Euan Blair | The Future of Education & Skills

0:00Welcome to Jimmy's Jobs of the Future. Today I'm joined by Euan Blair, who I think is building one of the most interesting companies in the United Kingdom at the moment. Multiverse is an edtech platform that is leading the skills revolution. It's valued at over a billion pounds. We talk today about scaling a company, what it's like going from zero to one, far beyond. On to today's episode.

0:28your multiverse was incorporated at the end of february 2016 and i was just thinking like how much the world has changed since then if you could go back and tell yourself one thing when you incorporate the company what would it be do you think start thinking about ai and its practical applications right now because that's going to be hugely valuable uh what can i say i mean probably just that the journey itself and i was actually thinking about this in the context of when i look at the kind of people we're hiring for now as opposed to the people we're hiring for back then at the early stages because you know zero to one one to ten ten to a hundred a hundred to a billion you're looking at very different types of people.

1:13You know, we have more specialists, we have more people who have experience of kind of hyper growth because you do need that around you. But one of the biggest things is finding people where the journey is the thing that attracts them more actually than the destination or the mission. Because having people who are sort of really dialed into how do we make this a generational European success story is a really difficult thing to find. But when you get those people it helps the company just absolutely fly so when you are trying to hire those first 10 people what are the skills that you're looking for at that stage so you're you're basically looking for scrappy generalists yeah you're looking for people who can do a bit of everything who are prepared to um kind of pivot immediately if a new opportunity comes up who can experiment but Interestingly, because we made an unusual decision for a sort of tech startup scale up in entering a heavily regulated market through the apprenticeship system.

2:13We also needed people with genuine regulatory expertise, right? Because there's a huge kind of compliance burden that we have to sort of comply with. And so it was a mix of kind of people who understood that and then people who were kind of scrappy generalists who wanted to take a chance on something where you couldn't pay them very much. you couldn't commit to anything though i do remember looking a number of those early employees in the eye especially ones who are leaving some quite interesting jobs as we started to get a bit momentum and say trust me when i say the business will be valued at least a billion dollars one day right like this is this is going to be a hell of a ride and what were the inflection points along well firstly what gave you the confidence that that was going to be the the case right because we talk a lot about having more of that optimism right in the uk how did you know at that point it was going to be worth it so i think there are a couple of things one was just the opportunity was really obvious to me every companies that historically viewed kind of skills and education as a sort of niche employee benefit right we're gonna we'll buy all of our employees a library card to access a coursera udemy audacity whatever else they're going to teach themselves some stuff it might keep them here longer it might make them more productive but actually approaching it through the lens of what are company's biggest needs and how do you kind of build a skills infrastructure to help them go and sort of address those needs.

3:40And that felt like a really compelling way to kind of view the apprenticeship system, right? How do you create the scaffolding on which companies can grow and be successful? And so it was that sense of the market's not addressing this issue. There's an incredible opportunity. if regulation actually plays in our favor because there's a need to go and invest money in apprenticeships, let's turn this into a kind of engine of sort of growth and opportunity for companies as well as individuals. It was kind of putting a B2B lens on what had traditionally been seen as a B2C challenge. And when did the apprenticeship levy, when did that sort of make a difference?

4:15Because you're right in saying that, you know, unusually for a UK tech startup, you're in quite a regulated space. Like how much did that impact your thinking? Well, it meant companies would basically have to have a conversation about, are we going to do something about this? Do we just write this off as a tax or is this something that can genuinely be useful? And so that kind of created a bit of a pull. But the other thing was, everyone accepts that Britain historically hasn't spent nearly enough on skills. We've spent a lot on kind of technology tools, but we've spent very little on equipping people to use them.

4:50You've had a two-decade relentless march of kind of every conceivable type of software tool, and yet productivity's flatlined. And so it was a skills problem, quite manifestly, and an enablement problem. And so it was a combination of genuine market need, a potential market incentive, and then the fact that the space was sort of dominated by quite old legacy companies that were private equity owned and just kind of eking out a bit of EBITDA margin every year, but not really changing or disrupting their delivery models. Yeah, when I was trying to explain to the team about what you guys are doing, I was a bit like you're kind of like the monzo of the education world.

5:27And if they're challenging NatWest, Barclays, et cetera, you're kind of challenging the universities a little bit. Would you describe that as fair? Yeah, I think that is quite accurate. I think the other thing is just sort of any time you can look at a market and say, what would this look like if we flip the current model of delivery? Again, as I said, instead of asking what things might people be interested in learning and providing them with those things, saying what are the things that companies need to be more successful and what do their workforces need to be more productive and how do we kind of productionize that?

6:01Yeah. And what do you think they need to be more? Well, and look, it's been interesting. You look at our evolution. We've hit a number of inflection points. One was around data when we went in a big way into kind of data literacy and data skilling. We'd speak to organizations in nearly every industry, particularly in the public sector, where they'd say we have reams of data. We're inundated with it. Everyone wants a dashboard. Everyone wants something to kind of give them real time information on what's happening in their organization. but we have very limited insight and actually very rarely translates into action.

6:34A lot of organizations as well would have quite small, discreet, expensive data teams and data scientists who were working on really basic tasks because there was no sort of self-service data culture. And so embedding data literacy in an organization is a really powerful way of sort of taking that rich seam of intelligence that exists and using it to provide better products, better services, better outcomes, better margins, all of the rest of it. And then, of course, we've really hit this inflection point around AI because, in theory, AI is a solution to a long, long list of problems for most businesses.

7:10But actually, the more I would speak to CEOs, the more a lot of them thought it was a bit of a solution looking for a problem. Because when it came to practically applying it within their organizations, people lack the skills and capability, right? outside of kind of summarizing meeting notes and sort of doing some fairly basic plumbing tasks with AI and automating, which you could already do, people weren't equipped to go and start to build agentic capabilities or new features or focus on kind of how we actually going to grow revenue as a business or into new markets. And that's where the AI training and skilling piece has just become so important.

7:45And talk to us about some of the other inflection points along the way of that we sort of got diverted off the people because you're about 700 people now or so yeah 800 so what what were the kind of hiring moments that changed along that journey you've talked about zero to ten what was it like at 50 100 were there any moments along that way where it just you felt the sort of change in it well a big a big kind of inflection point and one i think of as kind of the real sort of when the company really started with a vengeance was when we first raised venture capital funding from Lightspeed, right?

8:17We raised our seed round in 2018,$4 million. That allowed me for the first time to have some firepower, to go out and hire people with more of a track record in tech and more experience to kind of hire people I wouldn't have been able to afford before or necessarily persuade. Because when you get validation from one of the, you know, Lightspeed have been one of the most successful investors in the world over the last 20, 30 years. it makes a massive difference to your ability to kind of attract talent and you you go through as i said this kind of team of scrappy generalists some of whom do manage to evolve and stay with the company and develop and i'm delighted that we have some people who've been here since our first year yeah which is is pretty cool but you can bring in people who've seen a version of the movie before because one of the hardest things is trying to build without a clear sense of the pitfalls and you can get some of that from investors and boards and advisors and other things but actually having people who are dealing with operational challenges that are a version of operational challenges they've seen before is super valuable and when you're a first time founder by the way you need that in spades and how's your kind of like how have you developed personally over the 10 years as well right because it's a very different job you know inspiring 10 15 people to inspiring 800 i think the the some of the hardest things around developing at first accepting that you can't solve every problem the business has and i i pride myself on being in the detail and kind of being able to really dive into a problem really understand what's going on in every facet of the business but that stops scaling at a certain point and you need to find people you can trust who can sort of take some of that burden from you but also start to make decisions there with more limited input from you at the same time you have to be very careful you don't sort of tap too far in that direction right brian chesky you know recently coined the term founder mode right and i'm i'm a big believer in that that especially at growth stage and then you know if you look at everything that's happening in the market at the moment as a founder c you you still have to be prepared to trust your gut, be decisive in ambiguity, deal with imperfect information, and you can't just outsource a long list of decisions or functions.

10:36So you've kind of got to balance the constant being in the detail and kind of actually being a driver of a lot of things that happen in the company while not sort of getting so sucked in that you can't see around the corner. I think honestly the hardest thing for an entrepreneur in a sort of hyperscale up is chewing gum and walking at the same time. It's basically figuring out where you go next while making sure that the core machine is flying and you're nailing your kind of immediate targets. And one of the things when you raised some of the bigger venture capital rounds as well was to expand to the US.

11:10Now, company accounts show that kind of revenues in the UK are still around 80, 85 percent. What have you learned about trying to export to the US? I've learned it's hard, right? And it's something that, you know, we knew that before we go in. There's a long history of British companies expanding into America and the expansion not going exactly as they planned. I think interestingly for us, it was a good lesson on the fact that it's clear there was market failure around the education system in America. College is incredibly expensive. Most people don't complete. Most people don't get a degree within five years.

11:46Employers aren't finding it useful for them. But interestingly, they're also not fixated on solving that problem. And short of any kind of government focus and will on the issue, you know, in America, they spend a thousand times as much on higher education as they do on workforce education, which is really out of whack. No defined definition of an apprenticeship in any real meaningful sense. It was much harder going to persuade companies to take that seriously. Do you think the UK or US in that sense is more meritocratic?

12:23Yeah, I don't think the US index is more meritocratic than the UK. They've got a lot of structural problems around access to education. The state provides a lot fewer services and often a poorer version of state education. So I actually think Britain scores fairly well on that. But I think there is more of a cult of individualism in America that makes sort of collective action to address societal problems more challenging. And it's interesting if we look at the apprenticeship landscape in America, most of the apprenticeship provision is driven by not-for-profits and charities. It's sort of almost a social good rather than actually how do we make this a foundational business practice.

13:04and it's you know a long time ago right businesses and and early in america's history businesses literally built cities and facilities and infrastructure to house their employees and give them places to live while they worked for them and there was a really powerful version of the social contract there i don't think that exists in the same way now and and with ai it's it's unlikely to revert back to that true what are the other cultural differences when it comes to business between america and well so we're actually quite unique despite nearly all of our revenues the UK and Europe, pretty much all the funding we've raised has come from America.

13:38And that is in part just because there is not just a sense of kind of dynamism, but a real willingness to take risk and to go big or go home. You know, I would have conversations with growth stage investors in Europe who would say things like, I want to, I really want to get certainty that I'm going to get at least three times my money back over the next three to five years, right? I went and met Mary Mika, who invested in the company and has just been an incredible force and fount of advice in so many ways. First thing she said to me was, Ewan, how do you make this a$500 billion business? And it's that difference in terms of let's build something generational and profound that can impact the world and make us all a giant amount of money on the way, as opposed to let's do something that doesn't go horribly wrong.

14:26And we can all sort of say, yeah, we tick that box. and just the way they approach the journey, the risk calculus and everything else is wildly different. We're getting more of that in Europe, particularly at early stage. I actually think the challenge is at growth stage. And how did you sort of level up that? I mean, you were clear there, you were talking about the first 10 employees, you were saying that this could be kind of a billion dollar company, but how did you kind of level up that ambition as you go along? So you sort of, you're looking at the market, You're saying, how big could this thing be if everything went right?

15:02And, you know, AI for us has been a massive boost for a number of reasons. First, it sort of catapulted the skills element from something that is a nice kind of thing for us to do for our employees, as we were saying earlier, into something that is kind of existential, right? The companies who adopt AI fastest will be the ones who will win the biggest. There is no question on that. And how are you going to adopt AI if your workforce isn't skilled and trained in it? So, you know, in the early days, we were selling to HR managers and early talent and the sort of people function. Now we're selling to CIOs, COOs, CEOs and boards because they're saying, hang on a minute, this is this is kind of foundational to our business strategy.

15:46And we really need your advice and input. And if you think at the moment, billions has been invested in the foundational model for AI. And we're already seeing those commoditized, right? But you've got GPT, Claude Anthropic, everything else. Billions is being invested at the application layer. Microsoft Copilot, Gemini, and Harvey, long list of applications. The adoption layer has been largely neglected. Now, you know, last year, Accenture made more revenue from AI than OpenAI. and in large part because everyone is trying to figure out what do we actually do with this. That is an incredible market opportunity.

16:24And for a business that's built its kind of reputation in literally training people through a series of use cases into how to do their job better and get more productive, right, all on the job, it is a very short hop for us to actually start to say how do you drive more widespread workforce transformation and in the AI era. It's a massive thing. And so that, by the way, just hugely expands the size of your market. So it's interesting on that. What do you think about the wider societal impacts of this potentially? Because Scott Calloway had a really interesting point in his email this week, which was just about basically how a lot of these companies have increased their valuations by trillions overall.

17:05And essentially, you're not going to get quite that much productivity increase from it. What you may be able to get is by cutting costs out of the business. And you're seeing the likes of Microsoft, even with this incredible open AI product, still shedding jobs. And just what do you think could happen over the next 10 years when it comes to the wider workforce? So, by the way, Scott is an angel investor of multiverse and has often given some great advice on these areas. It's very hard to predict the shape of the workforce, but kind of over the next 10 years. The reality is AI is both the biggest threat and the biggest opportunity for the workforce.

17:42Because if you think of the kind of macro environment and take the debates happening in British politics at the moment as part of that, there is a lack of resources, right? Financial constraints. What should the state be providing? What should people expect from their governments? And how do you manage that through the tax system and everything else? AI could usher in an era of abundance across the entirety of the public sector and that is net good for everyone but at the same time it could wipe out jobs if people aren't brought on the journey so the obsession for me has always been how do we maximize the benefits and build that kind of optimistic case scenario and make it a kind of base case scenario and minimize the downsides and the only way to do that is through training and skilling because if this just becomes about how do we reduce costs and kind of just strip them out from the state and the workforce in general we could end up in a race to the bottom very quickly but if we make this about how do we grow the size of the pie rather than cutting off the crusts it's a very different conversation because every business could generate massive returns if they get the ai piece right yeah but it's not going to be a tooling issue it's going to be human issue so and that that benefits humans right train people instead of focusing on basic automate tasks that can be easily automated and aren't really hitting the bottom line have them focus on the next frontier of the business right let's let and our my favorite customers in many ways are those that are actually probably felt behind the previous technology revolutions we've experienced but see this as a chance to catapult themselves into the forefront of this one yeah i i do worry that particularly when i when we go back to that incorporation date i was thinking gosh how much the world's changed over the last 10 years and we've got our five-year anniversary as well it's all right yeah it's just like yeah i was just one of the things that worries me is the costs on business have got so much more partly because you have also had a pandemic yeah an energy shock in that as well like it's just really tough and i worry that the temptation for a lot of businesses is to cut off the crust as you say rather than grow the pie well i think that that is the easy route if we don't offer them an alternative and again if you could show them how their employees can become much more productive can do things they couldn't do before you know we've we work with 100 nhs trusts now we help one digitize their patient records and they're now seeing double number of patients a day we've helped trust solve enduring bed occupancy problems because they couldn't map which beds were occupied and where they had capacity and everything else were helping children's services in local authorities triage troubled families better so they know who to spend their time with like the there are loads of really good positive cases that don't require more money yeah and you know given actually the debate in british politics is basically well the public wants more and better public services but doesn't want to pay any more for them yeah right well the only way you solve that is ai and the only way you're going to do anything meaningful with ai is if people actually on the ground in these jobs are trained to spot what are the great use cases what are the opportunities how do you do it ethically how do you do it in a way that protects data and privacy and everything else this is the biggest reskilling need we have ever faced certainly in the kind of modern post industrial economy and what does government need to be thinking about with when it comes to because when we interviewed peace carl just before summer just been made business actually from technology of secretary but it's the classic government uh move around but like he said that he thinks agi will be here by the next general election and when the pm was on as well he basically said you know he didn't think that was going to be far off either and i just thought this could be such a fundamental change to our economy like i'm not sure westminster has quite got its head around it or anyone yet frankly so this is the thing i mean even just talking about agi there are multiple different definitions of what AGI is.

21:38AI is already better at humans than a number of tasks, right? Even pretty super intelligent humans. So I, you know, I slightly put AGI in the kind of marketing category, right? You know, who's going to declare the moment? When is it going to happen? I think the bigger question is, what is the role of humans where AI is actually doing a lot of the things that humans previously saw as solely their purview? And that is a really good question, discussion to have but again the reason i come back to the kind of the training and skills need is look most organizations don't even have a clear sense of what their workflows are right what are the various kind of jobs people are spending their time on what are the activities they're doing even at a manager level people have very poor visibility of that but if you can start to kind of automate workflows and break that down you get a clearer sense of where are the opportunities for what could people be spending their time on what are the things that we can be doing if we're building kind of the next version of our company, AI actually could act as a force to help even very long established old companies act like kind of brand new insurgents, because it just opens up things they can do with the data they've acquired historically, with the kind of intelligence they have, with the domain expertise.

22:50I mean, AI can be an incredible sort of generalist powerhouse, but actually domain expertise is something that's still really valuable. And you can use AI as a co-pilot to apply it in ways that you wouldn't have thought possible before so you know i i think predicting is hard one thing we can say with certainty is train people in it yeah no doubt about that where have you used it for as a founder ceo or personal life like i find it fascinating talking to people about this so many things um and and generally by the way claude for writing because i just think i i like its writing style preparing board packs we basically have built our own board agent we can challenge kind of we had a board meeting last week and we went into it with a list of questions that we thought the board might ask based on kind of historic board packs all the information would fed them some of the other things they were doing in the world some of the other companies they're involved in and it was the most comprehensive briefing i'd ever had and and really easy to interrogate um i use it a lot around kind of taking meeting notes and basically going back and saying right what are the big themes that i can take out of this week sort of just aggregating everything together and then pulling out insight because when you end up in back-to-back meeting you often miss the insight yeah being able to kind of have that played back to you in the form of a podcast or a kind of high level summary means you miss a lot less i found that really valuable my personal life i know you've got young kids i've got a two boys one about to turn six one about to turn four i've used chat gpt to build kind of mini i'm a big nerd by the way dungeons of dragons adventures for them which is great and they love um tell kind of creative stories test them on facts everything from you know what's the fastest mammal to kind of what um is there any possibility that a new species of giant squid will be discovered and and they love that sort of thing yeah and so it's kind of using it as an addition you know in work for me as a founder ceo it's probably a combination of a hyper effective exec assistant and a chief of staff that sort of doesn't miss anything which is interesting does occasionally hallucinate so of course you gotta you gotta sort of be be alive to that and in my personal life like an incredible kind of partner for entertaining my kids right yeah yeah it's and and the i'm fascinated to see the shape of AI in early years education because again training is required because it can absolutely be an excuse to just outsource any kind of critical thinking yeah but if you use it as a genuine partner in a foil yeah how do you think the working world might have changed I obviously think about this a lot with kids and so on when they're entering the workforce in kind of 15 years time like what what do you think like how do you

25:43it's interesting already it is easier than ever to set up your own business because you know you can build what a team of 10 engineers that needed four million dollars of venture capital funding needed before you could now do in a couple of weeks on your own you know you don't even need to be able to code, right? Low code, no code solutions can help you build a kind of early version of a product that you can then refine. You can digest competitor analysis. You can get sort of instant feedback on potential market and customer reactions. So I think we'll probably see more entrepreneurs, more people building businesses, which is broadly a good thing.

26:23I think the nature of the kind of jobs people go into, I think will be the biggest challenge to the kind of education system, because I do think the thing we have lacked in kind of the K through 12 education system is a reimagining of what it should look like in an era of AI. You have teachers right now who are saying, what do I do when I've set a kid and an essay project? They've turned in something that's great, but suspiciously great because, you know, the punctuation is perfect. I've never heard them write in this tone of voice before. They've got subheadings and footnotes and other things I wouldn't normally expect.

26:57right now you can on one hand say okay well this is a problem how the system doesn't work let's ban ai or you can say how do we actually reimagine what we're teaching people and how we're incorporating into our lesson plans and other things so that that isn't an issue yes you know wrote exams and wrote homework and the kind of easy stuff has to disappear and this requires a reimagining and then my hope is as well there is more of a focus on how you use education the school system to prepare people for what comes next in the workforce because most schools don't do anything around financial planning or literacy never mind taking care of yourself cooking cleaning other things there are lots of basic life skills that are a long way away from being automated and the reality is if you speak to any employer they will always hire people who are especially at the early stage who are prepared to work hard who are desperate to learn and will come up with creative ways of solving problems right so long as you're kind of inculcating those skills the labor market outlet doesn't have to be bleak um and there's every every single kind of technology inflection point has ushered in a new era of jobs that no one's even thought of yet what do you think so we've asked there about the feature and what what would you and blair be doing if you were 1920 today do you think would you go to university yeah i've thought about that before probably not because i didn't find the kind of i didn't find it all that exciting just kind of learning in an academic classroom environment there was stuff i was really interested in and loved learning but the kind of broader experience um was not particularly appealing to me i i'm not sure as to whether or not i'd go and immediately set up a company and also you know i felt one of the most powerful things for me was going to morgan stanley going through that kind of um graduate scheme there and basically learning how to work incredibly hard right yeah all the way through the day and the weekend and whatever else you were needed learning to find kind of joy in things that often seemed incredibly bright you know building presentations and financial models that would never be used by anyone um but there was a kind of discipline in it and um learning how to kind of build relationships with often quite important senior people at a kind of early stage and networking in the office i will say by the way the work from home thing can be absolutely dangerous at the start of your career yeah it is you know you need connections to people you need to learn to collaborate you need to learn from someone next to you can do some of that virtually but we we have to make space for it yeah i'm anyone that is working from home more than one day a week in the first seven years of their career i think is a great mistake it is the is honestly one of the most profound taxes you will end up paying on your kind of future earnings potential and success because you miss so much not to mention just the kind of rituals of being in an office and and the sorts of things that exposes you to.

30:09What is the multiverse policy on it? So we're three days a week minimum. Yeah. And we have quite a few employees who live outside of London. At the moment, our kind of headquarters is in London. And there is travel involved for a lot of people, but we've made it work because it's important. We found that, you know, through COVID and everything else, well, we did grow and we saw a lot of success. problems occurred between teams and in kind of cross-functional collaboration that wouldn't have if we'd had people in the same room things that could have been resolved in an hour took several days to resolve and multiple kind of passes of google docs and coda pages and other things whereas sometimes just being able to bring people in a room and get them together and work on solving a problem is magic and gets you there much quicker and i think particularly the big thing for Whereas with AI and the pace at which the world is moving, you lose so much context through not being together.

31:07Like it's made, perversely, AI and its capabilities have made in-person interactions that much more crucial and the human components that much more integral. Yeah, I think it's an interesting example there, Google Docs and so on, of where technology can end up creating more bureaucracy. Like an ease of bureaucracy as well, I think is a real danger for companies. Right, because who doesn't want a chance to have their document in theory reviewed by multiple different people and refine it to the perfect version? This is why a lot of people very sensibly talk about the 80-20 rule. And it's also, again, in an environment or an era where we have, in theory, more intelligence than ever and more access to data, actually one of the most important skills is being decisive in ambiguity.

31:52understanding that as you scale and you bring in more senior people you're really paying them for their judgement so you don't actually necessarily want them kind of spending too much time worrying about every possible piece of information they could get access to you're actually saying look you've seen some things before make a call you have this fantastic 13 skills of the future that you've recently done and so I'm going to read them out I'm not going to test you They fall into four buckets, right? So cognitive skills, analytical reasoning, creativity, systems thinking. Bucket two, responsible AI skills.

32:30AI ethics, spotting bias, et cetera, cultural sensitivity, which I think is going to be huge. Third one, self-management. So curiosity, self-regulated learning, determination, and detail orientation. And then fourth bucket around comms, empathy, tailoring comms, and exchanging feedback as well. I just wonder, particularly on the third one in terms of self-management, like how much of that sort of curiosity, detail, orientation, determination can be taught? Yeah, look, I'm a believer, by the way, nearly everything could be taught. I actually think being exposed to something, working with something, there's a reason, by the way, we use human coaches at Multiverse, despite the fact we have an incredible AI agentic coach called Atlas.

33:20Because humans can help you learn things, right? You can interrogate things together. You can use AI to help learn things as well. I think the issue is less around the teaching and more in that knowing, again, as you read through those 13 things, these are not technology problems. that we're talking about here. These are kind of questions around judgment, around bias, around cultural sensitivity, around awareness of what's happening in the world. The AI will, for the foreseeable future, be pretty limited at. And so if, as a human, you're trying to get the most out of AI, as a co-pilot or as a colleague, you need to be able to display these skills.

33:57Now you can test for them, and you can absolutely learn them. I think that the big thing as well, though, detaching skills from context is not all that valuable. If you look at how we work with organizations, We kind of run a three-stage process. We diagnose the gaps between the goals of the organization, the capability of their people. We then quantify those gaps. And then we prescribe kind of skills and learning pathways for people to follow to basically meet those gaps. And there is a massive difference between how someone needs to demonstrate kind of judgment and work with AI in one role, like a clinician, for example, versus an operations manager or a warehouse worker.

34:31and so kind of being context aware is really important and people learn much faster when they're applying a kind of theoretical skill to go and solve a real world problem where they they're kind of dialed into the context and they're sort of working on this so i i actually think kind of contextual application is the most important thing in learning skills and if you get the contextual application right you can learn almost anything yeah and that's the cultural application as well i think it's going to be really right really important like when i read that i thought that is going to be the number one thing yeah yeah and and oh by the way challenging ai because the reality is ai um is certainly the current versions of models indexed towards being massively supportive of whatever issue you have right to a somewhat sycophantic and slightly off-putting degree you have to be able to challenge and they'll respond to challenge but if you don't know how to do that you don't know how to set up the right prompts if you don't know how to incorporate the human judgment element and all these other you know cultural awareness as we said um you end up you end up getting bad outcomes pretty fast yeah it's it's interesting on that because i thought that when i the empathy point because that's often something for the last five years it's been trotted out as a thing that's going to be really needed and i was like i just thing is over indexed on it's become too empathetic actually i agree and like my dad always used to say you know the difference with sort of management partly is you need to know when people need a glass of whiskey and when they need to kick up the ass yes yeah that's it and yeah i think that's right this is you know look um when does someone get the hair dry treatment and when do they when do they get supported it sort of it sort of depends doesn't it and and again that awareness of context and what is in the best interest of the individual I'm serving as opposed to how do I make them feel good is something AI really struggles with.

36:25Yeah completely um we've got a few they're not really quick fires as such they're more just sort of shorter um ones um what's out of those kind of skills what do you think your superpower is like what has been sort of enabled you to kind of get this far building a business?

36:46So, and I think it's important for any entrepreneur, and I've touched on this a bit before, but being able to make decisions fast within perfect information. I think that is really, really important because so often there will be, you know, if every market opportunity was massively obvious, then, you know, there'd be very little chance to kind of go and build a new market or spot something others haven't. um so you have to have a high degree of conviction to basically say um in this area i might be right and a lot of other people might be wrong yeah and that you know you have to then go and kind of follow that and follow it aggressively and if you can bring venture capital and other things to go and support you so you can go and fund it you've got to be prepared to take bets with a genuine mix of intuition gut calculation and what's happening in the world like that being kind of decisive within perfect information is just essential.

37:41Can you give an example of where you've got that wrong at some stage? Sure. I assumed massive market failure around kind of the education system in the US, the piece around apprenticeships. This would be a very straightforward entry into a new market and would be able to kind of dominate it very quickly. And it didn't work out that way in part because, as I said, knowing that a structural problem exists and that businesses will actually put their hands in their pockets of pay to fix that structural problem. It doesn't always work that way. Which country excites you most in Europe to expand to?