Summary of the Interview
In this interview, Philip White, Founder of Audacia, a 100 person technology consultancy based in Leeds specialising in data, AI, software engineering and cloud, speaks with David Savage, Tech Evangelist at Harvey Nash. Together, they explore what AI skills really mean for organisations today, how businesses should approach AI adoption, and what the rapid pace of technological change means for workers, regions and the future of the tech industry.
Philip challenges the idea that “AI skills” are a single, monolithic capability. Instead, he highlights that AI is a multifaceted technology and that many organisations rush toward AI without understanding what it is, what it can do and how it applies to their business. He argues that successful AI adoption depends far more on strong delivery capability, business analysis, testing, data quality and security awareness than on the latest tools alone. These traditional foundations, he says, are what separate productive AI projects from failed proofs of concept.
A recurring theme is the need for organisations to decide where they want to sit on the innovation spectrum: do they simply want to be adopters using off the shelf tools, or do they want to become genuine innovators using their own processes, data and culture to build transformative AI solutions? Philip notes that misalignment across leadership teams is one of the biggest blockers to progress, as organisations often pursue AI for different, sometimes conflicting reasons.
The conversation also explores the human impact of AI. Philip acknowledges that AI will replace certain tasks and may reshape the job market, particularly affecting junior roles that traditionally serve as stepping stones to senior positions. There is uncertainty, he says, about whether society will demand more output as AI accelerates productivity or whether it will simply reduce demand for human work. What is clear is that organisations must communicate openly with employees so that AI is not viewed solely as a threat, but as a tool that can enhance skills and create new opportunities. Transparency is essential to reduce fear and build trust.
The discussion also touches on AI growth zones, particularly Leeds’ ambition to become one. Philip believes there is real opportunity for the region, provided investment sticks locally, talent is retained and academia, business and community work collaboratively.
For individuals thinking about their careers, Philip suggests focusing on the application of AI rather than solely on tools or theory. Understanding what AI is good at, how it solves problems and how it integrates into processes will be the most valuable capability in the next 18 months and beyond. Experience and interpersonal skills will also remain critical in a world where not everything can or should be automated.
Key Themes Discussed
- What “AI skills” actually mean and why delivery, data and culture matter more than hype
- Why organisations must understand their goals before adopting AI
- The importance of experimentation, agile practices and failing fast
- Risks to junior talent and future workforce structures
- Leeds’ potential as an AI growth zone and how to retain value locally
- The role of employers in supporting workforce readiness
- The shift from experience to aptitude and the value of interpersonal skills
- How individuals can position themselves for future career success
Tech Flix 6 documentary
This interview is part of Harvey Nash’s latest Tech Flix documentary, which explores the AI paradox: AI is scaling, skills are not. The film examines how AI is transforming work, education and regional economies, while asking a fundamental question:
Are we preparing people for the future, or leaving them behind?
Watch the full Tech Flix documentary here.
Full Transcript
David Savage 00:00
First of all, thank you for giving us some of your time this afternoon and inviting us into your office.
0 mins 5 seconds
Do you want to just start by telling us a little bit about Audacia and what your role is here?
Philip White 00:11
So, I'm the founder of Audacia, we're a technology consultancy based here in Leeds, team of about 100 people and we specialise in skills across data, AI, software engineering and cloud.
David Savage 00:23
Not always based here in Leeds though, kind of an arrival to the city having been based in the southeast or were you always in Leeds? Always in Leeds. Always in Leeds. So you had a...
Philip White 00:33
So we did have, for a period of time, a London-based office. Right, now a lot of our London work is on-site.
David Savage 00:38
What about Leeds is good for a tech startup?
Philip White 00:41
I think it is the community. I think Leeds has always been a place where any kind of advice helps support. I don't know a single occasion where I've gone to another technology business in Leeds and they've not spared an hour or two for a coffee to share thoughts.
David Savage 00:58
Are you a proud Yorkshireman? Is this through biased lenses?
Philip White 01:00
I think I am a proud Yorkshire. I'm from further north originally. I'm from up near Durham.Yarm. Oh yeah, I'm from any castle, so. Oh, there you go. So halfway between the two. So I'm like an honorary New York.
David Savage 01:12
So you're not overly biased, but you still think Leeds has something special.
Philip White 01:16
Yes.
David Savage 01:17
Well, we're here to talk about AI skills. AI skills feels really vague, especially when we talk about an 82% rise in the level of demand for AI skills, as our report does.So from your perspective, as the leader of a digital tech organization,
1 minute 38
what capabilities should we be really talking about within that very broad term of AI skills?
Philip White 01:45
I think there's a big issue at the moment in people going out and assuming AI is one thing. AI is a multifaceted technology, and I think there's a combination of the drive to AI being a big fear of missing out, coupled with a big push from vendors, making organizations go we need to get on this AI bandwagon without actually realizing what it is, what it can achieve, what it can do. So I think in terms of the skills that you need, I think organizations probably need to slow down rather than speed up.I think they need to get a set of people in a given organization to tackle AI. I think they need to understand what AI is, what it's not, what it can do, what it cannot. I think a lot of organizations are going, let's buy a language model product, tick an AI box, we've got some workforce improvement over there. And then at the end, it's following the same risk we've got. It's exactly the same problem we had with things like no-code platforms, where people are dismissing all the things that go around, all the stuff that goes around the outside of a technology project, the management, the testing, the planning, the business analysis. It's exactly the same problem.
2 mins 51
So I think the big thing organizations need to do is step back, look at what it can do, come up with a plan for how to, once you've understand the application of AI, the applied AI, then you pick up the skills. The key skills are delivery. When we look at successful and not successful projects, it's in AI, it's rarely technology, it's predominantly in delivery, so good delivery mechanisms, some old-school, classic, agile delivery skills, deliver something fast, fail fast, work collaboratively, release incrementally.Strong business analysts, I think they are key skills at the moment, understanding, we're going to use this to automate a process, what's your current process, not really sure. So some really solid skills around BA, secondly. I think three, testing is going to be a key one, especially for big language model-based solutions. They're really hard to create test cases for reproducible test cases. And then I think four, it's organizations need to have a key focus around security. So I think just having somebody in the organization that is capable of making sure that you are not introducing potential security risks.
David Savage 04:02
Listening to you there, that all sounds quite traditional in a quite reassuring way. I think in the industry, the dialogue seems to be pushing away from traditional hard tech skills and listening to you there sounds like I'm talking to someone perhaps five years ago and I mean that as a compliment in a positive way.Is that a fair reflection?
Philip White 04:23
Yeah, absolutely. AI is just a technology. It's an exceptionally powerful technology, but if you're looking to implement it, you should implement it just as you would in another technology.
David Savage 04:35
So when you talk about organisations and those that perhaps aren't thinking about it in slightly more traditional hard skills within the tech industry and therefore creating a narrative perhaps of something that's a bit mysterious or a bit unknown, what kind of organisations are we perhaps referring to or is it not one style of organisation?
Philip White 04:58
Now, so, I mean, if you look at our customer base, so we've got, you know, FTSE 100 corporates that we're working with, we've got public sector health. So we have a broad range of customers tackling AI and, you know, yeah, the key, so it's the delivery and the data, I think, is the, you know, people with good data are getting ahead and people with good, well, people, data and culture, sorry, yeah, the delivery data and culture are the three things, I think, that make the big driving force.If you know how to deliver a project, well, you will deliver it. If you have good data, that's where, yes, using an off-the-shelf, it's the same with any technology, if you're using an off-the-shelf package, you're at the same starting point as everybody else. If you've got loads of proprietary information, AI loves data. If you can use that to, you know, what are we trying to do? You're trying to do things better or faster or safer with AI. And if you've got your own data, you can really hone in on getting the most out of AI.
David Savage 05:53
But there are some organizations out there who don't understand this and therefore are creating a bit of a narrative problem.
Philip White 05:59
Yeah, absolutely. I think it's this, because there's a lack of understanding, I think organizations need to decide where they want to sit in AI world first. Do they want to be innovators or adopters? I think this is the focus.What are you trying to do? Do you just want to make sure that you are using AI products? Then going out and buying some co-pilot licenses will tick that box and find some work for better emails getting written or some sense of automation? Or do you want to show we are going to be cutting edge in this space? We've got loads of really good skills. We've got loads of really good processes. We've got loads of really good data. We can use this to create some super powered AI technology.
David Savage 06:35
I think we're getting at something quite interesting there because our research when I say our research Nash squared says that 50 percent of the organisations that we talk to are not investing in their staff. And when I talk to people at events and ask why it's this we don't really know what to invest in is it perhaps not that's not the right question that's not the right answer perhaps.And that your point that you have to decide what kind of a user or a doctor or innovator that that first step is getting missed.
Philip White 07:06
Absolutely, but where do you want to be on the innovation spectrum and what are you looking to achieve? What's the outcome you want?Quite often we'll see dozens of people, so we'll day-to-day go from boardrooms doing AI strategy, and peeling under the surface, the one thing you find is everybody will be, we need to do AI, but when you peel it back, quite often when I say AI to you, what do you mean? Well, I mean co-pilot. Well, I mean, it's like we've done linear regression over here. Getting consensus on the technology is a challenge. Getting consensus on where they want to be on that innovation spectrum. And then three, having a perception of why we're doing AI. You know, a lot of people will sit in this innovation empowerment. Imagine what we can do. Imagine the services we can give. And then on the other end of the spectrum, in the same room, people have agreed that they need to do AI for six months. We'll suddenly turn around and go, oh no, that's not what I thought we would do. I thought we were automating processes to cut heads. So getting that alignment is critical as well, first.
David Savage 08:00
Who do we need in the room to help get that alignment? Who are the voices that give clarity to that whole kind of process and discussion?
Philip White 08:12
I mean, it's too easy to say us, but people with experience delivering AI projects, I think it's people who have that experience, because obviously the big risk at the moment is a lot of AI proof of concepts and not a lot of AI in production.People with experience getting a concept, a use case, through validation, through the technical choice, technical implementation, out and in production and being used in the wild.
David Savage 08:39
Are there any voices around board tables that you think are unhelpful when it comes to making sense of what AI means to an organization and therefore what it means to the staff in terms of skills?
Philip White 08:52
That's a great question. Now, I don't think that, I mean, speaking anecdotally from personal experience, I don't think we've met any specific roles. Definitely met personalities that are a problem. Definitely where there's a view of, if I look at projects we've run, where we're looking to improve and do things more efficiently and there's been this, you know, really powerful human in the loop augmentation approach to AI, that works really well. I think if you start with an end in mind and that we're replacing these 30 roles with AI robots, those, you know, you need to have total completeness in a process run by AI before you can even think about doing that. They tend to be less successful.Definitely, so a classic issue we see from people is if there's a mindset that it should be delivered as a kind of a classic waterfall project, I need to see, so two really good examples of approaches. I need to see, here's a budget, we're gonna spend half a million pound on machine learning and AI and language models. What's the ROI going to be? And the technology, it just doesn't work that way. You know, this mindset of, this goes back to this agile concept. Organizations that are genuinely agile are winning at AI, not the ones that buy the best AI technology.People that have the mindset of, have a use case. I want to look at three, because obviously AI is just a big selection of ice cream flavors. You know, there's so many different ways you can use, there's so many different applications and tools within the AI set of toys. Which one you use is gonna decide the outcome of success potentially. So if you take your use cases, what we might do in a proof of concept world is try and solve the same use case with different AI technologies. And you then play survival of the fittest at each gate. That trial and error mentality that's quite common in agile based organizations feels very strange to someone that's releasing funding and once guaranteed ROI early.
David Savage 10:49
Listening to you, I can't help but feel that there is a situation when you have an organization that therefore is agile in its approach and is fluid and flexible. And AI exists properly alongside people and it is not this thing that individuals, employees should fear.And yet so many organizations, and I'm talking very anecdotally here, friends that I'm talking to at the minute, across the board on making redundancies and the word that people who sit outside the technology industry keep coming back to when they talk about those redundancies is AI efficiency, et cetera. How do we get to a point where AI isn't seen as a negative, generally speaking, but is seen as something a little bit more hopeful, I think, which is what you're painting?
Philip White 11:35
Well I don't know if we can say it's not going to be negative, I don't think we know yet, I don't think we know the impact. We're seeing the way we develop software potentially taking half the time. I don't know if the market will then take twice as much software or there will be half as much demand. So I think there's a big uncertainty around what the impact of being able to do things faster will be.
David Savage 11:56
Okay, if I play devil's advocate for a minute, and I cast AI as a negative on jobs and opportunity for people, and that leads to a reduction in the number of opportunities, particularly a junior level for an organization, has the leadership of many organizations, and I know this is very difficult for you to answer because obviously I'm asking you to talk for a lot of other organizations and that's hard to do, but do you think the leadership of organizations has really thought that through in terms of the long-term implications that it would have for not just the number of talented individuals in the market, but the shape of their very organizations?
Philip White 12:39
Again, anecdotally, I don't think we necessarily know exactly what's coming. I think you talk about the triangle to diamond organisation.So is it going to be... So we are seeing that starting now. So if we're looking at tasks that junior people are doing are being replaced potentially by technology, the big unknowns we have is what happens to that narrow band of junior people who don't gain the experience to move into the mid and senior level people, or will AI overtake that middle senior band? We won't need as many people in that space. And in the end, your triangle goes from diamond to an inverted triangle, and we end up having leadership, heavy organisations. We don't know, I think is the honest answer.
David Savage 13:24
If you kind of think that through, what could the implications be for an organisation such as your own if you end up with a leadership-heavy organisation?
Philip White 13:33
Well, I think it's going to be based on the demand as to how big that tri... It's not going to be about the shape of the triangle. It's going to be how big is the triangle to sustain demand.You know, this all comes down to... You know, I'm talking purely in technology space here, in writing, say, writing software. AI is going to generate an expedite, the rate at which you can create software. My key question is that we don't know in the future is, will the market adapt to say, well, actually, we can have twice as much now, or will the market be able to say, actually, I need to spend half as much?
David Savage 14:07
Outside of technology, we talk more broadly about the job market as a whole. The government talk about this idea that within two and a half hours they can give people the foundational knowledge to close the gap between a 55 year old to a 35 year old and that will give them the skills to thrive in a digital AI economy. What's your reaction to a statement like that?
Philip White 14:31
Again, it depends what we mean by AI, and it'll depend upon the organization's argument. Are we saying that you sat at your desk will be able to use Copilot or ChatGBT better? Or are we saying actually you will be able to spot and adapt the way you work due to including AI technologies in part of your processes? It goes back to that innovation adoption spectrum.
David Savage 14:56
So what do you think individuals can do for their own careers sake?
15 mins 02
What should people be focusing on if they want to make sure that they are attractive to employers perhaps 18 months down the line?
Philip White 15:11
I think any kind of knowledge on that fundamental knowledge of the application of AI, not necessarily AI tools, not necessarily AI theory, but the application of AI, what does, generically, what does AI do? What problems is it good at solving?It's good at forecasting things, it's good at interpolating information and analyzing existing data, it's good at automating or augmenting certain processes. We've seen it recently with the big generative model that's good at creating content. Having an understanding of what the applications are, getting an understanding of a set of tools that sit within each of those spaces will be useful. Because I think it's a specific cross-section, I think, are at risk. I think if you look at the more, you know, we're talking about more mature people in an organization, if they carry with that experience, I still think there's a huge amount of value in experience as we don't yet have complete trust in models. I think that probably the high area of risk is more junior in terms of the hierarchy of an organization and more mature, they're probably the individuals, I think, that are probably kind of key risk of needing to upskill.
David Savage 16:21
If we talk a little bit more about Leeds briefly, the government talks about AI growth zones as a big part of its AI strategy for the country. This idea that they don't want to leave anybody behind in any part of the country.And Leeds is very keen on the idea of becoming an AI growth zone.
16 mins 39 to 16 mins 52 *David Savage*
You talk about Leeds being a community. What do you think being an AI growth zone would add to that? Do you tangibly think it is the kind of policy that can create opportunity or is it something that just sounds quite nice?
16 mins 52 to 17 mins 47 *Philip White*
It has the potential to, definitely. I think a lot of the folks around the AI growth zones, you know, where the big money is, is around infrastructure, data centers. The big unknown is how much of that capital will stick, how much of that, are we gonna invest in data centers and have the revenue profits booked offshore, are we gonna see that money flow into the UK? Will it create jobs or will it just move jobs around the country?If it's successful, how do we retain the talent that we create, how do we retain the IP? It's all well and good if we have a number of really solid AI-based startups that benefit from having closer access to exceptionally powerful GPUs that have been spun out of this. How do we nurture and grow those businesses within the UK and keep them here, longer term?
17 mins 47 to 18 mins 06 *David Savage*
What do you think the ingredients would be to make sure that there are jobs in a region like West Yorkshire that give opportunity to the people of this region as opposed to here's some infrastructure but the money is going to a hyperscaler, jobs going offshore, et cetera?
18 mins 06 to 18 mins 27 *David Savage*
I think, again, it comes back to that community. If you can tie in the community with academia, and it depends on the space, you know, if you've got the transport, you know, if you've got decent transport, you can support bigger communities.But I think, you know, public, private academy, academia, I think if you can tie them together and get them to work collaboratively, that's gonna be the key.
David Savage 18:27
I think within this conversation, there's a little bit of an exploration of where the responsibility should lie when it comes to the upskilling of individuals. Should it be government? Should it be companies? Should it be individuals?
18:40
As a business leader yourself, do you feel any responsibility for the upskilling of the people who work for you?
Philip White 18:47
Yes, I think it's a mixture of both. I think it's a communication and a support thing.I think it's helping our job as a responsible employer is to do the best for the career path for any individual that works for them. So yes, I do think employees have a duty of care to support people in what will be a big shift in the workforce over the next five, ten years.
David Savage 19:11
When we're talking about this more broadly again, there seems to be this resolve that it's a shift from experience to aptitude. And that is where a lot of that readiness of the working population would appear to be taking place.Is that something you see mirrored within your own organisation or is it still very much skills based at the moment? In terms of the type of workforce.
19 mins 38
So I suppose within your business, it's very much the technical roles would still be the hard skill roles. But if you think more broadly about your people who work in HR or your people who work in marketing, how are you helping them, I suppose, adapt to the changing workplace?
Philip White 19:58
I mean, we're fortunate here where most people, there is quite a lot of cross collaboration, not just between teams, but between entire functions. So I actually think most, from an internal adoption of AI, marketing is probably further ahead than most other departments, you know, we'll use, you know, we can take a tech talk that we've run online to then transcribe it and within 20, 30 minutes, we've got technical articles and blogs for the website coming out of it, all automated through AI agents. So I think it is a cultural thing. I think if you've got a culture that is looking for new, innovative ways of doing things, I think you're probably gonna be on a really strong footing.I think it's the organizations. I mean, if we look at the specific skills, there are, we don't, if you look five years ago, 10 years ago, which skills, if I were to write down a list of skills at risk because of AI, there would not have been the skills that I would write down now. I think there are definitely people being caught out since the introduction of these big models that we've seen come out. Professional services in entire industry seems to be one at risk, which is frightening when you look at how much of the UK economy is based on professional services.
21:11
So when you go back to AI growth zones, are we funding our own demise here?
David Savage 21:16
One last quick question then. Out of interest, do you have kids?
Philip White 21:19
I do, I have two of them. How old are they? They are seven and four. Okay.
David Savage 21:24
I was talking to someone in technology a few weeks ago and they said, you know, I've got teenage kids. I haven't. They have. And they said, you know, five years ago, I would have told them, you know, software engineering, you've got to learn coding languages. That's going to be what sets you up for your career. And they said, now I wouldn't tell them that.I wouldn't really know what to tell them. When you think about your own kid's future and you think about how they make sure that they are competitive in the kind of the workplace of the future. What do you think you'll be telling them?
Philip White 21:56
So I still think there will be a significant demand for technology roles, but they'll evolve. I think we'll just see them evolve faster than we have done before.I mean, I started off life in technology. If I worked in technology 20 years ago, half of my day would be looking after memory in a piece of software. That just isn't done now. People don't do that anymore. So I think we will see, naturally, as long as, and this goes back to, as long as universities and education more broadly can keep up with the rate of change, that's the big problem we've got, is can the people who are teaching people to get jobs move at the pace of the technology? Because we've got people coming out of an education conveyor belt with a degree that was relevant three years ago that doesn't exist anymore. The technology may not be in use anymore. So I think, yes, I'd push them into technology. I think person-to-person skills are more important than ever. People still want to be and work with people. So I think I'd be pushing them in the realm of a, yeah. Technology, I'd be quite keen if they wanted to. I mean, it wouldn't force them to work in software engineering. But keen to work in technology, but interpersonal skills. You know, we were talking about how we support people through their careers. There is sharp words. If people don't want to attend meetings, don't want to attend video calls, just want to communicate via a text prompt in Teams. Those, you know, we've had some sharp words saying, you really are at risk of becoming replaced by a bot. Get a room, talk to people.
David Savage 23:22
Our report talks about this 82 percent rise in AI skills year on year compared to last year. And there does seem to be a lot of hype around how do we drive return on investment from AI. Again, I think only a third of organizations who have implemented AI projects of some kind have seen return on investment.Do you think it's a bit of a red herring that this is just so new to a lot of organizations at the minute? Most organizations are still going through digital transformations, never mind rolling out AI. We run the risk of fixating a little bit on these big numbers because it's telling us a story that isn't really going on. And this is just a natural maturity process that will iron out in the next two, three years.
Philip White 24:09
Yes and no. I do think the AI wave is, yeah, you know, one of the spectrum, it is just a technology and I think people are getting excited about a technology. But on the other end of the spectrum, it is a much more powerful technology than I think people are ready for. And it is relatively in its early days.You know, if we look at the step away, you know, AI has gone through various waves over the last 20, 30 years. But, you know, Google's research around what eventually became, you know, the transformer paper that came out, attention is all you need. This is where all these language models would come from. That was the inflection point and that was the point at which this technology is set off on the bounce again. And it doesn't show any signs of slowing down. So I think, no, there is not a hype. I think people are struggling with it because they, you know, what is ROI? How are they measuring it? I mean, anecdotally, I spoke to somebody in a medical field that had run, said, you know, AI does not work in this space, just doesn't work. And I was fascinated because I was, you know, is that a case of the AI technology as it stands? Or do you think AI will never get there? And, you know, we went through in the 18-month trial. You know, you have given it a go. That is fair. It has not been a toe in the water. But when I had to go under the surface, it was like, well, how did you, you know, how did you define what success would be on day one? How did you identify this use case as being a really good use case to run AI against? And I got nothing. And how did you determine what the ROI might be? It was like, oh, well, it was a, you know, it was a product that we were approached with. I was, okay, what made you pick that product? No, they said they had run an 18-month trial for free. So when you look at how they are being applied, you peel back some of the layers. They are just not, what I talked about earlier, that was small, incremental, you know, risk is good if you fail quickly. None of that was being applied.
David Savage 25:54
I think we've got what we need.
Philip White 25:57
Did I wonder about there? No, no.
David Savage 26:01
intro again. Yeah. Yeah. I mean, the intro, to be honest, the intro was just wobbly.
Philip White 26:08
When I I meant like verbally was I wondering I know I'm physical no
David Savage 26:12
Verbally, no, no, it was perfect. All right. I'm just kind of having a look at what this kind of...
Philip White 26:25
If you're all right for time. Yeah, that's fine. OK. Yeah, needs to be out by about six. Fine.
David Savage 26:35
It's a little bit of intro again, yeah Okay, I mean the intro is basically just chitchat one warm-up. So
Philip White 26:47
Well, I'm warmed up for the warmer, that's fine.
David Savage 26:51
Look, it's been a pleasure to speak to you today. I suppose if I had one more kind of throwaway question, how do you feel about this whole piece?You know, we're talking to people over the next couple of days here in Leeds, and we've got a day where we're talking in London coming up next month, around how we need to think about AI if we keep in mind the person, the individual, and making sure they aren't lost in all the hype around AI. What would you want people to take away from if that question is being put forward?
Philip White 27:23
I think it goes back to that both ends of the spectrum in that people need to understand what AI can do and what it can't do. It's about the technology that's coming, but understanding its limitations, understanding where it's going to impact. But then on the other end of the spectrum, it can do certain things very, very well and it will do them and it will replace certain skills in the job market at the moment. So I think all of this must be around transparency and communication.If we talk to an organisation and we're saying, how are you communicating your AI strategy? Well, we're not telling it enough. If you don't talk about it, people will be assuming the worst. Best thing is open, transparent communication.
David Savage 28:04
Thank you very much for your time.
Philip White 28:06
Thank you very much.
