27):
Fall short of the ROI you established or what your stated goals were. Some just outright fail like 20, 30% or basically a failure. You may even just roll it back and go with the old software, right? Most major software implementations fail to meet whatever the goals were, something like 70%. And the biggest reason we here for that is change management, not 70%. It's like 20% that actually they end up maybe rolling it back or call it a true failed deployment versus falling short. So if you think those type of numbers, why would it be any different with ai? Because it's a change management issue more than anything. It's a person issue, it's a planning issue. Rarely did you just pick a software that's completely garbage and can't do anything. You went through a decent enough process, you didn't manage to change management and the implementation and changing your processes and what I worry about is with how fast AI is moving, are we really taking the time to do that?
(00:24:22):
Prime example, I want to say it was meta, so they wanted to overhaul their, I think it was their AR process, we'll say it was AR for the example. They spent 14 months before they did anything with ai, documenting, streamlining, figuring out what the process map should look like for using ai. Then they built it. They went from something like six days to one day, but they spent 14 months to get there. And so with how quick this is moving, I think everybody's worried about being left behind and there's going to be a lot of bad decisions made along the way. I know we're getting a little away from modelling,
Co-Host 1: Giles Male (00:24:55):
But it is such a good point and again, I think hopefully when any company goes through that process, what are we doing? What are the workflows within a workflow, what are the steps that might be an area where you can get value from using ai. I think for a lot of companies, if you go through that process, a lot of it's going to hit on automation, which you could have done 10 years ago. It's only going to be some things where you go, yeah, we need copilot studio to be running autonomous processes every hour of every day. I bet you for so many companies it would just be get your data in the right state, manage it properly, get the right security around it, automate as much as you can, whether it's power query, python, power bi fabric, whatever it might be, and AI can be the icy on the cake.
Co-Host 2: Ian Schnoor (00:25:46):
You raise an excellent point and I've heard this as well, people are working hard to investing time in building skills and using the code components of each one to automate things and you're absolutely right. These things were all automatable for a long, long time using other tools, but now people sort of know that they're automatable with ai. It's funny, right? I may not have even been the best way to do it, to automate it.
Co-Host 1: Giles Male (00:26:13):
Well especially, I mean if you've looked at, I'm not a co-pilot studio expert at all, but that is the proper programmer side where you're dealing with multi-stage autonomous agents but there is a deterministic layer to it which is basically power automate with other bells and whistles. So part of copilot studio, you've got this probabilistic agen layer but then you've got all these workflows which is pretty much power automate.
Host: Paul Barnhurst (00:26:43):
Agreed. So I will say mine finished. Mine finished about a couple minutes ago.
Co-Host 2: Ian Schnoor (00:26:46):
Mine finished two minutes ago, so mine took around 18, 17, 18 minutes. I don't know, I think
Host: Paul Barnhurst (00:26:50):
Mine was around 14 or so. It might've been a little earlier. It took me a minute to realise that it actually finished, so mine was pretty quick. There are some issues but is been running,
Co-Host 1: Giles Male (00:27:02):
Mine is having a really good think. It tells me it's still on step one,
Host: Paul Barnhurst (00:27:07):
You might need to rekick yours off.
Co-Host 2: Ian Schnoor (00:27:09):
You
Host: Paul Barnhurst (00:27:09):
Might need to put a prompt in there saying
Co-Host 2: Ian Schnoor (00:27:12):
Has it built anything yet, Josh?
Co-Host 1: Giles Male (00:27:14):
Yeah, it started to build schedules which look fine.
Co-Host 2: Ian Schnoor (00:27:18):
Okay, so it
Host: Paul Barnhurst (00:27:18):
Is building then at least it's,
Co-Host 2: Ian Schnoor (00:27:20):
It's working
Host: Paul Barnhurst (00:27:20):
Away.
Co-Host 1: Giles Male (00:27:22):
I think it might be having an issue but we'll see. I think it new to go through yours first and then I'll see how far I can get.
Host: Paul Barnhurst (00:27:29):
Real quick before we go through the models, we mentioned Fable five and I think one thing worth mentioning that I think everybody needs to consider about is the pricing with Fable five. There's two things they announced. They gave us a two week free window till June 22nd and then they said model fable five is a hundred percent consumption based, so none of it sits in your subscription model and it was $10 for every million input tokens and $50 for every million output tokens and they use twice as many tokens as Opus 4.8. I'm already hearing all these people talk about how they're burning through tokens. There's a whole, I think it's very similar to Excel. When people first learned Excel, how many models were there that were just horrendously big and calculated really slow? Oh I got to go to lunch while it calculates and I'll come back and now over time we've learned to optimise.
(00:28:25):
The problem is that didn't cost you anything but time in Excel. That optimization could cost you a fortune with ai. So it's something you need to keep in mind as you're learning, especially as what philanthropic and OpenAI are both going public and I guarantee you as soon as they go public all of a sudden it's not going to be about growth. It's also going to be what's the path to profitability and one of those key elements is going to be price increases. We just saw it with Fable five, we just saw it with copilot, so I would love to get your thoughts on that kind of element. You could take it wherever you want and then we'll get into the model. So Giles,
Co-Host 1: Giles Male (00:29:02):
I mean broadly pricing. So yeah, I said earlier I was kind of working towards that 22nd of June deadline because I get the feeling fables going to be really expensive but Microsoft as well. So they've just announced that cowork again, which has kind of come over from philanthropic that's going to be on a consumption basis. So I think it's going to be a really interesting period of time over the next few months where the really early movers with cowork through Microsoft may have transformed so many of their processes to get onto cowork through Frontier, which until two days ago was just part of your $30 a month enterprise subscription package and now it's all flipped and for the most part it's all going to be consumption basis on top of your $30. So that's going to be really interesting for companies that have maybe got rid of people and gone all in early on ai that bill's going to start not just creeping up but probably shooting up really quickly.
Host: Paul Barnhurst (00:30:06):
Well it's when philanthropic started to charge separately for agents a little while back, same idea.
Co-Host 1: Giles Male (00:30:12):
I can understand it from the other side where this is expensive. So if you just imagine I never quite understood how you could have things like cowork and even a lot of the other declarative agent stuff, which still costs token effort. How can you just have almost limitless usage of that? I think it was inevitable this was going to happen. I think it's just the start of prices going very, very high.
Host: Paul Barnhurst (00:30:39):
I'll let you go and I'll share one other example. I'll spend 30 seconds. There's a data centre they're building here and this just talks about the cost. Just think the energy costs and this is just for military for the US that they want to build here in Utah and the energy for it is twice the energy used by the entire state of Utah, which is nearly three,
Co-Host 2: Ian Schnoor (00:30:57):
Hang on an AI data centre. They're looking,
Host: Paul Barnhurst (00:30:59):
It's a military ai, it's going to be used for the US military, but it's an AI data centre. 10,000 acres, it will use nine gigawatts of data of energy, which is twice the entire state of Utah's energy usage just for one data centre for the US military.
Co-Host 2: Ian Schnoor (00:31:17):
And did they talk about where are they getting energy from? How are they powering it?
Host: Paul Barnhurst (00:31:22):
Yeah, there's a lot of ongoing debate about that. They're tapping into a gas line coming from our good friends to the North Canada and they also said they're going to use some renewable energy. There's a lot of concerns I have about it. I will go there. But just think of the cost. That's one data centre, just the US military and it could power roughly 6 million people or almost 2 million houses.
Co-Host 2: Ian Schnoor (00:31:47):
I don't think there's any particular economic model that would point to where the pricing will cap, where will the pricing? That's unusual to have a product where there is competition obviously, but yet pricing. You could be diving into a product that for which pricing continues to increase and then your only option will be to stop using it and hire a person instead, which it may be a lot cheaper. And actually people are starting to talk about that.
Host: Paul Barnhurst (00:32:15):
Yeah, exactly. There's some cases where it's cheaper to hire somebody is what people are going to start finding. So they'll have to settle into that balance. You're going to say something Giles.
Co-Host 1: Giles Male (00:32:25):
I do think there's a lot of headway for, I hope Microsoft don't listen to this, but I think there's a lot of headway for prices to increase because even if you are, let's say the output from your AI tool genuinely saved three or four people's worth of FTE effort, you're getting into very significant six figure sums that you are saving that could go into those kind of token costs and subscription costs. So I don't know, I guess every company's going to end up looking at going, all right, how much do I get from this? How much do I get from humans? Probably need a bit of both. Where's the balance?
Host: Paul Barnhurst (00:33:05):
Yeah, there's a CFO who recently said ROA could double the cost on everything and I wouldn't even hesitate to pay it.
Co-Host 1: Giles Male (00:33:11):
Yeah, I've said something similar
Host: Paul Barnhurst (00:33:13):
Because the value they're getting, so everybody's going to have to make that decision.
Co-Host 2: Ian Schnoor (00:33:17):
Yeah, I don't think everyone would feel that way on our team. We got certain low level licences on one of them for everybody. If they doubled the costs, if they doubled or tripled or quadrupled the cost, no, we would look at, I don't know what we would do. We would look at only certain people. We are not like a lot of companies, we're not going to just double or triple or quadruple our AI costs. Every time a firm says yeah, we're just doubling the price, that's not going to work. So we will have sort of, there is still some elasticity somehow, somewhere where it is on the curve there will be everybody has a
Host: Paul Barnhurst (00:33:53):
Limit.
Co-Host 2: Ian Schnoor (00:33:54):
It's just going to be, yeah, I'm out. I can't Will that CFO, the CFO who said that they'd be happy to double their price, would they feel that way if that meant a thousand people on their team were using it and the price doubled and then doubled again and then doubled again? I know wouldn't everybody have a threshold at some point
Co-Host 1: Giles Male (00:34:14):
That's
Host: Paul Barnhurst (00:34:14):
Risky. At some point it's cheaper to hire people,
Co-Host 1: Giles Male (00:34:16):
But also the people cost is to a certain extent more predictable and in your control
Co-Host 2: Ian Schnoor (00:34:22):
Way more predictable.
Host: Paul Barnhurst (00:34:24):
There's a reason I don't have a robot doing all my meals upstairs. Even though you can buy some can't afford it.
Co-Host 1: Giles Male (00:34:31):
I'm pleased to announce Claude is on step eight of eight. So we have made progress in the last five minutes.
Co-Host 2: Ian Schnoor (00:34:36):
It's unbelievable. So I use Opus 4.8 but through copilot it ran 23, it took 17 or 18 minutes. It's done a nice job It and it did 23 steps along the way. No 29 steps. It did 29 steps to get a model. So I don't know, but yours is still working. Giles, it hasn't given up on you, it's
Co-Host 1: Giles Male (00:35:00):
Nearly done. It's on the print setup of checks. It must have read some FMI documentation about the importance of print setup.
Host: Paul Barnhurst (00:35:07):
Mine was, I think it said it had eight steps, but it's no longer, I can't see how many steps it did. Now it doesn't show it. That's kind of disappointing. I'd want to see the steps when it's all done.
Co-Host 2: Ian Schnoor (00:35:16):
Mine I have to go to the very top of the, and it tells me up there what it did, how many steps? 29 steps. Yeah,
Host: Paul Barnhurst (00:35:21):
No, I know it. I'm not seeing it in here. I thought it told me but I don't. Interesting. Let me add to stage. So I've made no changes. So you can see right now, and this is something I actually did with co-pilot's, new personalization. One of the instructions I put in is please autofit the width of each column when you're done because it does this all the time. I don't know if you guys have noticed this but right, so tiny thing, let me just expand it out.
Co-Host 2: Ian Schnoor (00:35:49):
That's a beauty. I encourage you to hand that into your boss that way with all those
Host: Paul Barnhurst (00:35:54):
You could see it said, but it did mention it fixed the print ranges
Co-Host 2: Ian Schnoor (00:35:57):
Nice.
intro (00:35:57):
So
Host: Paul Barnhurst (00:35:58):
The will print, you can't read it but it will print.
Co-Host 2: Ian Schnoor (00:36:01):
So first of all, some decent points for formatting. Not bad. I mean it's made an effort to try and do something presentable.
Host: Paul Barnhurst (00:36:10):
Let me make it not quite that big. Let's go there. So how they laid it out, we can see 26, 27, 28. It did a summary. It didn't do it on the model sheet. I would've liked to see, well it did bring it in here as well. It looks like
Co-Host 2: Ian Schnoor (00:36:27):
Very interesting that it collapsed the columns and made them.
Host: Paul Barnhurst (00:36:30):
So it looks like interesting enough it did it here. This is its summary. Okay, so let's go. I don't like the order. It's model assumption summary.
Co-Host 2: Ian Schnoor (00:36:43):
Not at all. Like
Host: Paul Barnhurst (00:36:44):
Summary assumptions model.
Co-Host 2: Ian Schnoor (00:36:46):
Exactly. It went opposite for some reason. Okay, so
Host: Paul Barnhurst (00:36:48):
I'm just going to switch that around real quick just as we read through this and I'll put the case at the back. Alright, so let's go ahead. It provided a summary clear enough. It did. I would say decent. Okay. EBITDA margin, ebit, netcom and in cash. Total debt net capacity utilisation. Let's see. See,
Co-Host 2: Ian Schnoor (00:37:08):
Can you click on one of the best case numbers? Go down and click on some of the best. I wonder what it did click.
Host: Paul Barnhurst (00:37:13):
It looks like all the numbers are coming from below. So it put a summary up here.
Co-Host 2: Ian Schnoor (00:37:18):
Well I'm very curious to know how it's linking to.
Host: Paul Barnhurst (00:37:21):
That's what I'm trying to get to. Base
Co-Host 2: Ian Schnoor (00:37:23):
The best and the worst at the same. It was in row 80. So
Host: Paul Barnhurst (00:37:25):
Lemme take one. Let's just, oh,
Co-Host 2: Ian Schnoor (00:37:27):
41 and then 81 B 41 control
Host: Paul Barnhurst (00:37:29):
Bracket. Give me a second. Alright, so net revenue here. So did the what? That doesn't make any sense. Am I? Okay, so hold on. B 41, let's go to B 41 and yeah, so it did a calculation here for net revenue B 38. So units sold times gross price minus freight divided by a thousand.
Co-Host 2: Ian Schnoor (00:37:57):
Yeah, you don't really want to see any calculations like this on a summary sheet.
Host: Paul Barnhurst (00:37:59):
Correct. And so then this one, so it looks like where's
Co-Host 2: Ian Schnoor (00:38:05):
That coming from? What
Host: Paul Barnhurst (00:38:05):
You did here? Is it that the calculate, so this says capacity utilisation, but it's where it did the calculations for the base case. Then it did the calculations for the best case here. It did 'em off the assumption
Co-Host 2: Ian Schnoor (00:38:19):
It's basically built three separate models. It didn't know how to run a base best and worst off one model with scenario management or sensitivity tables or switches. It basically reran all the calculations. Three different ways to get a base correct.
Host: Paul Barnhurst (00:38:32):
That's what it looks like it did here, but it linked into the model.
Co-Host 1: Giles Male (00:38:37):
Okay, so it's given you a switch on the model has it?
Host: Paul Barnhurst (00:38:40):
No, no, no. Not a switch. What I'm talking. So this beginning revolver links to the model. The model doesn't have switches. It's decided all three of them the bank debt. Well that makes sense. The bank debt revolver from 25. Your beginning is always going to be the same because it's
Co-Host 1: Giles Male (00:38:56):
An actual Yeah, but what's it done on the forecast? It's
Host: Paul Barnhurst (00:38:57):
A terrible way to try to trace it out.
Co-Host 1: Giles Male (00:38:59):
But what's it done on the forecast? How is it picking one of the three models with the three outputs?
Host: Paul Barnhurst (00:39:04):
It just linked it to the schedules. It's built on this page. So it did the work on this page of one model then it recreated three for each of the cases. Are they consistent? Not sure yet.
Co-Host 1: Giles Male (00:39:18):
Okay.
Host: Paul Barnhurst (00:39:19):
Your look
Co-Host 2: Ian Schnoor (00:39:20):
Done.
Host: Paul Barnhurst (00:39:21):
A
Co-Host 2: Ian Schnoor (00:39:21):
Decent job on the formatting I guess presentable,
Host: Paul Barnhurst (00:39:24):
Right? So if we go through this, let's just run through its main model and kind of flow. Okay. Do
Co-Host 1: Giles Male (00:39:28):
You mind zooming in a little bit just for my eyes?
Host: Paul Barnhurst (00:39:31):
Yeah, yeah, I'll zoom in a little bit. How's that?
Co-Host 1: Giles Male (00:39:34):
Yeah, great.
Host: Paul Barnhurst (00:39:35):
Okay, so what we have is we got our income statement that's all linked below. That all seems to make sense. I'm not going to go through numbers, right? I just want to see it all appears to be linked.
Co-Host 1: Giles Male (00:39:47):
Well you've got some interesting current tax numbers going on there.
Host: Paul Barnhurst (00:39:51):
Yeah, we'll look at the schedules. I think it did this right in the sense of you got your cashflow statement, it's all linked below. You have your balance sheet, it's all linked below.
Co-Host 2: Ian Schnoor (00:40:03):
It's using the wrong colour coding but it's okay. Balance sheet
Host: Paul Barnhurst (00:40:07):
Doesn't balance. It does. Nevermind. I thought it didn't for a minute. I was like okay, so it balances does it does just
Co-Host 1: Giles Male (00:40:17):
Go up
Host: Paul Barnhurst (00:40:17):
A little bit and then what it did months
Co-Host 1: Giles Male (00:40:18):
Going up a little bit just to the cash. What was the cash line?
Host: Paul Barnhurst (00:40:21):
Yeah, give me one. It'll go up to that. First. You want to see cashier,
Co-Host 1: Giles Male (00:40:25):
It's always the first thing that I used to note to is sometimes you had numbers on cash and the revolver and god knows what else.
Host: Paul Barnhurst (00:40:32):
So the cash goes to zero and stays there, which is a problem for all five years
Co-Host 1: Giles Male (00:40:36):
Leaning on the revolver. That's the check that I wanted to see. So you never go cash positive, you don't have a minimum balance. Yeah,
Host: Paul Barnhurst (00:40:43):
I mean you should be able to paying it off but yeah, get it. So what it's done, which I do is this is one of the few times I've seen where it did not do any calculations on these schedules. Now it did create three calculation details for each scenario, which is bad.
Co-Host 2: Ian Schnoor (00:40:59):
Sorry, it did not build, I
Host: Paul Barnhurst (00:41:02):
Don't see any calculations taking place outside of
Co-Host 1: Giles Male (00:41:06):
Yeah, the structural. Yeah that's good
Co-Host 2: Ian Schnoor (00:41:08):
On the three statements.
Host: Paul Barnhurst (00:41:09):
It builds
Co-Host 2: Ian Schnoor (00:41:10):
Where the schedule. That's good. The schedules are down below.
Host: Paul Barnhurst (00:41:13):
Now we're into the schedule. So what I was going to do is I'm just going to hide these columns so I can bring this over a little bit and kind of freeze the pain maybe here just as we're moving. Alright, so that's done. So alright, so went off the assumptions and did the math there for units sold, which I think there was a capacity thing. So that seems to look roughly right, the revenue, okay, it goes down. I think that looks freight and warehousing. I mean as I look at this first part goes up to a hundred percent capacity utilisation and stops
Co-Host 2: Ian Schnoor (00:41:51):
It itself. It feels right. That's what I would've expected for this case.
Host: Paul Barnhurst (00:41:54):
So I think the revenue schedule just a high level without going through every row looks fine, right? I don't see has any real issues. Now let's go to the cost schedule. It's doing a nice job. Raw materials, what did it do here? Alright, assumptions doing a nice job. It seems reasonable at a high level. I mean costs are going up as revenue.
Co-Host 2: Ian Schnoor (00:42:18):
Yeah,
Co-Host 1: Giles Male (00:42:19):
I think it's fine. You've got a single cell inflation assumption somewhere that I think it's just linking everything to Yeah,
Host: Paul Barnhurst (00:42:25):
Yeah. We can look at, so if we look at the assumptions, let's just take a minute. So they did the assumptions for base best and worst they did gross price multiplier, volume growth, annual growth thereafter, I'm not sure. Okay. They did 2026 and then every year after inflation they made assumptions for each case. They made assumptions for D-S-O-D-I-O and DPO. That's a reasonable set of assumptions to vary. So that all seems fine. Annual case inputs, these were the inputs that were given. I liked that it gives you the notes. Case B 11 gives me the cells that it took it from so I can validate it. So I like the assumption page there. Same with the operating. It says hey, I took 'em all from there. So it tells me what's the no assumption based on run rate and it goes through and gives all its assumptions. This page
Co-Host 1: Giles Male (00:43:20):
That looks fine. I mean you could tic and say it's got calculations on the inputs but
Co-Host 2: Ian Schnoor (00:43:25):
Love. Yeah, I mean I don't love that but the
Host: Paul Barnhurst (00:43:30):
Historical driver calculation, I wouldn't have.
Co-Host 1: Giles Male (00:43:33):
Okay, that
Co-Host 2: Ian Schnoor (00:43:34):
Pulls that right out of the case. The 5 94 was in the case study. It just hard kitted it in. That's not very good.
Co-Host 1: Giles Male (00:43:39):
Yeah,
Host: Paul Barnhurst (00:43:39):
So that's what we have. Anything else? You guys there you can see that. Do you
Co-Host 1: Giles Male (00:43:44):
Know what I know
Host: Paul Barnhurst (00:43:45):
Why it's doing an index. Oh it's just pulling it from above
Co-Host 1: Giles Male (00:43:48):
For these one. It looks pretty okay. I mean probably a bit of an improvement on a lot of, well actually a lot of improvement on what we saw last year. Still some of the same minor issues I guess some of the formatting, some of the choices to hardcode numbers pulled in different directions. Not perfect, but I mean you've got a balancing balance sheet and the numbers looked sensible.
Host: Paul Barnhurst (00:44:13):
Yep. So any other, just as I was scrolling down, how did they do their depreciation? Oh
Co-Host 1: Giles Male (00:44:20):
So that's interesting. They've got a semi fixed range. I mean that's a reasonably intermediate modelling technique to semi fixx. A range to do depreciation over time.
Host: Paul Barnhurst (00:44:32):
Pretty. Yeah, I would like to see a but it's not bad.
Co-Host 2: Ian Schnoor (00:44:35):
What's the new asset depreciation?
Host: Paul Barnhurst (00:44:38):
All they did is they basically did a sum of the two years and divided it by an assumption.
Co-Host 1: Giles Male (00:44:43):
Yeah, but look at the next column over go to the next over to the, right now it's
Co-Host 2: Ian Schnoor (00:44:49):
CapEx. Yeah.
Co-Host 1: Giles Male (00:44:50):
Go over to the right and then hit F two. So that's what I'm saying. That's impressive that it can do a modelling technique with a semi. You're
Co-Host 2: Ian Schnoor (00:44:58):
Saying
Host: Paul Barnhurst (00:44:58):
You did it as
Co-Host 2: Ian Schnoor (00:44:59):
A
Host: Paul Barnhurst (00:44:59):
Rolling
Co-Host 2: Ian Schnoor (00:44:59):
Each
Host: Paul Barnhurst (00:45:00):
Year.
Co-Host 2: Ian Schnoor (00:45:01):
It's a very subtle technique to lock the first cell reference and not the second one so you can Yeah, it was smart
Host: Paul Barnhurst (00:45:06):
Enough. So yeah,
Co-Host 2: Ian Schnoor (00:45:07):
That's good.
Host: Paul Barnhurst (00:45:09):
Alright,
Co-Host 2: Ian Schnoor (00:45:10):
Great start. Do you want to share your
Co-Host 1: Giles Male (00:45:13):
I can go to next. I think Paul you need to,
Host: Paul Barnhurst (00:45:16):
I stop sharing and share you, which I believe is this one here.
Co-Host 1: Giles Male (00:45:20):
I need to bear with me. I need to share my go
Host: Paul Barnhurst (00:45:25):
Quick. Well you're going to share overall if you were handed this, I mean we haven't gone through it in detail, but just initial thoughts, kind of final thoughts. I think for me a couple things I really liked that at least did individual schedules. That's an improvement over a lot of what we saw last year. I hate the way it did scenarios of detailed case calculations and so there's linking issues. I would've done a little more detail on the depreciation. So there's some messy things, but overall it's pretty good, right in the hand of the right modeller, you could take this and finish it up.
Co-Host 2: Ian Schnoor (00:46:00):
I'm not supposed to see this. It's done an amazing job in 20 minutes or 15 minutes and yet I beg people not to ever hand that in to your client or your boss and people will, it is not going to be a good day when you hand in something like that, it's off to a great start and now you'll probably spend an hour or two hours learning it, understanding what it did. So you can answer the questions about what's going on, how is it built? You're going to find issues you want to improve formatting issues. Like you will spend time and you can either iterate with the tool to have the tool do it or you'll just do it yourself. But you need to know what to look for and you need to know where there are challenges. So yeah, I mean I'm not surprised it looks, and
Host: Paul Barnhurst (00:46:43):
One last thing I do like that it has a check section. There's a few more I could add, but it has a cash roll forward check and a debt schedule tie.
Co-Host 2: Ian Schnoor (00:46:51):
Click on one of the balance sheet checks in the third year, let's say. Is it? Yeah. Great. Okay. Charles, were you
Co-Host 1: Giles Male (00:46:59):
In all of their defences in terms of formatting? We are not throwing full capabilities at this because you do have custom instructions and skills files and all of this stuff that you can add. So we could, I mean 20 minutes, like you say, we could
Host: Paul Barnhurst (00:47:16):
Improve this quite a bit with skills. Do
Co-Host 1: Giles Male (00:47:19):
You know what? That would be a really interesting next episode we go away, we do the same thing but we put everything on it. We get our custom skills and instructions in there and better prompting. That'd be
Host: Paul Barnhurst (00:47:33):
Interesting. We should do that at some point. I still got to build out those skills.
Co-Host 2: Ian Schnoor (00:47:37):
So that takes a lot more learning, right? Correct. It's one thing to know how to just put in a pretty basic prompt like we did. It's a whole other skillset to know how to go in and build skills and have a library and a now you're, anyway, that's great. Giles, how did yours do an
Host: Paul Barnhurst (00:47:53):
Example for, I know someone build 81 skills for a project they're doing and others built. 19. That's a lot of time. Alright, we're going to turn it over to Giles now.
Co-Host 1: Giles Male (00:48:01):
So I haven't touched it. I haven't changed the order of anything. So we will just, I'll tell you why don't we start. So I have a summary.
Host: Paul Barnhurst (00:48:08):
You got summary last as well.
Co-Host 1: Giles Male (00:48:10):
I got summary last. I actually think the order's okay because it's basically you enlarge it a
Host: Paul Barnhurst (00:48:14):
Little bit from
Co-Host 1: Giles Male (00:48:15):
My
Host: Paul Barnhurst (00:48:15):
Old eyes.
Co-Host 1: Giles Male (00:48:16):
It's treated the original input tab as the output essentially. So actually you do have assumptions, schedules, output and summary. I think that's fine. It's given me a nice little summary. Five year cumulative numbers and then the final year, which is fine just at a high level. So this is my five year forecast here. Formatted nicely. I think. I mean this. Let me just go to, sorry about that. I just wanted to see if you look here on the right all formula driven to the,
Host: Paul Barnhurst (00:48:55):
Although it used a let formula, did
Co-Host 1: Giles Male (00:48:57):
It
Host: Paul Barnhurst (00:48:57):
In the bottle.
Co-Host 2: Ian Schnoor (00:48:59):
That seems a little bit, where did you see a let formula?
Co-Host 1: Giles Male (00:49:03):
But
Co-Host 2: Ian Schnoor (00:49:03):
You know what
Co-Host 1: Giles Male (00:49:04):
Might, that's for the cash. So if you were going,
Co-Host 2: Ian Schnoor (00:49:07):
Didn't see that though. I'm curious
Co-Host 1: Giles Male (00:49:09):
Probably here.
Co-Host 2: Ian Schnoor (00:49:10):
Oh my gosh. Holy. That's what it did on your, wow, that's what it did on your, this just, wow.
Host: Paul Barnhurst (00:49:19):
Honestly that's that's not all that bad if you understand it.
Co-Host 2: Ian Schnoor (00:49:22):
No. Well it just completely proves and makes my point every single time. This is not how you want to build the revolver section on. There's so many issues with this. First of all, that should not be right on the financing section of a cashflow statement. And second of all, really, are you going to use a let function, one of the newer, more complex functions in Excel when your client or boss might just want to see a simple step-by-step add up of what you're doing? This is going to be hard to audit and understand. I'm not liking it even though it's probably right. Ja, you're going to say something.
Co-Host 1: Giles Male (00:49:56):
Well say it's really interesting. Obviously I agree. This is not how I model this at all. What we saw almost consistently last year was that a lot of that final cash balance or revolver balance stuff was solved in the financial statements. So you could argue if you're going to go down that route, I mean at least a let is trying to make the formula a bit more readable. But yeah, I mean I wouldn't do this.
Host: Paul Barnhurst (00:50:20):
Wait, it has a note there. What does the note say? Put a note on that cell. You didn't
Co-Host 1: Giles Male (00:50:25):
Put that there, right? No, that's a flag.
Host: Paul Barnhurst (00:50:30):
Oh, you add? Okay. I thought it was, was there an explanation note or something? I was going to be impressed.
Co-Host 1: Giles Male (00:50:35):
You
Co-Host 2: Ian Schnoor (00:50:35):
Got excited there Paul, for a second. You thought, hey it's,
Host: Paul Barnhurst (00:50:38):
I did. I was like, wow, it's now giving me explanations. So
Co-Host 1: Giles Male (00:50:41):
Other than that, obviously that is a big, I think area of concern. It's pulling from the schedules. I do, wait,
Host: Paul Barnhurst (00:50:49):
Didn't this hard code a zero up there, instead of pulling from a schedule down, down where there's all zeros. I thought that was a hard code
Co-Host 1: Giles Male (00:50:59):
Balancing
Host: Paul Barnhurst (00:50:59):
Balance sheet. Go up for a second. There was a keep going right there. One of those looked like it was all zeros.
Co-Host 1: Giles Male (00:51:06):
There is, well
Host: Paul Barnhurst (00:51:09):
I think it's the common shares issuance that was all zeroes, which
Co-Host 1: Giles Male (00:51:12):
Is fine. I think if you look at the output, this broadly aligns with what Paul's model saw. So you don't have a cash balance. You are drawing on the revolver year after year. Little bits look sensible, like the common shares don't change. The retained earnings is going up. You can see the amortisation on the senior debt. So that's coming down every period. That looks good. I think that's good. And then schedules wise, yeah, I mean again, I'm a fast kind, but I would much rather see flags and blocks, better blocks than this. It's pulling a lot directly from the assumptions, which again is very similar to what Paul saw, which is not my preference, but it looks, I mean on the face of it it's go back. But
Host: Paul Barnhurst (00:52:00):
You with skills give it to use the face, the fat nut face, the fast standards.
Co-Host 1: Giles Male (00:52:06):
Yeah, you absolutely could. But that all looks pretty consistent and what are my assumptions look like? So here's big improvement on your one. If this does work, I've got a scenario switch, I'm assuming if I change that to two it goes to best. How's it doing that? Exactly, exactly how I would do it. Index, you could use choose index, switch, whatever you like. Probably not switch but choose or index.
Host: Paul Barnhurst (00:52:32):
Yeah, you could. I wouldn't use switch. You could use a nested if Giles
Co-Host 1: Giles Male (00:52:38):
You could, but I think that's impressive. So it is got a proper scenario tool. I haven't got time to check everything. It's doing the same sort of thing with inflation here of various things. Linking back to an inflation rate formatted, this is very much in line with how half of the Excel financial modelling world formats. It's always either blue fill or yellow font. I think I'm pretty happy with that. It took a lot longer than the two models you ran, but I think that's looking pretty good.
Host: Paul Barnhurst (00:53:11):
So I think what I liked, what I like did a better job with its base best and worst case, no question. I think they were both similar on the other one area I do like on the assumptions page that mine did is it told me where it pulled it from the document and I didn't ask it to do that. It came from this cell or we used averages so I could at least trace it all out on the whole, I think this one followed a little more, some best practises. Although did it give us a check section like the other one did.
Co-Host 1: Giles Male (00:53:40):
I haven't seen one. And again, I think that's a huge plus point for yours. The fact, I mean I do have the balance sheet check. Yeah,
Host: Paul Barnhurst (00:53:47):
You have the one.
Co-Host 1: Giles Male (00:53:48):
I don't have anything else. I didn't think.
Host: Paul Barnhurst (00:53:50):
I probably maybe lean slightly yours so far, but there's plus and minuses in both of 'em
Co-Host 1: Giles Male (00:53:55):
I think because of the scenario. Exactly not that I'm, I'm not trying to defend my tool. The tool, not at
Host: Paul Barnhurst (00:54:03):
All.
Co-Host 1: Giles Male (00:54:03):
I agree with you. The scenario thing is probably a big thing because that's kind of fundamental to how do you actually build scenarios in a lot.
Host: Paul Barnhurst (00:54:10):
Agree. If there's a flaw in the one I did, it's the way it did. The scenarios is just a mess. That's a D minus type work, right? Just bad
Co-Host 1: Giles Male (00:54:21):
Two out of two balancing balance sheets even that's a step up from when we do you remember when we started months
Co-Host 2: Ian Schnoor (00:54:28):
Ago, they couldn't even build anything. I mean it was awful. It was horrible six months ago.
Host: Paul Barnhurst (00:54:34):
Remember one time nsaid, are we really going to show this? I'm embarrassed to do this. That tool has shut down so we could name it now we won't but to where we're at now. So it's a huge improvement. Alright, why don't you show us what you got. We'll spend a few minutes there and then we will wrap up.
Co-Host 2: Ian Schnoor (00:54:50):
So mine is definitely formatted the worst out of the three of yours, but there's something similar. Again, I used Opus 4.8 right in copilot and there's something similar feeling about our two trials. There's a switch over here. It's going into, well it's going into the switch is actually only going into sales prices and sales volumes. So here they are and you use it. Choose function. Okay, I like that. Very nice. I looked at these formulas. They're mostly fine. I mean this is 2026 and they're not using colour differences. It's okay, not super clear. If I go to the schedules then there's a sheet of schedules, not very long. It just has some very simple schedules with pricing, some cost assumptions, long formulas here. Look what I did for depreciation. It's okay. Let's see. I mean this is all good working capital. You do
Host: Paul Barnhurst (00:55:43):
A depreciation schedule. Did yours do one? Think that's the first one that actually did a schedule by year. Did your Giles
Co-Host 2: Ian Schnoor (00:55:51):
Mine built a summary sheet? I don't really understand. It doesn't look very good. It built a summary sheet for five years under these five items. So anyway, the best and worst cases are just dead numbers. So we need to realise that, that they're not actually adapting. They're dead numbers. It typed in, I don't know. I mean listen, I think it's off to a nice, the financial statements look fine. They're a little, they're not quite client ready yet, but they seem like they're okay. All these cells are links. I looked at it quickly before, it does look like after we ran it, I was looking at it. It does look like it's built some pretty good formulas. I
Host: Paul Barnhurst (00:56:33):
Like the cash available before
Co-Host 2: Ian Schnoor (00:56:35):
Revolver
Host: Paul Barnhurst (00:56:36):
That helper sell.
Co-Host 2: Ian Schnoor (00:56:37):
It's right, absolutely. But
Host: Paul Barnhurst (00:56:38):
It
Co-Host 2: Ian Schnoor (00:56:38):
Looks, I haven't found any dig error. I have only been looking at it for 10 minutes, but I haven't found any big errors. It's off to a nice start. Here's got the base best and worst and it's typed in dead numbers. There's no formulas here. So I don't know. I don't like that it is a base case, so I just have to trust that they're working base best. Worst. I just have to trust that they're working properly. I don't know. I would be happy to start to use this as my starting point. Sorry Charles,
Co-Host 1: Giles Male (00:57:08):
Isn't that fascinating though, which was kind of why we did this deliberately that you and I have both used the same underlying LLM is 4.8, but mine is directly through the Claude add in and yours is through the copilot add in and yet it must just be the additional context and whatever else is going on in that processing step for and after.
Host: Paul Barnhurst (00:57:31):
I think more than that, a lot of it also comes back to one, what skills it's using. So what skills does copilot have inside? What tools are it's using, how it's pulling the data. But the other, it speaks to the fact that this is probabilistic in nature and that's why. Have you guys heard the term AI a harness? No.
(00:57:52):
Giving AI a harness, it's basically, it's everything beyond the prompt, all the context you're giving a files and the Claude skills. They're basically saying because if you give a harness, if you're climbing or whatever you're doing, a harness restrains you and protects you. And it's kind of similar to that to ai. It needs that additional structure and guidance or we end up with these three very different versions. I'm sure if we all build a lot of skills, we could get much closer to each other between these three models. There'd still be ones that would be better and differences where we could close some of that gap. But that's a whole thing people are expected to learn. Now you got to know how to model really well, but now go ahead and figure out how to do AI really well and then figure out how to be a good auditor now instead of someone else auditing it. You got to audit the whole thing, not that you didn't audit before, but in a different way.
Co-Host 2: Ian Schnoor (00:58:41):
And you're going to have to know your
Host: Paul Barnhurst (00:58:42):
Own work. And auditing someone else's is very different.
Co-Host 2: Ian Schnoor (00:58:44):
It's very different. Our brains, for many of us that have been building things or it's a very different learning process in your brain to build and to fight the struggle. And the fight is important when you're, that's the creative process and a lot of people's learning is absorbed and it happens through osmosis during the struggle, during the fight of figuring it out. And that's important. So I know for me, that's how I often learn things by fighting and struggling and learning through the wins and the losses to get there. And then I'm very, very smart about a topic at the end of it. For me, if you gave me a fixed model and you just said learn it in the same level of intimacy and just review it, I don't know that my brain can actually get to the same level of confidence with it or that I can have the same level of deep understanding with it by trying to understand something now I'm going to have to because the way the world's going and we will, but we all have to find a way. You're right to look at this and still rip it apart, understand it, know what it's doing right, what it's doing wrong, and be able to defend it as if it's our own. And that's a bit of a big mind shift. I don't know if you agree with that, Giles, but that's a shift for me. Yeah,
Co-Host 1: Giles Male (00:59:56):
I do agree. And the more we look at this, I think we've all been on a journey since. Is this the wrap up session? Is it worth sharing? I feel like we're in the wrap-up session, Paul. Yeah. Alright, let's share our closing thoughts. I think back to when we started, and it was longer than six months ago and at that time I didn't,
Host: Paul Barnhurst (01:00:17):
It was a year ago now when we first approached this just, yeah, nine, 10 months. First episode we did was October. Really? Okay. Three weeks, which means we recorded it in September, which probably means we first discussed it in August, late August. We were planning year 10 months
Co-Host 1: Giles Male (01:00:34):
I think back to when we started. And this whole concept of agents that you can draw on within an app and it wasn't there. And that's why I think we started with third party tools. They were the only ones that had put that layer on. So the world has completely changed. I think we can see clear improvement. The fact that I think we just all had balancing balance sheets. Awesome. And the more we get into this, the more I'm convinced that focus, this is not a criticism of what we've just done, but the focus on can it build me a financial model so that I don't have to do any of the work? Just feels so it's not the pot of gold at the end of the rainbow for me. And that's partly now I am looking at that space and going, God, you've got teams up to their eyeballs and all this other stuff. Do you remember Ian Bennett talking about all the value you could get from all of the other areas of a modelling project? And I think I'm probably even more convinced that that is where the value is at the moment at least. And what figure all that stuff out. If you're a big company and then by the time you've done all of that, we'll be on Fable. We'll be on Mythos version nine and you won't have to get out of bed. Right?
Co-Host 2: Ian Schnoor (01:01:46):
You'll think it, you'll just, we'll have
Co-Host 1: Giles Male (01:01:49):
Think
Co-Host 2: Ian Schnoor (01:01:49):
It. You'll think something and it will develop and yeah.
Host: Paul Barnhurst (01:01:53):
That's a scary thought. If it can read my thoughts. Any last thoughts from you on this end? I think we got Giles dissertation. Let's move to you.
Co-Host 2: Ian Schnoor (01:02:01):
No, I agree. I agree with Giles. I think this is a different world than nine months ago when we started, when we didn't even, it was only at the end of the year. Remember it was at the end of last year when the LLMs announced that they were starting to incorporate, it wasn't even considered in the fall. We were looking at all these other tools and I haven't kept up with how they're all performing, but I suspect it's a challenge for many of them where they need to differentiate themselves. Obviously if they're going to survive now, people are desperately learning, going deeper, as you said into ai. They're trying to code, they're trying to learn skills, they're trying to kind of get under and think about it in terms of automating their day and their monthly close and their life and their emails and connecting it all together.
(01:02:45):
We weren't talking about anything like that six, nine months ago. And from what we're seeing today, obviously these three models that we built today, I don't know if there's a real winner. I mean they're all similar I would say, but they are much better than what we looked at nine months ago. But compared to the models from February, March a little bit, the balance sheets balance, which is nice, but we're still with all of them. None of these is ready to hand in. And that's even scary if people think that they have a models, right? Because you're missing the point of the learning process I think. And
Host: Paul Barnhurst (01:03:20):
Yeah, I don't think there's a market improvement like you said from Opus four seven to what we did today,
Co-Host 1: Giles Male (01:03:26):
But I'd be really intrigued if you two are up for it. I think we should do, whether it's the next episode or whatever, I think we should do it like, okay, let's give it our best shot. Let's actually put all of the custom instructions we would think of and skills files and whatever else. I'd be really intrigued to be like with our best efforts, how good could we get it straight away? But
Co-Host 2: Ian Schnoor (01:03:49):
That takes a lot of skill on our part to know modelling, to know accounting, and to know how to really be a strong musician with this tool to work it, to really work our models and get, that's going to take a lot of time and effort to get there. And I'm not saying we shouldn't, we should, but that's not free. That comes with investment, right? That's a real investment.
Co-Host 1: Giles Male (01:04:14):
We're just invoicing Paul for our time. We Isn't that
Co-Host 2: Ian Schnoor (01:04:17):
Correct? Right.
Host: Paul Barnhurst (01:04:18):
Hours and hours. I'm pulling it to Anthropic.
Co-Host 1: Giles Male (01:04:21):
Yeah, let's get Nico to paper for it. Were your closing thoughts Paul?
Host: Paul Barnhurst (01:04:24):
I think it's similar to the two of you. One thing I want to say is do you remember when we tested copilot online, right? That was an episode. Think how far we've come from that. A lot of people have badmouthed, copilot. I've been guilty of it in general. I think Claude's a better tool right now and I think most people do. That's pretty clear. But how far it's come, that's what I keep telling people is you can get value out of it. I don't care which one you're using. I think one message is you can get value out of all of 'em. Two, none of them on their own are production ready. Could you potentially get there with skills for certain tasks, for certain things? No doubt. Full financial model maybe depending on what you're doing. But even if you do, you still have to know how to audit it.
(01:05:09):
You still know I have to check it. I am just like the two of you. I say AI is a magnifier. The better you know what you're doing, the more you can get out of it. The one other thing I'm so sick of is the AI slop. We're seeing all over the place of do this in five minutes and I will repeat what I've been saying all along. Go ahead and build the model in five minutes, get fired in 10. It's just stupid If that's the way you think these tools work. It's not, no, there's not a magic easy answer to anything. And that's true with modelling, but you can get huge benefit if you're willing to pay the price, I think is the message.
Co-Host 2: Ian Schnoor (01:05:41):
Yep, I agree with that. Jan. Always great to see you.
Host: Paul Barnhurst (01:05:44):
Good to see all of you. That's a wrap. We'll do this again next month. So thank you for joining us everybody. We hope you made the whole hour with us.
