Host

Philippa Lamb

Guests

• Daniel Clark, business trainer and writer
• Julia Penny, Director, JS Penny; ICAEW past-President
• Jenny Blewett, Head of Finance, Costello Medical

Producer

Natalie Chisholm

Transcript

Philippa Lamb: Hello and welcome to Behind the Numbers. I’m Philippa Lamb. Today we’re talking about continuous learning around AI and tech. For accountants, getting the most out of these technologies involves more than simply training people how to use them: judgment, communication and granular accountancy skills all come into play. So how can firms identify their training needs, and what are the best ways to deliver that training?

Julia Penny: If we say, Oh yeah, I did a prompt… Oh yeah, I used it to help with Power BI… I helped use it with this… all of a sudden, that knowledge being shared around the whole firm, the whole team, the whole finance department, is hugely powerful.

Jenny Blewett: Literally asking it question after question after question – how do I get from here to here? – and it will structure its answer, and gradually the more you ask, the more it will tailor its answer to your needs.

PL: Jenny Blewett is here in the studio with me. She’s Head of Finance at Costello Medical. She’s here to share her experience of training teams to interact with tech and AI. We’re also joined remotely by Daniel Clark. He’s an accountant and a business trainer. He specialises in helping firms and professionals work effectively with fast-changing technology. And our third guest, also remote, is Julia Penny, a fellow trainer and a former President of ICAEW.

Hello, everyone. I thought we might make a start by asking you to draw on your collective experience of training accountants around tech, and I’d like to ask you, how aware do they tend to be of their own training needs? We know about the wins, and as you say, the wins can be great. But Daniel, I think it’s fair to say, accountants are often risk averse. And there are risks attached to using AI in the course of their work.

Daniel Clark: I don’t think this is unreasonable at all. We are trained to look at risk and be very conscious of it and think about this, and that’s absolutely right – that is part of our role. I think where you get an issue, certainly in relation to learning needs, is where the concerns about the risks actually put people off from learning more about AI, because there’s absolutely a learning need there. But my general experience is that a lot of accountants, obviously, are very aware of AI – It’s impossible not to be, it’s talked about all the time – but there’s a lot of people at the stage of “I’m not really too sure how this applies to me, what benefits I can get from it.” A lot of work is needed to start at that level and explain the concepts behind AI and how it can be used, and potentially ways of getting started and experimenting.

PL: As you say, there’s a lot of specifics around AI and tech training, aren’t there? But Julia, I know your experience of training is very broad. Presumably it’s the same, in the sense that some people have more of a learning mindset than others?

Julia Penny: Oh, absolutely. Some people will just know the whole time: gosh, I must learn about this, this is a new thing, this is coming up, and keep their eye on the horizon as well as what’s happening today. But other people might focus more with their CPD needs. They know they have to do continuing professional development, but they think of it in traditional terms: auditing, accounting, tax. And actually today, while you need that, if those are areas that you’re covering, you need to keep up with the technology. You need to look at the opportunities, so almost be ahead of the technology. I think that openness of attitude is absolutely vital if accountants are going to get the most out of their CPD.

DC: Just to emphasise that point – and maybe at risk of stating the obvious for an audience of accountants – continuing professional development is part of being a professional, it’s part of what we do. That’s why it’s a requirement of the Institute and of every other accounting institute as well. There’s actually an ethical principle of professional competence and due care. So, from every point of view, it’s something that absolutely should be in our culture, and often is.

Jenny Blewett: I think there’s a huge increase in awareness needed. It’s understanding the benefits you can gain by really understanding technology. The main thing I’ve seen is massive, massive efficiency savings – by doing clever reports, by implementing scripting, by using AI. In the NHS we were able to save weeks out of our month-end schedules. Even working in a very tech-savvy environment, I’ve still taken eight hours out of our month-end schedules at this stage. Not only am I taking them out, but I’m seeing my team take that time out of the schedules as well. We’ll cut a task from three hours to three minutes by really thinking about anything repetitive and removing it.

PL: I know, Daniel, that you feel to get the most out of AI systems, tech systems, you do need to teach – or at least, refresh – allied skills. I’m thinking about things like relationship building, communication…

DC: Yes, I think the headline here is that a certain amount of tasks are going to be made much more efficient, as Jenny has said. There are things we can do much more quickly with automation, with AI and so on, particularly where something is very repetitive or very rules-based or very structured in its nature. Those tasks are increasingly going to be done by technology. So what that means, in my view, is that accountants need to be focusing much more on other human skills that technology can’t do. It’s all the good stuff that we do at the moment. It’s things like building relationships, communication, explaining things to your clients or to your stakeholders in ways that are meaningful to them and that they understand, building trust in those relationships. It’s all stuff we do now, and in my view it’s only going to become even more important, because that’s where we’re going to add more and more of our value, as opposed to perhaps, ‘I know the detail of the rules around accounting’ or ‘I can run this particular process.’ That perhaps is going to be less value-added going forward.

PL: I’m also thinking around… Accountants getting their arms around AI and tech, that’s a collaborative learning process, isn’t it? Because it’s not as straightforward as learning some other core accountancy skills, and so the capacity to interact with others, talk about your experience, your difficulties – that’s all part of it, isn’t it?

JB: Yes, I’d say that was part of it. I suppose a nice way to venture into using AI if you’re unfamiliar with it and you don’t want to use it on large models or train it or do anything massively complicated… I came with an example because I was thinking, how do you embrace AI if you’ve never used it? I had a BACS file out of the bank in CSV format. It’s awful, and accountants will be used to that – we do a lot of work in Excel. And you can put the structure into AI and say, pull out what I need, pull out the account number, pull out the sort code, and give me the Excel formulas to do all of that. And then if you want to go a stage further, you can say, give me the Visual Basic or Office Script to do that – because I do script it – and it will do all of that for you, so that you can very quickly transform data, and it’s teaching you how to do it.

Another really nice use I found this week was I was playing with some pivot tables and adding slices, which I’ve done a million times, but I wanted to apply the slices, which is how you cut and slice data effectively. I wanted to apply it to three pivot tables at once. So I was like, can I do this? And AI is the place to go to find the answer.

Similarly, my other area is Power BI. When you want to show really effective management information, and you might not be familiar with everything you can do in all the various things like Power BI or Tableau, AI can take you from Excel to that. Literally asking it question after question after question – how do I get from here to here? – and it will structure its answer, and gradually, the more you ask, the more it will tailor its answer to your needs.

PL: I guess that’s what I’m getting at: this idea that you’re interrogating the technology, but it’s also about interrogating colleagues and others, isn’t it, Julia. Do you understand what I mean?

JP: Absolutely. I think what Jenny’s highlighting is that you’ve obviously got really good grasp of a lot of these things, you’ve been working on it, doing all sorts of things, but within a firm, within a finance team, wherever you’re working, there’ll be some accountants who are happier with the whole thing, some accountants who will have tried one thing with AI, some will have tried another thing. And actually, if we share this knowledge, we really sort of super-charge our ability to learn more things.

Even on simple things. I mean, our phones have huge amounts of computing power, even today. And sometimes you find that you’re sharing little tips with colleagues on something really basic. But now we’ve got this whole AI landscape that we can explore. And if we say, Oh yeah, I did a prompt… Oh yeah, I used it to help with Power BI… I helped use it with this… all of a sudden, that knowledge being shared around the whole firm, the whole team, the whole finance department, is hugely powerful and a really good way to leverage our CPD.

PL: And obviously you can share this stuff digitally. We all do, we all use a variety of platforms to talk to our colleagues about stuff. Is there a potential loss, off the back of… I’m thinking about hybrid working. This is sort of stuff you used to pick up, perhaps more easily, by having someone sitting next to you.

JP: If I can pick up on that, because I do think that the hybrid working, it has great advantages. I’ve actually hybrid worked my entire career, apart from the first three, four years…

PL: Yes, me too.

JP: …so I’m really used to it, and I’m really used to working in that environment. But for a lot of young people particularly, as you say, how do you pick things up when you’re not used to seeing colleagues around? And you can counteract it: you can have things like Teams rooms open or Zoom rooms open, and have discussions to make use of chat facilities. So you can do it, but it’s a different way of working, and you need to ensure that you’ve built that ability to communicate, to collaborate, even if you’re doing it remotely. And maybe think about, actually we should get together every so often, and maybe we have a day or two or whatever when we really concentrate on team building and team sharing.

So different things will work for different people and different teams, but I think we really need to be cognisant of the challenges and also the advantages of things like hybrid working.

JB: I think I’m really lucky in the environment I’m working in where I feel we’ve got it right. We have an AI taskforce with people who are super-excited about AI and implementing AI that come together regularly to discuss the latest developments. We have a prompt library, so that we are efficient in how we use AI. But then within just our finance team meetings, we always cover what CPD someone’s done. And it doesn’t have to be AI, it doesn’t have to be tech, it can be soft skills, but we always bring that to every single meeting and ask someone to share a development. And we also make sure that the teams have time specifically to go and attend a webinar or to develop themselves. Particularly with all the ICAEW resources that are now available – the webinars are incredible that Ian Pay is currently publishing.

PL: I’m wondering, Daniel, if there aren’t some misconceptions still floating around about who’s good at this stuff. Getting back to the thing about learning, specifically how to interact with AI and tech. There’s a trope, isn’t there, that younger people are better at it? Have you found that to be true?

DC: Not really. The whole debate around digital natives is a very, very interesting one. I’m not going to go into the detail here, but…

PL: Well, no, give us your thoughts, because I think it’s relevant here, isn’t it?

DC: There’s been quite a lot of research on this, and I think the headlines are, firstly, there is absolutely no evidence of a big cut-off around 2000 or whenever it was, that people before and after are fundamentally different. That’s just not the case at all. And while it is the case that clearly people who’ve grown up in the last 20 years have grown up in an environment of very different technology to what those of us of a slightly more vintage generation have grown up with, there isn’t really any evidence that young people as a group inherently get technology or can learn how to use new things or understand these concepts in a fundamentally different way to older people.

PL: But do you think that trope plays into… perhaps older people… I hesitate to use this phrase, but non-millennials – does it play into their idea that perhaps this is going to be harder than it actually is?

DC: I think so. I think there’s an awful lot… Julia’s already mentioned the learning mindset, and the learning mindset is something anyone of any age can have, and it’s what’s critical here. It can be very self-limiting to think, “Oh well, I don’t understand this stuff because it’s all new.” AI can be an incredibly complex and technical area, but actually, in terms of what you need to understand to use it effectively, and how you can use it effectively, I don’t think is that complex. I think it is something that that anyone can learn.

At times there’s even a feeling in some firms that the young graduates coming in now, they’re the ones who will sort this out for us because they all understand this stuff, and they’ll all do loads of projects and that will sort everything. This is totally misguided. Apart from anything else, the graduates coming in don’t have the business knowledge to do that. There’s no reason why they should, because they’re right at the start of their career.

But there’s definitely scope… I know some companies do things like reverse mentoring… there’s definitely scope for collaboration between different groups with different outlooks, different skill sets, to come to some conclusions here. If someone happens to understand, let’s say, loads and loads about AI, that’s not necessarily going to make them useful to an accounting firm, unless someone with business knowledge can help them make it useful. Clearly, if you have people who have really good business knowledge and have very good technology skills, fantastic. Those people are very in demand and that’s a good place to be.

PL: What’s your experience, Jenny, in terms of different age groups and their facility to get their arms around working with what is obviously very fast-evolving tech?

JB: I suppose my experience has been directly with coaching teams on all sorts of tech, and I’ve found it’s not about the age of the individual, it’s about the way their brain works, so I will tend to focus people towards their strengths. That’s how I work with my team. And some of them are in team management, and some of them are in tech, and that’s what we’re very focused on internally. I’ve learned from people that are younger than me, watching them, how they prompt, but equally I’ve learned from people older than me, watching how they might code something. So it’s really about the brain type, not about the age.

PL: Daniel, I did want to ask you – you mentioned in your earlier answer about the necessity of understanding the business environment, and tech skills are not going to be enough for anyone to be effective in the workplace, in this context. I know you have a concern about fundamental accounting skills and principles fading a bit into the background in terms of learning and development?

DC: Yes, and this is very much a concern within the profession, and people I speak to in the profession – and this predates AI, but AI perhaps super-charged this trend – but as more tasks become automated, you have software that will produce your accounts and do your reconciliations and stuff like that; there’s a contrast there to what junior members of staff do compared with 20, 30 years ago, when they might have sat there with a spreadsheet, or even with receipts, and put together a set of accounts. It was a painful and tedious task, but you came out of that really understanding how a set of accounts was put together and what makes them up, and understanding the real detail in that. There is a concern that with some of those routine tasks, or many of those routine tasks being automated, some of that is being lost.

It’s definitely something the profession is focused on. It’s definitely something ICAEW is focused on, is ensuring that through the qualification and through the work experience that people get in the first few years in their career, that they do develop that fundamental understanding of how accounts are put together, how debits and credits work, how accounting works, because without that you’re not going to be a good accountant, no matter how good you are at some of the other areas.

PL: Julia, the importance of old-school skills?

JP: Well, I think that basic understanding of what makes up a set of accounts, the debits, the credits and so on, is absolutely vital. But we’re moving very rapidly to a world in which accountants, when they come in, they’re not going to spend years just doing that basic, if you like, bookkeeping or basic accounts prep stuff. We need to be imaginative as to how do we accelerate it so that they get onto the really interesting stuff really quickly? Even things like gamification of some of those sort of things, those basics… I don’t know, maybe we’ll start seeing virtual reality or augmented reality used to help train up students really quickly, so that they can come in at that more senior level very, very quickly, like a lot of graduates going into other jobs in other organisations expect to start being in meetings with clients really, really soon after starting their work there. I think we’re getting rid of the – dare I say it – drudge work of accounting, but we need to bridge the gap so that when you’re going in at that higher level, you can also understand what’s behind all of the numbers.

PL: It’s mindset, isn’t it, that knowledge gap there. The understanding of how things work in the business environment.

JB: I think it’s still super-key having the ACA qualification. I see that day to day in business that the ACA qualified accountants will know what a debit or credit is. They won’t do silly things like put income to expenditure and put things in the wrong way around which fundamentally flaw everything.

PL: I’m interested to talk about delivery, actually, how to best deliver training. You just alluded to it, Julia – gamification and automation of all sorts of forms. What do you think, Jenny? You talked about webinars earlier. Do you watch them as a group?

JB: We do a mix of different things. I send out webinars to my team. I particularly like a batch of webinars from a lady called Leila Gharani. When I first joined my team recently, I gave them a library of resources that I would recommend they watch when they have time. Equally, we will take up the training firms when they’re offering webinars, or the ICAEW when they’re watching webinars, and incentivise people, sometimes with a free lunch, to come along and we’ll all sit together, watch the webinar, and then we can talk about our learnings from it. If someone finds a particularly good one, we’re always encouraging our team to then share that so the wider team watches it.

PL: So, shared experience. Good idea?

JB: Absolutely.

JP: Absolutely, it’s a good idea. But I also like… because I’m working on my own a lot of the time, I love YouTube – I hate to say it – or things of that nature. If I think, oh, there must be a way to do this, and I don’t quite know what it is, I’ll YouTube it, and I’ll get a little video saying, yeah, you do this, you do that, you do that. Things like, Jenny’s been talking about prompts and prompt libraries. For a lot of our listeners, they might not even be familiar with the term ‘prompt’, because those are the questions we’re asking the AI, can you do this, can you do that for me? And we’re prompting it to give us a response.

But there’s loads of…

PL: So, drafting a question to it, for people who haven’t done this.

JP: Yes. There’s loads of educational material out there, in little snippety things as well as… certainly the ICAEW has a ton of webinars and face-to-face type training things, which, yes, I do a lot of those as well.

DC: But I think what’s critical is, there are challenges. Technology is moving very fast and so on, but we are in an absolutely golden age of educational and training material. There is so much of it out there, more than anyone could possibly get their heads around, in whatever format you like. So the opportunity is there for everyone to construct their own portfolio of learning resources, which will depend on what you’re doing, what skills you want to learn, what your style is. Do you like watching videos? Do you like listening to podcasts? Do you like reading books? Do you like reading newsletters? Are there people whose style you particularly like?

Based on all of those things, you can tailor a very personalised journey. And this is what I would encourage every professional to do is make sure that they have a strategy for how they develop their learning over time.

PL: So curating your own learning, obviously, as an individual, as you say, it’s great and it’s empowering, but there are quality issues there, aren’t there, for organisations to worry about?

DC: There absolutely are, and that’s part of the curation process. Some are better than others, but I think judging that, judging what is actually good quality, what is worth using, that’s part of professional judgement, and that’s part of sort of how we assemble those things. And the team element, as Jenny talked about as well, if you’re lucky enough to work for an organisation like that, of course, that’s a key part of it as well. All of these are part of the portfolio.

JB: I love that AI constructs your learning for you. You could say, I’m at this point in my career and I want to get to this point in my career, and it can literally give you a path. And it writes an enormous path. You can ignore it or use it, but it’s interesting.

PL: So you’re using AI to tell you how to how to use AI, how to construct your journey through AI?

JB: I absolutely am. When I was looking into this role and thinking about my journey and where I wanted to end up, it was part of that process. Like thinking about where I was coming from and where I wanted to be.

PL: It is great in the sense that it completely personalises that journey you’ve just talked about, Daniel, doesn’t it? Whatever works for you is the journey it will give you.

DC: Absolutely. And by the way, AI can be part of that journey too. One of the things these LLMs [large language models] are very good at, on the whole, is explainers. If there’s a particular topic you want explained at a particular level, or using a particular example, in a particular context, it’s normally very, very good at doing that. And then you can ask it questions. You can go back and forth and drill down on that. And it’s often a very good way to develop your understanding of a topic.

JB: And it’s got so super-… the user interface is incredible now. And if you’re really at speed and you’ve not got confidential material, you can just photograph something and then ask it a question about the photograph, and it will read it and provide the answers. It’s crazy.

PL: I mean, just playing with AI, it seemed to me – so when I was starting to learn about it myself, literally just playing with it, and having the confidence to do that and asking it… and it’s very fun, isn’t it? And you don’t have to ask it stuff about work either, do you, to understand how the interface works? You can ask it to plan your holiday for you.

JB: Exactly. That was the example I was going to use. I was lucky enough to go to Singapore recently with work, and I asked it to plan me a full itinerary. And it did. It was an incredible itinerary. I can vouch for its quality, I thoroughly enjoyed it.

PL: Yes. Julia, do approve of this? I mean, it is very self-directed learning, isn’t it? But it kind of works, doesn’t it?

JP: Yes, it definitely works. And I think for those accountants who are sitting there thinking, oh gosh, but AI… um… er… risk and aversion and whatever: firstly, do remember professional ethics, because things like… okay, let’s bring that scepticism into mind – is this telling me something sensible? And if I’m going to put something on AI, is it a private AI, so a Copilot within your organisation, for instance, or is it ChatGPT where I’m giving away whatever I’m putting it on? As Jenny said, don’t put private information or a private photo or whatever there.

A quick thought about your professional ethics, as Daniel was saying earlier, allows you to put everything into context and say, okay, yeah, this is reasonable, this is not reasonable. Hey, yeah, that’s a good idea. And you can try things out, you can be taken on that journey, but you can drive that journey yourself as well, by deciding what you do and don’t do among what is suggested in the AI. Fantastic opportunities for learning.

PL: Yes, and as you say, completely resting on all the training you’ve had as an accountant, so you have a clear understanding of what’s feasible and what’s right in terms of the answer you’re getting. Because obviously, as we all know, AI can tell you nonsense.

JB: It can. I was about to say, actually, one of the things we always need to be mindful of – we say it in everything we do – it’s hallucinations, where it does just tell you something that’s completely wrong, but with utter confidence in how it’s describing it.

PL: And you need to know that as an accountant, don’t you?

JB: Absolutely.

DC: You do.

PL: You absolutely need to know that. So this is where, of course, the jeopardy comes in. I’m also thinking about leadership roles here, in terms of helping people feel confident to learn how to integrate this into their working day. How does the management team model that in an authentic way?

JB: I suppose I’m really lucky where I am. We have our AI leaders, and they’re going around our organisation and doing demonstrations and having conversations with the team about what’s relevant to them and how they can implement it in their individual functions. And then we have KPIs on AI usage as well within the organization, to see how effectively that’s working.

PL: This all sounds great, but I have to say most organisations are not at all like Jenny’s. How might that work in perhaps… I would describe it as a more mainstream organisation?

JP: I think, in terms of how you deal with that leadership aspect of learning about AI, implementing AI, when you’ve got, if you like, a more traditional outlook on everything tech-related, it can be a bit of a challenge, because you do have to get that top end, that management, that leadership team, to be on board with it. If they’re hesitant, if they’re very risk averse, if they’re like, oh my goodness, oh yeah, I don’t know, then it’s really going to hold back that finance team, really going to hold back that firm if it’s in practice.

I think it is up to that leadership group, if it’s a group, or the individual if it’s an individual, to really do some self-reflection and say, okay, I need to get confident enough, not to necessarily do all this myself, but to lead my team, to allow them to explore, to tell me what could be done, what can be done, how we can take the firm, the finance department, the process, whatever it is that you’re looking at, to the next level.

So I think, that’s a leadership skill. It’s not necessarily the AI skill itself. And I think that’s really important – to be cognisant of the need to do that, for all of the leaders out there.

PL: Do you agree, Daniel, that the training journey starts at the top?

DC: It certainly needs to involve the top. It can start different places. But it’s a very well-established principle of any change management that it’s not going to work unless you have buy-in from the top. And I think the type of understanding that a leadership team will need is, what are the opportunities – and risks – around AI? What can this do for my business, and where can we take things forward? And that allows them to set an example, to role model how they’re using technology, including AI, to set parameters, incentivise people to work on this, but at the same time put in place the guard rails and determine what is the risk that the organisation can take. That’s the role of the leadership team. And calibrating that is important – that you’re not taking reckless, unnecessary risks. But equally, just saying this is too risky, we’re not going to touch it, is really not an option.

PL: We’ve covered a lot of ground here, it’s quite a sprawling topic. Now, what is helpful to know is, I think you’ve all been involved with ICAEW in producing resources that members can use. Do you want to just run us through what’s there for members?

DC: There’s a lot. There’s a dedicated section of the ICAEW website on AI. I think a good place to start is with the Guide to Generative AI, which is a web resource that was put together a little while ago and is continually updated. That’s a reasonably short, very accessible, modular piece of guidance that I think lays out what everyone needs to know when they’re using it, including things like potential business uses, risks, legal issues, all of that stuff. So that’s a very good place to start. And then there are quite a lot of webinars, interviews, recorded resources and so on. Because everyone’s journey is different, I think it is a matter of dipping in and seeing what’s there and seeing what you think is going to be most useful for you. But there’s, yes, there are a lot of resources there.

JB: I’m going to plug one I was involved in, and it just gives accountants eight use cases for AI, so it just gives you an in-road to think about what you can do.

PL: Julia, have you got a favourite?

JP: Not a specific favourite, but more of a general point. I get daily emails from ICAEW and it shows me the training that’s coming up, what there is, the webinars, the factsheets. And I can pick and choose and think, oh yes, I should look at that… listen to this… watch that. And so my plea would be to make sure that in your preference centre, which is your membership dashboard, if you like, where you say what you’re interested in and how often you want to be communicated with by ICAEW, you’ve ticked all the right boxes, so that you’re hearing about this wealth of resources included in your membership. Don’t think, miss it out. Obviously, there are always things for people that need to go that much further, where they might want to pay extra for specific courses, but there’s a ton of stuff, which is all included in the membership. So make sure you hear about it would be my advice.

PL: Daniel, Julia, Jenny, thank you all very much indeed for taking the time to be with us today. We’ll link to all the resources that you’ve mentioned in the show notes for the episode. Obviously, listeners can also visit icaew.com/ai to get started on that learning journey we were talking about.

Behind the Numbers will be back in late February. Before then, join us for Accountancy Insights covering, as usual, the key news and updates you need to know. Thanks for listening.

17:55

Does this make sense?

Otter had ‘training fans’

Should we expand? [large language models]