Brainfood Live On Air - Ep 400 - The State of AI Readiness: Is Your Organisation Ready?

0:05and you see just basically more people fall through the cracks um yeah you do just you just you just high stress just walking through, you know, the certain tube stops or certain exits and whatever, you always got to think, you know what? You're an easy target. Yeah, yeah. So it's not brilliant. But there we go. We just hope for better days, can't we? All right, here we are. Welcome, everybody, to Brain Food Live on Air, bringing it to you every Friday, no fail. It is baking hot. I think everywhere where we are at right now is 30 degrees plus. I can see all the comments here in the chat. 35 degrees C in Amsterdam.

0:51That's nuts. That's where the canal start evaporating, isn't it? So yes, it's blazing hot. It's still about 34 here in the UK. Way too hot than it needs to be. And I'm just sweating through this process. But anyway, welcome everybody. I'm excited to have this chat with you. We're going to be talking about the state of AI readiness and is your organization ready? This is a topic I think it's been kind of bubbling in the back of everyone's mind for a while because you do see certain exemplar businesses publish stuff about themselves. Say, yeah, we've done this audit. We know where we're at. And then everyone else is looking around each other and thinking, yeah, we haven't done anywhere near that.

1:33So let's explore the topic. What is AI readiness? How do we get to AI readiness? Is it even viable or realistic or desired? We're going to be speaking with a bunch of really smart people into the panel, and we're going to have a discussion very shortly. But firstly, let me introduce my co-host for today. It is the wonderful AJ Lamont. I'm the only person who calls you AJ, aren't I, these days? You are. You are very much so the only person. It's because you've altered your LinkedIn to Amanda, just a plain old Amanda Lamont. No, that's ever since school, really. I've never gone by Amanda Jane, but I'm glad you took a shine to it.

2:13Why do I call you Amanda Jane then? It's kind of more the reprimanding name for my mom. Amanda Jane. My dad used to call me AJ, though, but it's also because he was AJ. So he was Andrew John and I'm Amanda Jane. Right. So my name in conventions is following your parents, which is rather weird. It makes you even more memorable, Hank. There we go. So it's Amanda Lamont, everybody. I've just shared her LinkedIn in the chat stream there. Make sure you connect with Amanda. Really cool cat in Berlin, right? No, you're in Potsdam. You're in Potsdam, Berlin. So it's the more beautiful neighbor of Berlin.

2:52I have to say I'm biased now since moving here, but obviously I spent like the last seven, eight years living in Berlin previously. Yeah, fantastic. One and the same. Everyone commutes between for work. So you don't really see the separate. Fantastic stuff. And a great commentator on LinkedIn as well. Lots of interesting things. Sort of you've been posting, Amanda. I think you've got a talent for that, by the way. Your observations are really, really good. And your communication is really good as well. So I think that you do more of that whatever you get up to. Good to hear. I'm glad to be here.

3:25Yeah, cool. Okay, we're going to review the newsletter in a second. But before we do that, we have to thank our sponsors, everybody. Because without our sponsors, we are not able to do Brain Food Live. every single week a company steps up and says hung you need to continue having these conversations because they are so valuable and interesting for the community uh here we're going to sponsor it for you um so today our sponsors are coder pad um folks if you're hiring for tech talent which i'm sure you are by the way it's the only kind of sector that is still uh sort of on the trend upwards uh you need to make sure you are hiring the right people um what's happened i think my internet just one second you froze yeah one second i'm gonna correct it don't worry you're back on i think it's just the heat melting literally melting at the cables uh but folks um back on with codepad uh are you using any kind of assessment platform uh to to assess your tech talent if you're not you should check out codepad codepad codepad.io um just sharing the link in the chat stream there go and check it out let me know what you're using in fact for your tech hiring assessment needs okay let's get on with this we're going to review the newsletter real quick before getting on with the debate Amanda what did you read last week from the newsletter was there anything interesting at all there was a lot there actually a lot to digest on a Sunday morning if you're one of those people who reads it on a Sunday like me I think the biggest thing we've seen around is the EUA Act now being enforced.

5:07I mean, I don't know how many posts I saw on that already. And I can't help but wonder, you know, what, this isn't new, this isn't something that we haven't heard of before. And, you know, it's an important move. But also, there's so many other annexes and other components of the Act that are still delayed until the end of 26, 27, and 28. So for me, the question is really how are we going about making sure that our organizations are more compliant? And what's the strategy there? What are the loopholes? And it makes me think of, are we going to see more compliance legal officers coming up in job ads, on job boards?

5:54Is everything just going to be left to the data privacy officer? and also as an HR manager at a new startups. I don't know, there's so many not zero to one, also less than 50 people startups that have been popping up everywhere. We know what it's like. You're hired as the first HR person. There's an EU AI Act now and you're scrambling in forums. Hi, has anyone done this before? What do I do to make sure that we're compliant? because founders are focused on revenue, impressing the VCs, and we've just got to take care of all that people stuff. So what implications are there for us? Where does it fill in?

6:39I think there's two implications, isn't there? I mean, firstly, expansion of responsibility for the small HR function, absolutely. Where else does it go? Even though I think the EU AI Act probably is targeted towards larger scale companies, So presumably you're going to get more beefed up sort of teams to help. And there's going to be a recruiting need, of course. I do think companies going forward will end up just carrying more legal people in their function, even though they may not be doing too much. It's almost like you just need to have that complement as redundant capacity in case something happens or even to demonstrate that you have this person here.

7:18So, yeah, I think there's some implications that way. I hope it's going to be a positive thing as ever. It's all about the implementation. And by the way, Amanda, you're right. It's not new. People have been knowing about this for a long while. It just came into force. And so in other words, components of the EU AI Act are in force as of 2nd August. And so, yeah, we should be seeing some legal cases being made pretty soon. And you've got to keep an eye on. Yeah, another point to that is the companies that operate globally, right? I think I actually saw in that compilation that LinkedIn posted of, you know, different thought leaders, some opinion pieces.

8:06There's someone who actually narrowed it down pretty nicely, who's in the construction industry, actually. So just as construction as a use case, looking at, you know, non-EU companies with EU-based connections, people doing work for EU clients, you know, how far this reaches as well. And I think there need to be a lot more guidelines, kits out there. You know, a lot of people who like to create the starter pack for a business who doesn't know how to navigate this. and again as a standalone HR manager and a startup just trying to get through the day this would be very helpful and by the way LinkedIn compliments to them they did put together a string I think LinkedIn do a good job of aggregating these stories into a single post and they produced I think there's some very good things there didn't produce they curated it the community members themselves and there was some good stuff in there so do check it out okay and by the way brain food live we're going to do a special on this in a couple of weeks we're going to get martin redstone in we're going to get natalie farmer in um and they're going to just tell us okay for ta this is what it means uh so you've got to sign up for that and make sure you're there um okay uh what else is going on i was interesting last week uh aj um i'd probably say the notion report that was on the how AI reshapes work and skills in Europe.

9:37I think that was the one. I was looking at the different, the spread of the different countries, right, in the EU. And I must say that everything seems to be quite prominent in manufacturing. So it seems like an easy one. Me being based in Germany, I've seen that manufacturing definitely automotive, it makes sense, semiconductor manufacturing. But also to see how many governments are also using this for administrative support, which is very interesting because I think there's a huge disparity across countries how we are automating such administrative services and how this is actually going to affect the reliance of professional services, especially in the UK, as you pointed out in your commentary.

10:32What does that look like then for employment and where we have a shortage of skills and where we're investing our money? It's a really interesting point because every economy is obviously like structurally different. And where we put, where we generate GDP, where we have employment in different sectors it's very different per country um and if we agree that ai is going to have a variable effect on different sectors then it makes sense it's going to have a variable effect on different countries um like i say i think the uk being like heavily financial services heavily professional services um we stand to either be a boom or bust in this ai kind of economy on the positive side if ai does turn into be augmentative to a lot of these white collar jobs and i think uk is going to fly if it turns into something that is in fact substitutive then i think we're going to absolutely have a huge problem because we don't have a lot of other stuff um outside of you know being lawyers and accountants and and all the rest of it so um so yeah very interesting um a couple of bit of information on that yeah and what about that article from lenny rochitsky right so how tech workers are feeling in 2026 um i don't know what have what have you seen uh from the people that you've spoken to i know you took on a couple of roles yourself last year yeah yeah this year time really flies um i don't know what you were hiring for but what what was the sentiment um amongst the companies you're helping out i think it's it's very much what lenny mentioned was exactly right and it's this description of this k-shaped economy where there's You've got a bunch of people that are winning.

12:15And for them, life has never been better. More productive, more money, more compensation, more demand. It's fantastic. It's a small sliver of people. And then there's everyone else who's thinking, right, cost of living's gone up. My wages haven't gone up. My job's getting eaten by AI. I'm not getting any of that. How does all this help? Recruiters, by the way, tend to fall into that second category because we know that recruiters are working more jobs than ever before. We're doing more interviews than ever before. We're doing more candidate processing than ever before. We're actually placing more jobs.

12:50We're hiring more people. However, job security hasn't improved. Job volume hasn't improved. And comp has not improved. So the story for us is very much AI is eating away at the work. we're not getting the productivity benefit that's going to the company um so anyway yeah k-shaped i think that's happening in tech as well um you've got a a a load of people that let's say are working for ai adjacent or ai related companies bully for you going great um everyone else um they're also under pressure because companies are saying you know what uh do i really need a team of 50 engineers maybe I can do it with 25.

13:35And so they're getting that crunch down as well. So yeah, not easy, not easy at all. So what do you think about the future of the workforce then? I like the question where Lenny mentions that those who are really confident in what their professional outlook looks like right now, they're excited about AI, but they probably wouldn't recommend their profession to people just starting out. What do you say about that? it's a brilliant question i'm going to pose it in my poll tomorrow i do a poll every Saturday folks it's going to be look would you recommend your career um to junior you or a new person coming in like if you're an accountant today and junior person came to you student wanted to be an accountant would you say yeah go for it um same with recruiters same with everything uh are you recommending and endorsing your own career to a new person coming in i'd be interested to see the answers to that um yeah the truth is i don't think anybody's that confident right um you can't be overwhelmingly confident about where your career is at um so i've got to figure out how to measure that maybe we put some sort of grade point on it um to see what how confident are you to recommend a career recruitment let's say uh to a junior person and we get some uh some some info that way It's just going to be tumultuous in my view.

14:55I do think sort of headcount per company is dependent on some technological decisions, which we are already making. I mean, I'm going to explore this in depth later on. I think we're way too early for this conversation, including my own way of thinking through it. But I do think basically local AI is going to lead to more headcount. But cloud deployed AI is going to lead to headcount zero. So I think it's a fork in the road for us. But I'll explore that further forward. We've got smart people to bring on. Maybe they've got an opinion on it. Okay, let's go. So let's bring these folks on. We are talking about AI readiness.

15:37Whilst, by the way, our crowd, our panelists are coming on, I want to direct everybody who's watching this to the right-hand side of your screen. If you're watching this on Crowdcast, you can see a poll there. I want you to click on that poll and then vote on it, please. And let me know how you feel about that. The question is quite simply, have you conducted an AI readiness audit in your organization? Yes, no, and don't know. Don't feel ashamed to say don't know. Maybe that's happened and, you know, it wasn't. Someone's just done it. and why should you know about it? Okay, let's get Joshua on here.

16:14Who else do we have? We have James. James is here. Let's bring him on. We also have, oh goodness, we're trying to find Richard here. There's Richard. Let's bring him on. Guys who are in the gallery, try and fiddle with your mic and your camera. Make sure they're switched on. As soon as I see they're both on, then I'll move you on screen. We should have Oliver here as well. Although I'm not sure whether it's Olivier or Oliver. Let me see if I can IDM. Oh, there he is. Cool. I see Richard here. Let's bring Richard on straight away. Brilliant. Hey, Richard, how are you? Hey, team. I'm doing great today.

16:58Glad to join. Great to see you, man. Yeah. Whilst you and the guys are getting on, can you quickly introduce yourself, Richard? Who are you? What it is you do? Yeah, I'd be delighted. I'm Richard Roseda. I'm the co-founder and chief product officer over at Icona Analytics. Icona, we're an HR transformation consultancy, but under the hood, we're a knowledge management operations startup. Some fun things going on there. But my background's in people analytics. So I worked at Facebook, Uber, Nike, got to see some great teams, got to see some teams explode. Definitely coming at this AI space from that data-oriented perspective and definitely a little bit of the knowledge management side too.

17:30Yeah, fantastic. Look forward to getting your input on this. You made a really interesting couple of contributions to some posts we put online, Richard, which is why I wanted to bring you into this conversation. So fabulous that you can make the time for us. We have James Hutter here as well. James, great to see you, sir. Why don't you quickly introduce yourself? Who are you? What it is you do? Hey, everyone. Lovely to be here. Thanks so much for having me. My name is James. I run a consultancy called Paradigm Junction. We tend to talk about doing technology disruption. so we work with companies or bits of government when when they're realizing that this isn't a case of adding new technology to a process that they already run it's a case of maybe thinking a bit deeper and starting that redesign from the ground up or where they've got those big new competitors so they're seeing that other folks out there are taking taking their customers or taking their market and delivering it in an entirely new way so we I guess mix change management with some with some market forecasting and of course a lot of helping people build their own worldview to understand, well, what is this technology?

18:29How might I use it? And I've done lots in the recruitment space over the last few years. Yeah, fantastic to have you here, James. Great to have your thoughts on this topic. We have Joshua Hoffman here as well. Josh, great to see you. Can you quickly introduce yourself? Who are you? What is you do? Hey, Joshua Hoffman. Quick announcement to make. For the last five years, I've been the CTO at Sharpest, an incredible company. And I'm now ready to find my next challenge. So I'm actually rejoined the market. So if there are any folks who are interested in that, then you know. But yeah, I've been a tech leader, a CTO, first a software engineer before that for my career working at companies you may have heard of like Tumblr and SoundCloud in the early days and spent nine years at Red Hat doing professional services consulting.

19:16Just love technology, fascinated with all the new stuff happening. Great to be here. Yeah, fantastic to have you on the show again, Josh. And we have Olivier Vidal here as well. Olivier, wonderful to see you. Can you quickly introduce yourself? Who are you? What is you do? I love you using the French pronunciation of my name. That's perfect. Everybody who knows me knows I'm totally English. I'm a HR tech product professional. After the last 10 years or so, I've been involved in building products that match people to work inside organizations and in a recruitment context and in a freelancer context.

19:55And now I'm building a practice focused on AI readiness, basically helping corporates to get their internal data, information, governance sorted out to actually get workforce tools to work. That's me. That's brilliant stuff. I'm going to keep calling you Olivier, by the way. I hope that's okay. uh i like it better look it's the name my father gave me it's just got anglicized over time yeah brilliant let's start with you because that you came up again with some brilliant commentary when i started this conversation a couple of weeks or so ago and it just like we've got to start really from the from the brass tacks like what is ai readiness um let's not assume we share consensus on that um what's your definition ollie i mean i think i think we are as a market we are using the term ai readiness to mean multiple things at the same time and hopefully at a certain like at some point soon the market will decide like what terms we're going to use for what you know martin who you say you're going to bring on soon i think is on there's a club of people who We talk about AI readiness as something that relates to the skills, aptitude, plasticity of the workforce itself, right?

21:18The people. And there's that bit. Then there is the kind of organizational and enterprise level. Do we have enough knowledge? Is it well enough organized? Are our systems talking to each other so that enterprise AI tools work, so that you can manage the workforce? algorithmically or with like levels of optimization tooling or real-time tooling whatever it might be i i firmly sit in that club at least that's where i'm you know that's where i'm focusing my time and attention and then i think there's a very legitimate other which is more technical which is asking well if a company is building proprietary ai uh is that organization mature enough to govern and provide guardrails around sort of citizen builders and people who are vibe coding or creating tooling.

22:19And I think that is separate from the question of how do you use off-the-shelf SaaS products that are increasingly AI-enabled. So I think there are three buckets that sort of are all being described as AI readiness currently. That's my best interpretation. question. No, really good. Let's go round on this because I want to just make sure you just surface all of the ideas and what it all means. Richard, go to you. AI readiness, what does it mean to you? Yeah, I love this question. I'll agree with Ali too. The market is still very open on what that term is. So at Icona, we are a practitioner consultancy.

22:56We all have 15 to 20 years of HR subject matter experience. We're practice and practical oriented. I think the painful truth here is that AI readiness is good HR. I mean, it's just, it's the unfortunate part of this. Like I wish it was a silver bullet. It was something new we could add, but it comes down to, do you have an HR strategy? Have you documented your HR operations? Do you have well-impleted technology? Can you access your data? If you have those four things, you are more ready for AI as an HR function than most HR functions that are out there, which are still struggling to figure out, hey, how does my strategy impact the business?

23:30What are we actually doing on a day-to-day basis? have we written it down? Do we have the technology in place? Do we have the data for AI? And the real truth of it, like we're hearing a lot in the market from a lot of consultancies, a lot of tech companies that say, hey, AI readiness is the thing we're selling. It's vitamins, it's exercise, it's eight hours of sleep a night. It's the stuff we all know we should be doing and we haven't done. And so the good news is I see AI as a lever we can pull to say, hey, if you want agents and you want AI and HR, you've got to invest in the fundamentals. and you've been deferring that for years, board, CEO, business partners.

24:03It's time to invest in these fundamentals and we can actually go after AI then. But it's not something different than really, really good HR when it comes down to it. Yeah. So basically access to data, data in silos, what's the data quality, all of those kind of pretty old questions really that maybe we, I don't know, I never sort of solved. We never did solve them really. and maybe we're going to pay the price of that a little bit because we're walking into this AI world now with very poorly prepared information that we now need to go and actually get sorted. And if I could add, I would also say that's not HR's fault.

24:43That's not because HR doesn't care about data. That's not because HR is not at the table demanding access or demanding investment. It has been systematic deferral of things that HR needed for decades. And it's time to pay the price for that. And that's not HR paying the price. That's the business actually has to pay the price to say, hey, if you want HR to be agentic and AI oriented, you have to actually invest in HR to make that happen. Yeah, really good. So I can see already something sort of emerging from this AI redness, like redness for what? There is a there is a transition that we're going towards.

25:15And therefore, we need to examine what conditions need to be those preconditions that are going to give us the foundations to leap into this transformation. So there's a motion, there's a momentum. AI readiness is about we're going to be getting into this phase and are we prepped for it. James, good to go to you because you've dealt with a lot of, I would say, let's say, stereotypically AI unready organizations, right? We're talking about your interaction with governments and stuff like this. I mean, again, don't want to lambast the public sector, but I'm going to use them as a negative exemplar of AI unready here.

25:53What's your take on the question of AI readiness? Is it a lot of like just housekeeping that we need to do in advance before we get into anything too esoteric? Yeah, it's such an interesting area. I'm inclined to agree with Oli, but I think I'd frame it differently from that being about readiness. Those are factors that your business might have that are good predictors of success. They're factors like having really good access to talent or a deep balance sheet or really good supplier connections. But I'm not sure I'd think about them as readiness. And when people talk about readiness and talk about all of the things which I agree with entirely are predictors of whether you'll do this well, but then frame it in terms of readiness, I often think that it's being used in a way that says, well, have we got a bunch of things we can claim we need to sort out before we get going.

26:41And it presupposes that AI is something you need to be ready for, rather than it being a change that's happening in the economy and a change that's happening to the way people work, where you need to be harnessing yourself to that change, and in whatever way you can, starting to learn and build capacity internally. Because the simple truth is, you can get all of those things right. And if you do nothing to do with the incentives to produce work, the structures of information, who owns work in your organization, how you're changing your processes, then once you've got those things right, you'll be really far behind because there'll be a bunch of people in your team who haven't actually got hands-on use yet.

27:18So I guess I would say, what is the main thing you need to change in your organization to start using AI? It's to say, well, is my team one where I can identify how we've learned and changed that we do something? Have we looked at something where we've gone, we could do this better and made a change in the last six weeks? And if you've done it in the last six weeks, then ask if you've done it in the last two. And if you've made a change in the last two weeks, then say, well, what have we changed today? But if you're this type of organization where you say, actually, we haven't changed how we do anything.

27:48We have continued operating, but we haven't made a change in response to learning in the last six weeks, then that's not an AI problem. That's a culture problem. It's just being exposed by a now changing work environment. So I'd say this is much more a question about incentives, much more a question about willingness to change than it is about data foundations, because those you can work in parallel, but the sort of ability to change the way that your work looks is for me what being ready means. And it's something that we can all do starting with a conversation. Yeah, really good. Ed, Joshua, you've got to come in on that.

28:22What's your thoughts? Oh, do we have sound? Sorry, guys. I've got silence if you're talking to me. As soon as James finished, the audio dropped. I don't know what's going on, but I'm really sorry about that. Can you hear me? You guys can hear me. I'm happy to answer the question. Sure. Yes, yes, we can hear you. Yeah, my take on AI already is a little bit different. I think of it more pragmatically in the sense of, I want to bring this in. What do I need to do to ensure success? And for me, I see really three pillars there. Some have touched on it, but on one pillar, you've got just do people know how to use it?

28:59Do they know what it is, what they can do with it, and how to apply it to tasks in a productive and efficient way? that's really important. There's the compliance side. If you are in an organization where you have strict compliance, you need to get ahead of that with AI and make sure that you help people understand. Because as we know, I mean, I've personally seen companies where officially they have rules, strict rules on how personal data is handled, but no one follows them and everyone uses chat GPT however they would like. And then the third area for me is really one I haven't heard too many people touch on, but I think it's sort of the policy slash etiquette area.

29:38And just as when we rolled out Slack to our organizations, we needed to have some guidelines for people on what's the etiquette around using this. I see this as a big need with AI as well. And for me, that's going to be things like, as an organization, we do not encourage you to have AI read an email and respond to it that you have not read yourself. Right now, this sounds very simple. And Richard, I see you're laughing. But yeah, I mean, people just there are organizations now where AIs are having arguments with each other because no one's reading any part of these emails and they're just letting them fire responses back and forth.

30:14And, you know, or second thing, don't send someone a message you generated with AI you didn't read. I know, again, that sounds simple, but this is the kind of habits that are starting to creep up. And I think this creates resentment and friction and just weird things in the workplace that we can get ahead of with just a clear statement of etiquette. You know, this is what we do. This is what we don't do. And then a whole chapter of that for me as a technical person and working with software engineers is What is our expectation around when a non-technical person uses some kind of AI agent to produce a code submission into the product and hands it over to the engineering team?

30:57Like, we should be ready with some thoughts on how we're going to address this. Is it completely not allowed? Is it encouraged? How are we going to deal with this? Because it's a very new reality that a product manager can go home over the weekend, work with an agent, come back with a huge code commit and say, here, I did it. And so these are the things I'm thinking about. Yeah, really good. I mean, the etiquette issue, I think, is worth a separate conversation on because I think everyone immediately resonated when you mentioned this AI-generated reply type behavior. It's outrageous, isn't it?

31:32I mean, Joshua has disappeared, but I'm addressing him right here. But it is outrageous. We feel outraged as someone's obviously not read our hand-tight sort of effort. And they've responded with some AI gen. It's like, okay, why is that weird? And is that corrosive to culture? I would say AI etiquette sort of ruling is probably a sign of some sort of maturity of your business. So perhaps it's along the way to AI redness. But I love the contributions everyone's made here so far because we're all getting a different picture of where we're at. Particularly on James's point, like AI readiness is a lot of it kind of like displacement activity to stop you from doing it.

32:19You know, it's almost like, is it performative? You know, it's almost like, are you getting ready to do your thing when you just get out the door? You know, you're spending too much time preparing for something when, in fact, it's difficult to prepare for this change. and you're better off just jumping into it, perhaps how you respond. Okay, open question here really for everybody. Where do you think companies generally are? Is there any company type that seems to be ahead of others when it comes to just being sort of ready for AI? Whatever we mean by that term. Is it stereotypically technology companies perhaps are ahead?

32:58Are there any functions within companies that are ahead of others? And if so, what are those functions? Again, is it engineering ahead of others? What are your thoughts? Anybody feel free to pop into a response here. Yeah, I think software engineering teams that have jumped into this have been doing it for a while now. There are some that are quite experienced, maybe even up to a year worth or more. And I've seen this kind of go nowhere. And I've seen it also go to very high levels of mature operation with it. So some of my colleagues, we sit and we discuss this a lot. and we talk about how exactly do we use this within our teams.

33:34And the more experienced ones have systems. They know the constraints they're going to apply. They know how people are going to limit task concurrency and not get burned out because the AI can code so much faster, but you need the human to review. I really see software engineering departments that got into it early are doing well and are pretty far ahead of it. I see other teams really experimenting. But yeah, I'd love to hear what others have to say to you. Yeah, Josh, I love that. I think what we see in the market is that there's starting to be more variance between HR teams. And something a little funny about people analytics, there was a variable study that said people analytics drives profit margins.

34:14And if you did good people analytics, you'd have higher profit. I've always thought that's flipped. If you have higher profit, you have enough money left over to pay HR to do people analytics. So what we've seen in a lot of people analytics functions is when you have a high profit margin, you tend to have a great people analytics team, which is sort of the rich get richer over time. We're seeing something pretty similar in AI, where the frontier labs, the data warehousing companies, the high tech companies, they've invested, they've worked with it with their software engineering teams and their HR teams are picking up that momentum and that tailwind.

34:44We're seeing a little bit of lag with the pharmaceuticals, the manufacturings, the different functions out there that know they need to do this, but they need to kind of leap ahead. And that's where we work with a lot of those teams that are kind of in that, hey, we've maybe missed the last era jump forward. We're going to try to leapfrog to this one. And there's opportunity there. So I think if you haven't invested until now and you're thinking, hey, it's probably time for me to get on this bandwagon, there's probably some lessons learned you could pick up from some of those early adopters and you could skip some painful moments.

35:13So I don't think it's too late, but I think we are starting to see some compounding effects of teams that have been on AI for going on four years now since ChatGPT3.5 came out. If you've really been investing since then and learning this since then, you're starting to build up enough of a momentum and headway that it's going to be hard to catch up soon. Yeah, I think... Go ahead, AJ. Yeah, I think also you need to think about, well, what actually predicts the readiness, right? So does the function already have a validation culture? So clinical validation, it could be financial auditing. are the errors that your department or function make cheap and reversible?

35:56And is there any feedback signal as well at all? So hiring is kind of a pathological case just from a talent acquisition perspective because you never observe the counterfactual performance of the candidates you rejected. So a screening model can be badly wrong for years without anyone noticing. And then also, what is the coherence of the data kind of estate, right? Like which this really correlates in with the maturity of your organization, how old or big the company is more than the industry. So I think what are we measuring ourselves against? And yeah, this can vary a lot between industries.

36:37I think that feedback measure is something I'd pick up on. If I was to pick the best predictor of whether someone or a team is going to be really heavily using AI tools and to sort of take use as a proxy for maturity, because until you get to the very high end where you can certainly use it inefficiently. but it is that feedback mechanism and its incentives that line up with getting things right and doing things quicker. I think there's this assumption that we all exist in this world where doing your job better is somewhere between sort of lightly encouraged and massively rewarded. It's very easy at the top end to think about people who are paid almost entirely on commission or to think about the massive rise of one and two person businesses that we've seen over the last few years clearly as a response to these tools.

37:25But it's actually at the less willing end that I think it's really interesting to ask the question. There are lots of places where people are penalized for doing their job more efficiently, for where it is seen to be harmful, not playing with the team, tool poppy syndrome, putting burden on other teams. There are lots of places where it is anti-incentivized to be making yourself much more efficient because you create downstream pressure for other people. And so I think, Hung, you mentioned work with government, I think one of the things that comes up time and time again is, well, if I work out how to do this in a new way, is that something that's going to be celebrated, championed?

37:59Am I going to be given any reward, not necessarily a financial one, but a praise reward? Or am I going to be treated with suspicion? Am I going to face all of the downside risks if something goes wrong and none of the upside risks? Because you can pretty clearly tell that someone's not going to adopt in those worlds. The most surprising success story I've seen was actually a small policy consultancy who um it was a fairly settled team they didn't change how much they were what they were doing much we worked with them about 18 months ago and they they knew very little if i'm honest and then one of the directors said that he wanted to try and start his own thing and they took him down to two days a week and he spent the other three days a week trying to start his own business and all of a sudden when when he was getting the profit motive you better bet that he learned how to use these tools and advanced his personal maturity massively and he's then been a really good function for not only bringing tooling back into the organization that they can lift and shift across and adopt, but then for diffusing that internally.

38:55I think it's a nice example of where when you're in an organization where you're being assessed based on how many hours you do doing the things you already do, then spending the time and effort to work out how to do something differently isn't rewarded. Whereas when you're in teams where you are getting that upside as you change, then pretty quickly people find incentives to teach themselves. It's not something that you necessarily need top-down teaching on. Yeah, very interesting. And I think also it kind of points to a few ideas that have been circulating around that maybe the optimal adopter of AI is that solopreneur type figure because they can do so with less constraint and also because they can make these cheap and reversible mistakes, Amanda, as you mentioned, because who's on it apart from them?

39:45Whereas if you're embedded and enmeshed in some sort of bureaucracy, then any decision you make may have all kinds of ripple effects that ultimately cause you to make very conservative decisions because the cumulative effect of all those relationships means that you can't really do much. So, yeah, really, really interesting. Oliver, let's bring you on this. Someone's raised their hand. Don't do that. It's me. It's me. It's all right. I'm starting to feel lonely. It's all good. Yeah, just step in, Oliver. Don't be polite here. I think it's clear that the term is so broad, AI readiness, that it's just difficult to navigate.

40:30At risk of adding another layer to that. I think within the conversation, people have talked about individual level changing how they work. And maybe your initial question was departmental, like how do you change? And maybe even industries, which ones are changing fastest? But I think from what I've read, and I can't quote these things verbatim here, but essentially there's a school of thought that says, yes, we can all get a degree better at doing our job by using AI to take over certain aspects of our role. Fine. James hinted at it. that the real potential breakthroughs are likely to come when we don't think about AI readiness as being within the constraints of our jobs as they stand today, but instead looking at how workflows are redesigned and probably orchestrated by AI.

41:33So AI is not just in the execution of individual tasks, but also plays a part in driving workflows and managing that handoff between human and agent in such a manner as the business creates value doing whatever it does, whether it's call center handling or whatever. It doesn't terribly matter. so I think the reason I mention this is I think that's the tipping point where AI readiness goes away from being cleanliness and vitamins as Richard was kind of alluding to and it much more becomes a kind of a category one problem that if you want to unleash the potential gains as a CEO of AI, you have to orchestrate man and machine.

42:30And as soon as you want to orchestrate man and machine, you need to know a whole ton more about your people and what they can do and what work looks like. So I think it's agents and the kind of excitement about getting them involved in workflows that switches AI readiness from being a cleanliness type conversation through to being like absolutely critical to a company competing. That's my little rant. Was that a good rant? I don't know. Did I go off on one? I don't know. Let's wait for Oliver's rant. Please comment below. What do you reckon? Do you need to hear more rants of this type? I think so.

43:12No, brilliant stuff, Oliver. I want to shift attack a little bit though, because people listening to this thing, this is all very interesting, but how does it help me tomorrow? um so i'm let's say the persona i'm imagining is somebody who's you know hr person ta person in a smallish type company they want to be able to say to their board this is the state of our company in terms of ai readiness they don't know where to start so can we give that person as some sort of like one two three four step a sort of approach to say this is how you get started with understanding your current state of AI readiness.

43:53I'm just tempted to say, is it just a survey you sent out? It's probably a bit piss-boot week, isn't it? I mean, is there a better way to do it than that? How do we assess AI readiness? How do we get started? Richard, your thoughts on that? Yeah, I love this question. It's one we think about a lot because surveys tend to assess self-confidence, they don't tend to assess actual work on the ground. So it's what do I feel about AI readiness is different than how ready am I? At Icona, we really lean heavy into a qualitative interview-based methodology. And this has really radically changed how I approach my understanding of work.

44:32Because I come from a deep people analytics background, very quant, very, hey, what do my systems say? How do I read off? Without talking to people, I just talk to the tech. We sit down, we talk to 40 or 60 people across the HR organization, across legal, compliance, system, security, different areas. We bring that all back together in a conversational stories. And it's a bit funny, like the newest thing we have going on with AI, we've got to go back to our oldest technique, which is you have to talk to people. And that's hard and it's difficult to do at scale. But the richness and the fidelity of that data stream is unparalleled, especially when you're dealing with something that's not predictable from historic.

45:09I can't look back through my data history or my past company performance to say, what can I do to predict AI readiness going forward? This is a fundamental change in how business operates. And to do that, you've got to get out of your office. You've got to go talk to someone. You've got to get on the Zoom calls. You've got to have those conversations. I don't think you could do it another way. And so I'm very much so a proponent of go back to the basics of get out there, talk to people, find out what's going on. And it has a secondary impact on AI, which is these AI tools understand outside of your company.

45:40They don't understand what's going on inside the firewall of your company. That stuff's not been written down. And so as you're doing these conversations, really keeping better records, building context layers, starting to store this in a way that's machine readable has a double play there where not only can you talk to people to figure out the readiness, if you document it properly, you're starting to actually take that first step, which is prepare tools and prepare AI systems to work off of that context. Right. Really good. So basically you've got to do the work. You've got to speak to the people.

46:11You've got to do the interviews. You've got to record the interviews, make sure they're machine readable. And then you can go back and you've got the qualitative and the quantitative at the same time. And you can pull out insights from that experience. Yeah. What question I think? Go ahead, Richard. If I have a theme here, I don't think there's an easy way out of this and i think especially when like a massive change comes to the enterprise you have to do the hard work which is which is hard and that's why there's so many phenomenal consultancies and groups popping up too to help out is like there's a lot of help that's needed right now to make this happen all right one one final thing from you richard what is the what's the question that's the zinger what what gets people talking about this like if i was to ask the one question that makes sense if i'm going to do this what is it man i you know what's funny is I did an interview the other day.

47:00And at the end of the interview, the person said to me, hey, I thought we were going to talk about AI. And the 25 minutes I was on the call, I was talking about AI. But I was talking about, hey, does your tech team talk to your HR group? What's the last time you got friction in a process? Have you written down how you do recruiting, step to step, process to process? And they're like, well, when are we going to get to AI stuff? This is all of that fundamental piece. And so unfortunately, I think, again, the marketing noise is so loud right now that it can feel like, hey, I'm missing something. There's something new I have to do.

47:30There's something different. And it comes down to I think it comes down to fundamentals still. And when you actually document, understand and figure out your business, these things get easier. And if you haven't spent the time doing that. So a particular zinger, what's the last time you had a hard time doing a process? and did your process get cheaper the second time you did it or was it as expensive the first time? It's not a very well-articulated one, but there's something good there. I think those kind of things, you can really find out what's happening within your business and you can figure out where you can apply AI.

48:03Really good. So the takeaway there is you do have to do interviews, basically, old school, as you say, very old school technique, but human-to-human conversations, never underestimate those. James, you've got your hand raised twice there. So I assume you have an urgency to make a contribution. So go ahead. I fear it might actually have been number two after Oli, but I'll take the opportunity to it. I can't read the numbers. I don't know what they mean. So it looks like you just like massively need to say something. So I've gone to use this. This is where I've got to spend my points. Now, I want to agree with Richard.

48:37I mean, the first thing we'll do in any organization we speak to is some interviews. And those interviews will also not feel like they're about AI because it's become such a hot button issue that people will tell you what they think they're supposed to say unless you can find more creative ways to ask them about how things are going. Often the blockers come out when they'll say, if you ask things like, if you told your boss that you'd produce this with AI, how do you think their reaction would differ? And then you can start to get into some of the attitudes that are shaping behavior. I'm inclined to agree that surveys don't have a huge amount of value.

49:10And the question that we'll often ask people on this, this links back to what I was saying earlier, an organization that is able to change things and to learn is an organization that will start adopting AI because it is the best solution for quite a few types of change at the moment, not all, but quite a few. And so if you were reporting to your board, I would say, hey, let's not get distracted by the technology we're using. But what I'm going to do every time you have a board meeting is write down the list of changes that we've done and what we've learned from them and whether they've got easier than last time.

49:39And I think if you're reporting, how are you changing what you're doing, you'll find that you're starting to make some changes that include AI. If you're focusing on the technology first, then you'll probably find lots of reasons why that change isn't the one that you can actually do first. Really good shift of focus. I love it. And yes, the numbering system is about ordering. So I do apologize to Ollie here. It wasn't insistence. I've just figured it out. I've been educated, so let's call that a plus. Oliver, your views on this. um how do you go about starting from zero um this person wants to get going with assessing ai readiness in their business i think it's i i loved richard's delivery and you know i'd definitely buy services from him but i actually think it's too big a question it's almost like such a big question is to be unhelpful and because it can go in a million and one different directions right it's just it's too damn hard um ai readiness is is a million and one things going Back to the point I made in the beginning, it's just so many things at once to almost just be meaningless.

50:42It's like getting a strategy consultancy. I think it's an awful lot easier to start with a use case where you want intelligence to be applied to a part of how you run your business. So we want to be ready to do automated resource management using an AI tool. We want to be ready to do automated workforce planning. like it almost it doesn't bloody matter which of those you begin with because a lot of the foundations are shared and as Richard was alluding you know the information architecture that is required in a business to drive AI that's kind of like a universal problem that that information there has to be there but if you try and do it in and of its own self without it actually being attached to a product or a user experience that you are invested in.

51:36It's just too damn hard. That's my personal take. But if a board has said we want to be more efficient in how we use our workforce and therefore we're ready to bend over backwards to get ready to do that, then awesome. then there's a much more narrow set of questions you can ask yourself. No, I love that, Oli. You're basically saying let's chain it down to a practical use case that people in business will know about. I mean, if you're working in a business, there'll be a thing that everyone understands, yes, that's a bit of a palaver. Can you then use that as your first experiment and just get away from the, chain the theory down to a practical use case that everyone understands?

52:22And somebody's probably bought a really expensive SaaS tool for like hundreds of thousands of dollars per year. Somebody signed it off and then they basically said, oh, my God, we can't actually make this thing work. Like, unfortunately, I think that's how a lot of AI readiness will come to actually pass. Yeah, I do. I do get that vibe. I think AI readiness may ultimately be something that is a little bit of a performance. We might need to go through that performance somewhat. but the tidal wave might just wash us out anyway. It's like, okay, just learn to swim is probably better than preparing a flood defense, so to say.

52:58Okay, Amanda, you want to say something? Yeah, just very short. I think some things, especially I like the use case point, something else you could think about is also within your own function, the override count per system. So how many times in the last week, in the last 90 days, did a human have to override this process? You know, what percentage of these systems actually has an accountable owner that can monitor this? Because often we implement a tool and we kind of just run with it. And then that also comes down to a time to detect failure kind of indication of how long between the time you implemented this tool to the first time that you or someone started noticing the errors, right?

53:49Because you kind of think, okay, it's going, people are using it, that's great. But is someone keeping an eye on it? And I think this is where the documentation is also very important. So if you have that basic framework in place, maybe that's a good place to start. Yeah, that's really interesting, Amanda. Do you have that type of framework in place, first of all? That's probably a sign of error-edness, So we might agree that an AI unready business would be someone who just buys a tool and deploys it without having some sort of mechanism in place to do the review and the override. And then the second aspect of that, if you do have that person or that system in place, then what happens with the frequency of overrides?

54:33And is there a trigger at which point you then review the usage of the tool? that's a sign of maturity even though we can understand that the incentives are probably not aligned for that system to currently be there because everyone wants the tool to just simply work and it may well be that you know if you ignore it a fact that it doesn't work then it you can sell that it does work um so um there is an element where the product itself or the tool itself once it's been bought it will it'll have its advocates irrespective of its function simply because vested interest has gone in, social capital has gone in, political capital has gone in.

55:11Joshua, your view, your advice on someone who wants to start about assessing the AI readiness of their business, don't know how to start. What's your input? Yeah. So, I mean, a lot of the stuff we're hearing, I think for those of us with a little gray hair would remember if we talked about big data readiness at one point in the past, very similar conversation, right? Are your systems well managed and organized? Do you understand your inputs and outputs. For a lot of companies, they're still not there. So this is really, to Richard's point about tech hygiene, all that stuff, I'd take a different approach for my advice.

55:46I think the most powerful thing a company can do today in any industry, but especially a company full of office workers, is actually take an L &D approach. And what I mean by this is bring in an expert, someone like James, someone like Richard who could come in and give a workshop to all of your employees who touch technology and to teach them, really choose a tool you're going to deploy. So for example, my company, we deployed a tool called LangDoc, which is just a European platform which lets you build workflows. I mean, there's a million ones to choose from now, but pick one, show people what it is, teach people how it works, and then put them into groups and challenge them to come up with ways that they could apply it, do presentations, rate them, just approach it as L &D and turn your workforce into well-educated AI experts in a way who know what it is, know where the rough edges are, know what it can do today and can start thinking about how to apply it to their own role and where the easy wins are going to probably be found, where the big organizational things are going to get talked about.

56:54Because then we've created a workforce that's AI literate. And when we start talking about the big organizational adoption and readiness, about bringing in those company departmental crossing procedures and policies and workflows that we create that go beyond just one person. When you have a company full of people who understand the mission and what you're trying to accomplish, they are going to be much more able to help with that, to solve problems quickly, to understand what to expect, to just know how to deal with it and know how to think about it. Because I think for a lot of folks, this is the gap.

57:28At least this is what I see. And a very powerful effect happens when you close that gap. Yeah, and that actually loops us back to recruiting and L &D and HR stuff, doesn't it, Joshua? Improve the human capital of your business on a general sense. And probably your AI readiness will increase, I guess. I mean, that's not bad. That's probably true. So it points back to upgrading the talent level of your business. um very very interesting um brilliant stuff obviously more to talk about on this topic i see on the chat stream as well some very interesting like sidebar conversations um about you know ai feature rollouts on existing products how we meant to assess them truth is we ain't um right i mean the the product's gonna say hey you got this new thing we're excited we're gonna say upgrade or install off we go we're using it uh how do we even figure out uh that's gonna be good to go and in fact if it is not good to go where does the liability fall um is it purely on the vendor or does it go to the operator didn't i read somewhere recently that it's going to fall into the operator as well as the vendor as well so it's not like you know um so every every single thing that gets updated suddenly uh needs to be reviewed auto-reviewed even before we accept um And that's annoying, but it's probably going to be more recruitment work down the pipe.

58:54Okay, that's about it, folks. We've got to leave it there because we're coming close to time. So let me thank everyone for joining. Richard, brilliant to see you on the show. Joshua, great to see you back. And good luck on the job search, by the way, unless you and I have a separate conversation on that, see what it is that you want to go forward and do. Olivier Vidal, what a great name. Is this man sort of misdirected in this career? Should he be a fashion designer somewhere? I think so. Great to have you on the show, Olivier. James, great to see you and your contributions here. Brilliant as well.

59:29Everybody have a good weekend. AJ, hang on for a second. You and I are going to end the show together, but we'll say goodbye to Richard, Olivier, Joshua, and Richard right now. Okay, folks, that's about it. I hope you've enjoyed it. Do bear in mind that we are going to be back next week. I think we've got a massive triple header next week, if not a quadruple header. There may be four Brain Food Lives happening next week. I know it's absolutely nuts, but we have to do them for a number of reasons. So let me just list these out for you. Firstly, we have a live demo of Emma, which is a new product coming out of the US.

1:00:06They're going to be your AI agent for HR, IT, and finance. Surajit Chatterjee, CEO of Emma, is going to do a live demo for us. This is the story of midweek live demos, by the way. CEO has to do it. Not interested in a salesperson, has to be the founder CEO. They're stepping up. Make sure you tune into that. It's also National Interview Week next week on Wednesday. I don't know where you know about this, folks, but the UK has got together. We're going to try and get 100 ,000 people jobs. How do we do that? We're going to just teach people how to pass interviews. And we've got a load of people coming in onto the show to talk about this particular campaign, but also to give interview tips as well.

1:00:47That's happening on Wednesday. Then on Friday, we're going to be talking about from talent acquisition to organizational performance. And this is where we're tracking the very interesting career path that we're seeing from a number of our community members that have expanded scope. I'm just thinking, oh, we're hiring external talent into our companies, and then we're going to hire more external talent to thinking about what is that external talent doing when they're internal talent. Can we think about organizational capacity? I was starting to see TA folks sort of expand the scope to take responsibility for that, which is an exciting direction for us, I think.

1:01:26Anyway, that's it. Make sure you follow the channel if you want to watch more of those shows and register independently. And each of those will be back next week for a massive marathon session. Okay, that's it. Thanks for watching everyone. Yeah, have a good weekend. how was that for you amanda you enjoy yeah it's a fun conversation i think we're all caught in the thick of this um and what what does it mean it does i mean we're all like in the mixer i call it offline of course not um no one's watching anyway um de facto offline um yeah listen really good having you on the show brilliant co-hosting by the way you're really good um and yeah we'll get you back with your fancy doing more of it i mean as guest or co-host as you prefer you know yeah it's really i like this group actually because um i mean i've seen more and more of this i think there was a reel i watched about um maybe you even shared it uh about the mckinsey consultant who was an ai expert right with um is it obi obi or some job search tool and it's like I'm not an AI expert it's like well now you are cool thanks McKinsey and you know there'll be a hundred thousand dollars for like an opinion on something I have no knowledge about so yeah sometimes I feel like that's what we're expected to be a consultant to actually have no idea what we're talking about but you know what the way I've dealt with this and people always say have you ever felt imposter syndrome i said listen mate i i am an imposter um there is no syndrome um you know i mean i literally don't know half the stuff i'm talking about and i don't pretend you know i don't pretend i'm an expert so i'm never caught out because no one ever thinks i'm an expert i say fair play me i'm just but you do a really good job of posing i think i'm giving a hot take and i can ask a question and that's it so that's my skill set um but you've got that You should do more hard takes.

1:03:28You should do more hard takes. I mean, you don't have to sit in the tube with, you know, your transport card as Mike. But you should do more hard takes, I think. Just pulling people aside. I need to revise my TikTok, which is a Mori Budden TikTok. Last time I posted on there was 2021, I think. So it's five years ago. Revise that, and I might well do a daily, you know, here's what I'm thinking. And then if people don't like it, don't like it. You can be a conference or meetup creditor where you kind of just walk around and randomly pull someone aside and ask them, put them in the hot seat, be like, hot take, what's your take?

1:04:05The hot takes, why not? Let me know, folks, if you want me to do that, I'm happy to do it. All right, AJ, listen, I'll let you go get a cool bath or something, and I'll catch up with you later. Indeed. Have a good weekend, mate. Cheers. See you later.