Annabel Gillard - AI Ethics & the Financial Services opportunity

00:00:00:00 - 00:00:07:09
Unknown
Hello, and welcome to the Impossible Brief. This is a podcast about leading when technology moves faster than organizations can change.

00:00:07:14 - 00:00:11:23



00:00:12:10 - 00:00:19:16
Unknown
My name is Ben Taylor. I'm joined by my co-host Eleanor. And we're joined today by Annabel Gillard.

00:00:19:18 - 00:00:21:07
Unknown
much for having me.

00:00:21:09 - 00:00:29:21
Unknown
Hi there. Annabel. So you're, you're an AI ethicist. You've got a master's in AI and philosophy and 20 years experience in asset management.

00:00:29:21 - 00:00:33:24
Unknown
And the investment industry, which is, an amazing combination of skills.

00:00:34:05 - 00:00:41:19
Unknown
And can I just ask you to explain to us what's what's an AI ethicists?

00:00:41:21 - 00:01:05:14
Unknown
ethics is a field that probably started about ten years ago, maybe a little bit less, just as the as the power of the new designs of AI started to become evident. And, and some particularly, there were a number of, particularly European academics who have a computer science background, but also some philosophy grounding who started to think about, hang on a minute, what does this mean, a human level?

00:01:05:14 - 00:01:35:23
Unknown
What's this mean at a societal level? What does this mean about the nature of what does what constitutes knowledge? What's consciousness? What's experience? Or all of these quite, quite big, big questions. And then as, as I, you know, as, as, as, as I this version of AI that we're experiencing now, of course, AI is not new. It goes back to the 1950s, but but the, the machine learning and the and the and the deep learning in particular, following, you know, the big data, revolution of the, of the noughties has meant that, you know, it's all things taken off.

00:01:36:04 - 00:01:56:23
Unknown
And so some of the early mistakes that were made, such as, you know, bias in data sets, I'm not saying that's a completely fixed issue now, but I think there's widespread awareness of it. Whereas that at that forefront there wasn't. And it was it was those, academics who were working at that intersection of, who had a computer science or an AI background and who then got into philosophy, who were the first to kind of identify on hang on a minute, there some human impacts on this.

00:01:56:23 - 00:01:58:09
Unknown
So that's where I think started.

00:01:58:13 - 00:02:09:18
Unknown
I think why we all today, AI ethics can mean anything from governance to, writing papers about consciousness to debiasing algorithms to anything else.

00:02:09:18 - 00:02:15:17
Unknown
it's one of those job titles, which is starting to spring up in, different organizations, and it means different things in different places.

00:02:15:22 - 00:02:33:01
Unknown
And I think probably do want to distinguish between what role is there, which bit is the governance role, is that kind of making sure that, you know, we've got the voting process in place, you know, who's doing the actually the upstream design thinking of we've got somebody who's in in the design phase who's thinking, hang on a minute, what about the human impact?

00:02:33:01 - 00:02:39:17
Unknown
Or have we thought about this as somebody who possibly isn't from a, a technical background, but actually brings that humanities perspective?

00:02:39:17 - 00:02:45:02
Unknown
So and for me, what it really means is it means thinking about the human impact of AI.

00:02:45:02 - 00:02:49:12
Unknown
what I mean by it is, is there is a kind of ethos of the business, the how are you doing things?

00:02:49:15 - 00:02:56:16
Unknown
and are we making sure that our actions are true to our intent? That's really what I think of as a, as ethics in a business context.

00:02:56:21 - 00:03:07:06
Unknown
So it's, business ethics first and should it should influence and impact your AI ethics in theory. Yeah.

00:03:07:08 - 00:03:34:23
Unknown
It's exactly, and I think the issue, the reason why it's, it's a it's it's been sort of an explosive field is that AI in itself introduces new ethical dimensions to some of the questions that, you know, it's not just business ethics as usual. It's different in scale is different in speed, is different in impact. And it amplifies some of the things that are already there, but introduces some new issues and also the issues that are relatively well contained when it's a human in charge.

00:03:34:23 - 00:03:45:24
Unknown
shouldn't assume there's any moral reasoning when AI is is doing a task that was previously done by humans. So sometimes we don't even recognize that there's a moral or an ethical or values dimension in a task.

00:03:46:01 - 00:03:47:13
Unknown
Yeah.

00:03:47:15 - 00:03:48:03
Unknown
it's only when you

00:03:48:08 - 00:04:00:24
Unknown
automate it, and then you and then something goes wrong and you realize, actually, there's an assumption in there that we had when a human was doing it that you can't just, you can't just automate it and expect it to be a replica because it's it's missing that fundamental kind of judgment.

00:04:01:04 - 00:04:06:23
Unknown
Yes. Yeah. So it's, business ethics first and should it should influence

00:04:07:02 - 00:04:10:13
Unknown
and impact your AI ethics in theory.

00:04:10:15 - 00:04:12:02
Unknown
Yeah, exactly.

00:04:12:04 - 00:04:16:12
Unknown
Well, I think it fits in, in, in probably two ways.

00:04:16:16 - 00:04:28:24
Unknown
if we adopt the sense of, of governance that where we are with AI is that we've had this hugely powerful, you know, transformation that's that's hit all companies in the last kind of five, six years.

00:04:29:01 - 00:05:02:03
Unknown
And we have a, a disparate regulatory response to that. So, you know, application. Well sort of technology regulation in the EU, no regulation pretty much in the US. And then outcome based regulation here in the UK. So you've got this sort of disparate range of, of regulatory features. But in practice, in honesty, I think regulation is really limited in its ability to give us the safeguards that we need for AI because regulation is is slow moving and specific and backward looking and jurisdictional and AI is none of those things.

00:05:02:05 - 00:05:03:24
Unknown
Yeah.

00:05:04:01 - 00:05:20:03
Unknown
And so if we actually want to get a kind of responsible AI that people feel safe, and trust, build trust in and they're happy adopting, then we need effective governance. That's that's agile, and that's, and that's practical. And, you know, and to develop a market around that,

00:05:20:03 - 00:05:20:08



00:05:20:08 - 00:05:30:00
Unknown
because areas such as insurance and assurance, are and actually particularly in the UK, areas where we have a real strength, we have a kind of global leadership, actually.

00:05:30:00 - 00:05:44:16
Unknown
And so insuring against AI risks is one, place at which we can start to ask the questions about, hang on, this is this, you know, is this data process up to standard, or has this been designed with enough rigor or what are the what are the backstops or can we explain this or whatever the question might be.

00:05:44:19 - 00:05:57:00
Unknown
So you can choose to get assurance for yourself, you can choose to insure yourself against others. And then the investor community can ask some of these questions and exercise governance as part of their shareholder due diligence.

00:05:57:02 - 00:06:13:14
Unknown
So I actually think financial services is one of the places where there really is some, some power, some kind of, you know, system wide power to actually help us move forward, and innovate in a way that's also, got some kind of safety and trustworthiness built in.

00:06:13:20 - 00:06:24:23
Unknown
Yeah. And, and so. And how are you applying this to the world of financial services? Where does that fit in?

00:06:25:00 - 00:06:29:11
Unknown
We want both.

00:06:29:13 - 00:06:54:08
Unknown
I think leaders are in a fantastically difficult position at the moment to balance all of these competing. I would say perfectly, perfectly expressed. Yeah. And impossible position because you're balancing all of these competing claims in an area of, you know, quite significant uncertainty. And with an environment that's moving at speed and the stakes feel existential. So it's incredibly hard at a human level to find your way through that thicket.

00:06:54:10 - 00:06:58:05
Unknown
Impossible. You might say.

00:06:58:07 - 00:07:09:01
Unknown
What I would say, though, is that while we do need to innovate, and while you can't, you don't want to be left behind. I don't think you want to be rushed into doing stuff that you you don't know what you're doing either.

00:07:09:06 - 00:07:19:11
Unknown
you know, one of the things I've just been reflecting on recently is that normally when there's with a sense of urgency inserted into, let's say, an email you receive, normally it's a sign to actually stop this.

00:07:19:13 - 00:07:27:04
Unknown
If you have your cyber training, one of the things they look out for is a sense of urgency, because it stops you doing your critical thinking. And I think we need to be very mindful of the fact

00:07:27:04 - 00:07:28:03
Unknown
Yeah.

00:07:28:05 - 00:07:37:18
Unknown
that if we rush a decision we haven't understood, we may not be able to want to do it. And taking a little bit of time to be intentional can quite often you quite often get the time savings back later.

00:07:37:22 - 00:07:51:15
Unknown
You know, if you've rushed into something and then you hit into it, you've hit a wall of problems later, and then I'm picking them. That can take much longer than having actually just spent 2 or 3 months upfront, being intentional about what you're doing, building in the right, framework. And then of, of you go to the races.

00:07:51:17 - 00:07:55:01
Unknown
The analogy that's often used in governance circles is,

00:07:55:01 - 00:08:05:12
Unknown
Why do Formula one costs go so fast? It's because they have the best brakes. And in fact, the better your brakes, the faster you can go. But if you just set off without having check whether your brakes work, that's reckless.

00:08:05:12 - 00:08:10:04
Unknown
You know, you might get away with it, but you really might not. And the cost may not be one you want to to pay for. And

00:08:10:06 - 00:08:11:02
Unknown
the other example

00:08:11:02 - 00:08:24:11
Unknown
making sure you've got someone in the room at the beginning, whether it whether you're doing product design or whatever, whatever, whatever way you're using the AI, when you actually at the design phase, having somebody in there, who isn't a technologist but is, a human specialist or an ethics specialist,

00:08:24:13 - 00:08:41:07
Unknown
Amazon, we're looking to get more, more talent into their, into their digital, at a point where they were growing faster than, than, then they could necessarily find the skills for and they thought, brilliant, why don't we get AI to go control the internet and find people for us? But without having specified upfront in the design protocols.

00:08:41:12 - 00:09:04:24
Unknown
But gender shouldn't be, a criteria. The Amazon book taught itself that gender was a critical criteria. And so even when they then, they then when they realized this was happening, they then started feeding in, you know, anonymized, CVS, but it was still using proxies, for gender. They spent, I think, something like 18 months and $3 million before having to retire the whole thing completely and start from scratch.

00:09:05:05 - 00:09:14:11
Unknown
And that would be an example of where actually having had, you know, an extra four weeks in the design phase would have saved you 18 months later, later on.

00:09:14:13 - 00:09:25:10
Unknown
actually is a great example because so many H.R. Firms are being sold to CV screening, AI tools back in 2018, 2019, 2020.

00:09:25:12 - 00:09:41:13
Unknown
And and incorporate them into their business. And at this point the system is now completely broken because not only have they been found mostly not to be doing what he had thought they were doing, and in fact, you're actually getting a really, that the algorithm is actually converging everything to the mean and you're getting one, see, that comes through.

00:09:41:15 - 00:09:53:11
Unknown
But also candidates. Now, just as the CV scanner can scan thousands of candidates, candidates can now send out using exactly the same technology, send out thousands and thousands of CVS. CVS are written to be optimized by the algorithm.

00:09:53:11 - 00:10:01:09
Unknown
if you'd actually started back in when it was 20, 18, 19 by thinking what is the problem we're trying to solve and how best do you solve that problem? And then

00:10:01:14 - 00:10:08:22
Unknown
would you use technology to help design the probability myself, rather than just, we've got to buy this tech and everyone else is doing it, so we better move fast, which is in practice.

00:10:08:22 - 00:10:09:08
Unknown
What happened.

00:10:09:12 - 00:10:27:06
Unknown
So I think it's a behavioral tendency to feel like where is the crowd going? I've got to be there ahead of it. But that actually rarely is the answer that that works out. I think, you know, we have to start and I think there's increasing awareness that we have to start with the idea of what's the problem with ourselves, what is it we're trying to achieve, and then how do we how do we want to do that?

00:10:27:06 - 00:10:52:15
Unknown
And then where does AI fit into that? How do we best design the AI? And also where do the people fit into that? And not just thinking about this, solely as an AI problem, but saying this is how do we how do we transform our business into something where the AI, the artificial in the human intelligence work together in a, you know, really effective and satisfying way to deliver, USP or to serve our customers as best as we possibly can.

00:10:52:15 - 00:11:07:16
Unknown
That really should be the goal rather than what it can feel like at the moment is FOMO on steroids. And I get it, because at a human level, it's unbelievably difficult to be trying to make these decisions at pace in an area where most likely you don't have expertise yourself.

00:11:07:18 - 00:11:12:12
Unknown
And so the people who you're listening to with the technical specialists, but the technical specialists are the ones pushing the technical solution.

00:11:12:12 - 00:11:29:09
Unknown
Actually, this isn't just a technical problem. This is an organizational problem. This is a leadership problem. And so that that that critical thinking, taking the time to think, you know, is really valuable. I think even if it feels incredibly difficult, you know, not managing that, that internal that that pressure that you're under.

00:11:29:09 - 00:11:44:20
Unknown
that's a place where services could help, because if markets were able to rival not to be indirectly applying extra pressure to leadership teams, which, you know, so I think boards can help as well, if they can try and create that space for leadership teams and to withstand that kind of external pressure

00:11:45:01 - 00:12:04:08
Unknown
to rush into what is probably what is, at the very least, a once in a generation moment. And it might be a once in a century, once to humanity moment. That isn't something you want to do with a shortcut because you felt like, you know, you had to. I think this is absolutely a moment for critical thinking and and pausing to consider what is, you know, where do we actually want to get to?

00:12:04:14 - 00:12:06:03
Unknown
that's a very good point about

00:12:06:03 - 00:12:18:05
Unknown
the boards giving the C-suite some space because everything you said we have seen. Yeah, many, many times the speed of which we're being asked to respond and to deliver

00:12:18:09 - 00:12:28:22
Unknown
for clients is it's days, not weeks, and it's huge amounts of output. You know, that that they're looking for from their own teams as well as

00:12:28:22 - 00:12:31:11
Unknown
from external teams.

00:12:31:13 - 00:12:31:21
Unknown
So,

00:12:31:21 - 00:12:33:22
Unknown
convincing them to take a step back

00:12:33:22 - 00:12:49:12
Unknown
do this a bit slow, more slowly is a very, very difficult conversation to have when there have pressures from shareholders, from the board. And, you know, that's going to the CEO and the CEOs passing that down to the rest of the C-suite who are passing it through the organization.

00:12:49:12 - 00:13:11:00
Unknown
Everyone, from what we've seen is running at this full tilt. And we have yet to find the, the, the line that we can give them to that will convince them to take that step back a little bit.

00:13:11:02 - 00:13:32:07
Unknown
whenever I've said that to people, it's actually resonated a little bit that in other parts of your life, if someone's pressuring you for pressuring you for action, that's normally a warning sign. Not not something to be encouraged. So I think that's a helpful reframe. I think the idea about having the video breaks and things that you can go helps as well, because it's actually again, thinking about do we, you know, we can actually overtake people in the next curve.

00:13:32:13 - 00:13:35:16
Unknown
It's not about winning in this portion of the race.

00:13:35:18 - 00:13:52:01
Unknown
What we want to do is win the race overall. And actually it's, you know, relatively unusual. And I mean, it's just a basic metaphor of its relative unusual that that the person first out of the blocks is necessarily the winner in any kind of distance race. Anyway. And then, so I think those two are helpful.

00:13:52:01 - 00:14:15:18
Unknown
I mean, many in the AI ethics community have been saying for some time nothing's, you know, people aren't going to slow down until there's disaster. But sometimes, unfortunately, you need to have a catastrophe in order for people to, to stop. You know, I really do live in hope that we can we can avoid that. This time. And I notice in the commentary, in the kind of in, in the market commentary that there is more general awareness that people aren't adopting this.

00:14:15:18 - 00:14:41:20
Unknown
In fact, a lot of the people numbers are going the wrong way, that there's a public backlash against some of this. So that's causing a little bit of pause for thought. I think that the changes in token, pricing and token usage and some of the stuff we've seen coming out is also a cause for thought. And I do wonder whether, you know, we've got two, three very big IPOs this in the next 12 months with, space X, about to list, and then OpenAI and anthropic due to follow.

00:14:41:22 - 00:14:57:19
Unknown
And the process of going through that IPO will put real attention on exactly what do those economic models look like. And then when you apply that to the rest of the economy, you know, again, that that provides perhaps more insight about where we might be going over the next 3 to 5 years. And actually, is that does that make sense?

00:14:57:19 - 00:14:58:17
Unknown
And is that what we want and

00:14:58:21 - 00:15:06:22
Unknown
more we can have these conversations collectively, the better, I think, because if people can see other people slowing down, it helps give the give the permission a little bit to do that.

00:15:06:24 - 00:15:13:06
Unknown
So I think the more of us who think that, the more the more, the more vocal we can be about it. I think maybe it helps give some cover.

00:15:13:08 - 00:15:16:22
Unknown
a part of the reason for the pressure is that technology is moving so fast?

00:15:16:24 - 00:15:38:13
Unknown
But in practice, organizations and people cannot move at the pace the technology is moving. And in trying to do that, you know, the gears are cranking. The system isn't work. We're not designed to do that. So the first question is, if you actually want to move at that speed, you probably need to think about organizational redesign because the organizational structure we have can't deliver that.

00:15:38:15 - 00:15:41:24
Unknown
one of the greatest opportunities that I think I brings

00:15:42:01 - 00:15:47:00
Unknown
So the ability to actually take a blank sheet of paper and go, wow, what is the thing we're trying to achieve?

00:15:47:06 - 00:15:53:10
Unknown
And then what actually is the most effective way of doing that? That's also best for all the humans in the system. And then let's design for that. I

00:15:53:10 - 00:16:04:07
Unknown
may be eight uses cases where you can. But the idea that we should be pegging our speed to change, I think, is the wrong frame. Again, we need to be actually it's it's it's what you're doing. And if what you're doing can be 100% automated, then great.

00:16:04:07 - 00:16:07:13
Unknown
Go for your life. Although even then you still have to go the organizational pace.

00:16:07:13 - 00:16:20:14
Unknown
But if you're actually building AI human together, you know you can't go faster than the human can go. And we shouldn't be seeing that as a use as a failing in the in the human that, you know, we all have we are

00:16:20:16 - 00:16:24:03
Unknown
we don't exist to serve the tech, the tech supposed to be there to serve us.

00:16:24:05 - 00:16:30:06
Unknown
would find that dystopian, and, I mean, that's where where I think the, you know, the Pope. I don't know if we'll get into the encyclical, but where

00:16:30:06 - 00:16:33:03
Unknown
another influential opinion coming in saying, hang on a minute.

00:16:33:05 - 00:16:37:10
Unknown
Let's just slow down here and let's think about what we're doing. And and the Pope, as I understand

00:16:37:10 - 00:16:52:18
Unknown
situates the starting point as the dignity, the the fundamental dignity of every human life on the planet. And we we that's the that's the stuff. Well, what's good for the humans in a, in a situation.

00:16:52:20 - 00:17:07:24
Unknown
And, and I think, you know, there's, there's firstly that point again about technology serving the human to human, serving the tech. I don't think, the Silicon Valley, the narrative that comes out of Silicon Valley is often that, you know, humans, humans have design flaws that need,

00:17:08:03 - 00:17:11:19
Unknown
designing around, you know, trying to remove all friction.

00:17:11:19 - 00:17:32:02
Unknown
And actually friction is how we it's often how we learn and form judgments and that there's a value to that. That's, that's wisdom. You know, Elon Musk is on record as saying the biggest, weakness of Western civilization is empathy. You know, and this is not I mean, that he's absolutely entitled his view. But I don't think that's the view of most people on the planet.

00:17:32:02 - 00:17:36:17
Unknown
And I don't think it's how most people want their employer to treat them

00:17:36:22 - 00:18:00:08
Unknown
Yeah. It's really interesting that anthropic, is aligning themselves to the Pope on the encyclical, because what the encyclical says is essentially tech should serve humanity and not the other way around. And that is the opposite of what the Silicon Valley founders have been wanting of AI. It's

00:18:00:10 - 00:18:05:03
Unknown
playing God. What if we could create the ultimate intelligence?

00:18:05:05 - 00:18:11:09
Unknown
And what the Pope is reminding us of is what for? What is AI? If not to serve humanity?

00:18:11:11 - 00:18:21:10
Unknown
We have at the moment a real rise in people wanting authentic authenticity, it authenticity, and to feel that they can trust, the people that they buy from and that kind of alignment of, of brands.

00:18:21:11 - 00:18:35:13
Unknown
even if you think about it from an economic angle, if we're replacing older people with machines, who's going to buy your products like where does it come from? It doesn't it doesn't actually make any, any sense. So, you know, they they're really super excited about engineering stuff.

00:18:35:15 - 00:18:46:09
Unknown
But but that doesn't mean that they are right about organizational design or that they're right about the impacts at a kind of social level.

00:18:46:11 - 00:18:53:09
Unknown
bias that when you're an expert in one field, you can often assume that that expertise, automatically, you know, that makes you an expert in other fields.

00:18:53:09 - 00:19:07:17
Unknown
Or when we listen to someone who's an expert in one field, we can, you know, Elon Musk or Boxer that are undoubtedly experts that develop, but that doesn't make them expert in how do you redesign an economy or had.

00:19:07:19 - 00:19:23:23
Unknown
Yeah. I mean, absolutely. And I think, look, he's brought and actually this pope isn't, isn't the first. You know, the Vatican has actually had quite a lot to say about AI already. If you go back to 2020, I think was when they first produced a really interesting, some really interesting.

00:19:24:01 - 00:19:32:06
Unknown
So pieces, yeah. They, they've been it predates this. Although of course, he might have been the one within the Vatican who was champion that I don't know exactly

00:19:32:08 - 00:19:46:13
Unknown
I just hope that with AI ethics, we don't end up in the same place as purpose and the ESG, which have become just a corporate communication thing. Tick the boxes. Do you want some awards? That's it's meet some random targets.

00:19:46:15 - 00:20:03:10
Unknown
There is just so much more we can do with AI ethics to make society and the world a better place. And I know it sounds idealistic, but I genuinely believe that AI ethics needs to be reshaping the way we do business moving forward.

00:20:03:12 - 00:20:12:03
Unknown
probably the, this is a big thing from the Pope, but there are at least two previous pieces from the Vatican about being intentional and thinking about the human as we design AI.

00:20:12:03 - 00:20:39:00
Unknown
So they think quite early saying we should think about AI ethics. And I think, you know, and again, I think although obviously, the Pope speaks to the millennials, it will have however many Catholics it is worldwide. I think the points apply even beyond, you know, whether or not you're a Catholic, you're of a Christian faith, but partly because they're grounded in some of those kind of, secular ethical principles that we that we talked about, but also because increasingly it's even been recognized as kind of good business.

00:20:39:02 - 00:20:56:24
Unknown
You know, I was talking about anthropology, anthropology with, blueprint for Better Business, which is a charity, that was set up after the, global financial crisis to say, have we lost sight of purpose? You know, have we mistaken in financial services profit as purpose? The profit is the thing you receive for delivering purposes.

00:20:57:01 - 00:20:57:15
Unknown
What do you.

00:20:57:17 - 00:21:13:17
Unknown
You know, what's the thing? What's up? I'm gonna trying to solve what's the goods you're trying to provide? But you know, what are you giving to customers? That's your purpose. And then as a result of delivering that purpose, you generate a profit. And if you're focused on only the profit, you can lose sight of your purpose and you can go astray, which is partly what arguably happened in the financial crisis.

00:21:13:19 - 00:21:29:18
Unknown
And that route, to profit actually. Also, you know, if you kind of think about purpose and people and profit that you end up in perhaps a more sustainable and sustainable place, way too caught up in some of these, you know, unhelpful bubbles and unhealthy bubbles.

00:21:29:20 - 00:21:50:20
Unknown
I think there's huge potential and there's huge potential to, to get rid of loads of bureaucracy and, and rubbish in the system. So I mean, you started off Ben, by saying innovation versus, you know, we need to innovate and then how do we think about legislation. And it's my view that that's a false dichotomy. You actually I think, innovation is supported by good regulation and is hampered by bad regulation.

00:21:50:20 - 00:22:06:14
Unknown
But no regulation at all. Also doesn't help, you know, would we all get on an airplane if we didn't have, you know, degree of confidence? Do you buy, you know, do you buy food if there aren't food standards? You know, Stuart Russell famously says, make a sandwich has is better regulated than the I. And it has, you know, much less power.

00:22:06:19 - 00:22:19:12
Unknown
So I think it's about having effective regulation. But I think the way we've in which we have, embodied, you know, regulatory environment within our organizations has become enormously cumbersome. And quite often it's not what the regulators want. I mean, the

00:22:19:12 - 00:22:24:22
Unknown
I speak specifically about financial services because that's more my, my background, has become less about

00:22:24:24 - 00:22:32:17
Unknown
how do we make sure we're compliant with the regulation and how do we manage risk effectively and much more about how do we minimize all legal risk to the entity,

00:22:32:19 - 00:22:42:05
Unknown
so that we've developed a very legalistic understanding about how you interpret regulation, which has sort of, suffocated companies, I think.

00:22:42:05 - 00:22:48:14
Unknown
doesn't mean regulation itself is inherently bad or even that that's necessarily what the regulators want. It means I think we need a system rethink

00:22:48:14 - 00:22:49:20
Unknown
other thing I quite like this idea.

00:22:50:01 - 00:23:05:06
Unknown
Some people have, every time you want to add a new regulation, you should take one away. I think it's quite nice. It's a good. But instead of what we focus on is a growing body that's being applied in a super kind of bureaucratic way. And those those don't need to be the case. And one of the eyes,

00:23:05:06 - 00:23:07:11
Unknown
if we think about this as a redesign problem,

00:23:07:11 - 00:23:14:07
Unknown
how could we redesign this so that we are managing risk effectively, honoring our governance requirements, but not getting caught in this,

00:23:14:07 - 00:23:15:23
Unknown
straitjacket of, of admin.

00:23:16:03 - 00:23:21:03
Unknown
that then becomes kind of an exciting design problem. And you can make governance something that is a competitive advantage

00:23:21:05 - 00:23:37:13
Unknown
it's been, implemented almost as a zero risk, zero exposure. Version of the regulation, isn't it? Which is, which is you can't leave the house with zero risk. So we, you know, we have to we have to accept some.

00:23:37:15 - 00:23:38:11
Unknown
Yeah, exactly.

00:23:38:12 - 00:23:39:08
Unknown
That's exactly.

00:23:39:11 - 00:23:40:07
Unknown
Well, so actually,

00:23:40:12 - 00:23:52:17
Unknown
zero risk isn't zero risk. Because I also think to your point here, I am saying, yes, we should all slow down and be thoughtful, but also the idea of saying, well, it's too risky. So therefore we will do anything that that is not the zero risk position. Actually,

00:23:52:17 - 00:23:54:21
Unknown
Absolutely.

00:23:54:23 - 00:23:55:24
Unknown
and somebody go, right.

00:23:55:24 - 00:24:09:14
Unknown
So that's what we've decided. AI is too risky. No, I and then under the table where everybody was sitting on their ChatGPT and, what have you, you know, that is also, you know, that's what you happen if you try and push all the risk out of a situation. And actually, it will it will find its way into, into the layers.

00:24:09:14 - 00:24:13:09
Unknown
But you can't even manage it or know what's going on. So

00:24:13:11 - 00:24:23:20
Unknown
Yes. It's really interesting that you're connecting the regulation to the legal aspect, because then regulation is about stopping risks.

00:24:23:22 - 00:24:47:19
Unknown
And so that's why it feels like is stifling innovation, when actually regulation should be a way to make sure that the tech is serving humanity. And through that lens, it's no longer about how many billion dollars am I going to pay. And fine. And it becomes about how do we make sure that we're building a better society whilst the tech is thriving?

00:24:47:21 - 00:25:09:17
Unknown
100% that I 100% agree with you. I think it's partly about, I mean, when I talk about, governance quite often, I talk about it's about creating environments you can play safely or you can explore safely. So it's partly that it's actually a, you know, if it's done well, it's about enabling innovation. And second thing is it's about thinking about that public interest.

00:25:09:19 - 00:25:17:11
Unknown
And part of the reason I think that we're seeing trust, you know, it's not I mean, trust is low in AI. So particularly in the UK and the US,

00:25:17:17 - 00:25:23:04
Unknown
the US has the lowest trust, according to not only the Edelman, barometer.

00:25:23:04 - 00:25:26:00
Unknown
They did a flash poll in November last year

00:25:26:02 - 00:25:33:14
Unknown
and also, Ipsos did a poll for the, Human-Centered, AI index that comes out of Stanford once a year.

00:25:33:16 - 00:25:45:21
Unknown
Both of those show us as having the lowest trust in in AI of any country in the world. I think these two things are related UK not far behind, by the way. But also trust is going south.

00:25:45:21 - 00:25:58:07
Unknown
It's, it's and that's not something you can just go we can sit it out. Firstly there's the public backlash. You know, we saw the commencement speeches this summer, two different universities booing speakers who talked about AI.

00:25:58:08 - 00:26:19:09
Unknown
We we're starting to see public backlash against data centers, for their water use, which, again, is actually a, you know, potentially a physical constraint on AI, development. We've seen, the, people who think AI is a bad thing go from a third to a half in the last three years. Again, that's a that's a sizable move.

00:26:19:11 - 00:26:39:24
Unknown
So again, get this idea that being, government regulation free is going to speed up innovation. Not if by doing so, you're actually leaving the public behind you. Part of what regulation is supposed to do is be able to communicate that public interest. And if the public don't feel what their interests are being expressed, you know, they have the right to resist it or even in some cases, acknowledge it.

00:26:40:00 - 00:26:44:00
Unknown
And I saw a piece yesterday about how there's quite sabotage in some places.

00:26:44:01 - 00:26:53:22
Unknown
the statistic that I found really chilling was that for adults below the age of 30, so that the the age group 19 to 29, the percentage you feel hopeful about AI is 14%.

00:26:53:24 - 00:27:02:16
Unknown
80 whatever that is 86% of adults below the age of 30, you know, don't have hope. And given that all anyone is talking about is an AI future,

00:27:02:20 - 00:27:14:24
Unknown
but when you start to think through the second and third order consequences of us going through this AI transformation, it raises some of these really big, profound, philosophical questions about what it is to be human and what it is, what sort of what it is to be human together,

00:27:15:01 - 00:27:40:22
Unknown
and we wonder why adoption of AI services in an enterprise level and things like that is not as high as we would expect it to that there is, there are all these pilots that going out that don't succeed. And when you apply it through that lens of there's so little trust, it's no it's no wonder

00:27:40:24 - 00:27:41:22
Unknown
What's in it for them?

00:27:42:01 - 00:27:51:16
Unknown
fundamental level. And that, again, is the thing I think the tech browser forgotten about is a moment we can see the benefit for them, but it's not at all clear that that benefit is going to be to be shared.

00:27:51:16 - 00:27:52:04
Unknown
And instead

00:27:52:04 - 00:27:57:02
Unknown
everyone can immediately see what the potential cost for these benefits people talk about.

00:27:57:02 - 00:28:15:05
Unknown
They seem to be accruing to other people, not, you know, maybe we get the health benefits. Maybe there's some solution to the climate crisis through AI that would be great. Collectively. But but, you know, there's a pie in the sky, whereas meanwhile, you can quite literally now say, I'm not gonna be able to ask for a raise because I'm just gonna say, you're lucky to have a job, and I may not have a job to, you know, two years time and

00:28:15:05 - 00:28:17:20
Unknown
that the downsides are very visible.

00:28:17:22 - 00:28:20:21
Unknown
And the upside is and

00:28:20:23 - 00:28:42:16
Unknown
the other thing I'd just like to say on that, Ben, is one of the one of the routes I've gone down since sort of exploring on this is, is our experience of transformation, because one way of looking at what we're doing at the moment is going through, you know, the mother of all transformation projects, that they beyond anything any of us have experience before, not not least not only because of the kind of profound level of it, but because it's happening everywhere,

00:28:42:16 - 00:28:46:16
Unknown
across every sector, across the whole of the economy and public services and everywhere else beyond.

00:28:46:16 - 00:28:53:12
Unknown
And I don't know if you know what our track record is like for transformation. But but it's it's not good. Go put

00:28:53:12 - 00:28:59:15
Unknown
it that way. If you look at how we do what our current transformation playbook delivers, it's a 70% failure rate.

00:28:59:15 - 00:29:00:21
Unknown
According to McKinsey.

00:29:00:23 - 00:29:03:20
Unknown
Well, and when you say we, who are you referring to?

00:29:03:22 - 00:29:09:17
Unknown
Gartner puts it at an 80% failure rate where technology is involved.

00:29:09:19 - 00:29:21:20
Unknown
this MIT study that came out that showed a 95% failure rate for, early AI projects. Now you can critique that. And lots of people have and I don't think the 95% number is robust necessarily. But what's interesting about

00:29:21:20 - 00:29:31:06
Unknown
the MIT study, is that the reasons why the transformations were failing are the same as the ones that McKinsey and Gartner point to.

00:29:31:08 - 00:29:32:24
Unknown
and you can boil it down to

00:29:33:03 - 00:29:34:04
Unknown
three things

00:29:34:10 - 00:29:47:01
Unknown
The first is if you are imposing a solution from above, rather than actually seeking to solve a problem, which is exactly what we've seen with AI, we are doing that left front and center. Instead of thinking, what's the problem we're trying to solve? What's the right?

00:29:47:01 - 00:29:50:19
Unknown
And is AI the right tool? How do you use it? It's just like you must use AI.

00:29:50:21 - 00:29:51:19
Unknown
The second reason

00:29:51:19 - 00:29:52:01
Unknown
is

00:29:52:03 - 00:29:58:19
Unknown
that you either haven't skilled the people or you haven't got the cultural buy in to go on this transformation together.

00:29:58:21 - 00:30:02:09
Unknown
you know, some AI training is out there, but most of the training is really generic.

00:30:02:10 - 00:30:03:20
Unknown
tailored, targeted guidance,

00:30:03:20 - 00:30:06:06
Unknown
is, I think, what you need in order to really get the benefits.

00:30:06:08 - 00:30:09:17
Unknown
And then the third reason why transformations fail is a governance failure.

00:30:09:19 - 00:30:13:03
Unknown
You don't have the mechanisms in place to play safely or explore safely

00:30:13:03 - 00:30:23:20
Unknown
one of the nice things about that MIT study is that the 5% who are reporting a success are ones who had designed for those things, who had said we made sure we had cultural by and by.

00:30:23:22 - 00:30:26:21
Unknown
And that's one of the reasons, again, why I say don't rush.

00:30:26:23 - 00:30:31:01
Unknown
This is every bit as much as a technical challenge. It's also an organizational challenge, a transformational challenge,

00:30:31:01 - 00:30:41:09
Unknown
What level of investment have you seen that's gone into technology versus transformational change and adoption? What's the percentage you reckon

00:30:41:09 - 00:30:46:23
Unknown
I would guess that if you looked at it from three different investments, the investment in the tech, I think that comes first by a country mile.

00:30:47:04 - 00:30:52:11
Unknown
Then there's the investment in the transformation that probably comes next. And then there's the investment in the human infrastructure, if you like.

00:30:52:11 - 00:30:53:06
Unknown
Yeah.

00:30:53:08 - 00:30:58:05
Unknown
whether through a mixture of training and cultural participation work and all the rest of it,

00:30:58:05 - 00:31:06:11
Unknown
that's just anecdotal, but I'm regularly hearing about people saying, oh, yeah, we've we've had some sort of AI webinar, but all of our other training budgets have been canceled.

00:31:06:14 - 00:31:19:07
Unknown
if we're thinking about this, is an integration between a human infrastructure and an artificial intelligence infrastructure. Almost all the investment has gone into the AI infrastructure and not into that human infrastructure that needs to whether whether you need to work together.

00:31:19:09 - 00:31:23:04
Unknown
And and part of that is being do we spend money on employees if we don't know if we're keeping them?

00:31:23:06 - 00:31:35:23
Unknown
But of course, employees can read that as well. So if you're not investing in me, do I have confidence in the future? Am I going to how quickly am I going to help to train these tools to possibly replace me? You know, it becomes a it becomes a vicious, vicious circle.

00:31:36:03 - 00:31:55:15
Unknown
So one key takeaway for leaders is whatever you are investing in, in AI implementation infrastructure, rethink about how much you need to invest in that adoption program training education across your business. Because it is it's probably a lot more than what you've got in your line item at the moment.

00:31:55:17 - 00:31:56:20
Unknown
I think part of the reason

00:31:56:22 - 00:31:59:18
Unknown
there's a trust gap here is because we're not being frank about it.

00:31:59:20 - 00:32:04:18
Unknown
what I suggest is that firms are actually honest with staff and say, we don't know.

00:32:04:20 - 00:32:19:17
Unknown
Now, the cost that that will give you in terms of maybe training some people that you might then ultimately lose, I think will be easily covered by the faster adoption that you will get and the and the and the better adoption that you'll get from the fact that you've got Buy-In of your of your from your people

00:32:19:20 - 00:32:22:16
Unknown
There's also a skills gap emerging in other parts of the business.

00:32:22:16 - 00:32:36:05
Unknown
We could be proactively reskilling, the workforce. And even if you end up reskilling and somebody go somewhere else, still at a kind of overall economic level, it will be better because the more people there are employed, the more people using your product, the healthier the economy is.

00:32:36:05 - 00:32:51:23
Unknown
So, so even if it means spending some money you didn't need to spend on training people, I still think that will be, that will be more than covered by what you will save on not having that kind of transformation, sabotage, as well as being actually straight with people. That's a trust building exercise.

00:32:52:02 - 00:32:56:24
Unknown
I've heard you speak as well about the five CS for leadership,

00:32:57:01 - 00:33:06:01
Unknown
it feels like good tools that leaders can use to help get to the point where they can implement what you've just spoken about.

00:33:06:05 - 00:33:20:11
Unknown
Thank you. Ben. Yeah. It's a it's a conversation that's been coming up over the last couple of years, really, when talking about this topic is, you know what? How do we lead through this? What are the kind of human characteristics?

00:33:20:13 - 00:33:24:24
Unknown
You know, I wouldn't suggest necessarily unless unless you're a tech company, that

00:33:24:24 - 00:33:28:10
Unknown
your leaders need to be tech specialist or experts

00:33:28:12 - 00:33:39:16
Unknown
think part of leadership is actually having a load of subject matter experts who you can call on. You need enough knowledge to be able to challenge them on that. But don't go down the route. I would say of thinking that you have to, you have to

00:33:39:16 - 00:33:40:02
Unknown
know.

00:33:40:02 - 00:33:59:11
Unknown
it's understandable because for most of us, the route to seniority has come through. Being an expert in something, it's come through knowledge, and we have tended to associate, seniority with knowledge. But in an environment in which information is freely and infinitely available, which is the one we're now living in, the real differentiator becomes how good the questions are that you ask.

00:33:59:17 - 00:34:07:18
Unknown
So cultivating a sense of curiosity, I think, is probably the most valuable thing you could do. Curiosity with

00:34:07:22 - 00:34:15:22
Unknown
So critical thinking. So, so, you know, being interested, being open, but but in a way where you're actively,

00:34:15:24 - 00:34:25:01
Unknown
challenging what you're hearing. But also the humility, because we're in an environment, again, where I think this is too big, too complicated, too fast moving for any one of us to keep on top of it.

00:34:25:01 - 00:34:41:09
Unknown
You and I spoke before about about how how we both like to say no, no one is an expert because it's not possible. So so what you what you can do to find your way through this is to build collective intelligence to to you have the bits, you know, and somebody else's, the bits they know. And so that requires a certain humility about the edges of your knowledge.

00:34:41:14 - 00:34:43:17
Unknown
So that's what I mean by curiosity. That's the see.

00:34:43:19 - 00:34:53:21
Unknown
And then the next see is courage. Because that is quite a courageous thing as a leader, to be able to say, I don't know or, you know, to recognize that humility.

00:34:53:23 - 00:34:58:23
Unknown
The other thing that's going to take courage is that whatever you think and whatever happens, we are in unprecedented time.

00:34:58:23 - 00:35:13:13
Unknown
So we're going to be quicker by necessity. We're going to be trying new things. And some of them won't work. So entering into that kind of courageous, kind of growth mindset recognition, it won't all work. But how do we take the learnings and move on? Rather than that zero risk

00:35:13:13 - 00:35:16:14
Unknown
that C, talk about is craft.

00:35:16:16 - 00:35:23:01
Unknown
And by craft, what I mean is both kind of creativity elements, but also that sense of mastery.

00:35:23:01 - 00:35:31:10
Unknown
for example, I, I would never use, ChatGPT or Claude to produce the final version. In

00:35:31:10 - 00:35:39:00
Unknown
I might use it to do some help me do some research to figure out where I should go, and then I'll try and do some of that thinking myself, and then I'll it will help me draft.

00:35:39:04 - 00:35:56:05
Unknown
But then the polish and the final refinements are something that comes from that, that wisdom and that hopefully. But anyway, the, the bits that you've learned, the mastery of your skills over it over time, it's kind of like something that's sort of artisanal, maybe. So craft is my is my third C,

00:35:56:05 - 00:36:01:06
Unknown
my fourth is compassion. Compassion for for three reasons.

00:36:01:06 - 00:36:10:19
Unknown
Compassion partly because my God, at a human level, this is hard. What we're doing, this is a lot of transformation, a lot of unknowns. We find that difficult. So so let's recognize that

00:36:10:19 - 00:36:19:11
Unknown
point two we are humans. And and the point for most of us about about life is is the other humans in it. And and our human experience.

00:36:19:13 - 00:36:21:21
Unknown
Yeah.

00:36:21:23 - 00:36:36:03
Unknown
thing is that compassion is something that is uniquely human. You know, I can, can, can produce a, a reflection of it. But that's not real compassion. Real compassion is something that exists between between human beings. And we can

00:36:36:05 - 00:36:52:12
Unknown
and then my final say is community. Community, partly because the studies show the longest running study on happiness out of Harvard, suggests that, the quality of our relationships is the thing that is the, the most accurate predictor of our of our health and happiness.

00:36:52:14 - 00:37:07:05
Unknown
The quality of relationships is, is crucially important. Community is also how we can hope to have this sort of collective intelligence and collaboration that I think we're going to need in order to, to cope with some of these kind of systemic challenges and community, because it's the thing that makes those other CS possible.

00:37:07:10 - 00:37:09:24
Unknown
Community is where you can exercise your curiosity.

00:37:10:02 - 00:37:28:22
Unknown
It's where you can build your courage, is where you can hone your craft, and it's where you can, give and receive compassion. So, so community is sort of the bit that pulls it all together. And, and I think those five CS are hopefully a helpful way of kind of reframing your thinking about maybe, maybe some different areas to, to, to lean into over the, over the coming years

00:37:29:00 - 00:37:31:13
Unknown
is not this is not just about absorbing more knowledge.

00:37:31:17 - 00:37:51:15
Unknown
You will never know enough, when I that's actually not the point. It's how do we how do we build on the knowledge of others? How do we exercise curiosity and make sure we're retaining that critical thinking and not just getting bounced into something because someone who knows more than us is telling us it's a good idea, you know, how do we make sure we we keep the whole, the whole community together because we are a social, very social species.

00:37:51:15 - 00:38:03:00
Unknown
We have, been as successful as we have been, partly because of our intelligence, but also because we've done it together. It's the it's the collective intelligence that has as won out. And I think that's where we need to to aim for the future.

00:38:03:04 - 00:38:18:05
Unknown
So I bet you find it conversational AI, which is a cross-sector conversation that needs to happen. It's definitely what we're trying as well to do with this podcast, The Impossible Brief. But the quality of conversation on AI is that you're actually bringing people together in the room.

00:38:18:10 - 00:38:23:15
Unknown
It's in person. There is no tech. Whatever is said in the room stays in the room.

00:38:23:17 - 00:38:41:20
Unknown
I mean, I think I was partly inspired and I should say I've got to, to people, to partners who founded it with me, which is, Tracy Woods and, Brian Ball, and, Brian Ball was our professor in our AI and AI philosophy masters. So it's a he brings the sort of academic heft, through it, through Northeastern University.

00:38:41:20 - 00:38:52:13
Unknown
And then Tracy and I, Tracy has a long, background in it, and then, and then consulting and, and then I background in financial services and, and mostly we wanted to keep the conversation going

00:38:52:13 - 00:39:15:18
Unknown
And that's where we start thinking how how do we kind of keep going with these kind of open, interdisciplinary, spaces? And then I went to entropy, and I think that also, that also planted a seed, because what I loved about entropy was partly that cross-sector element, making sure that you've got business and you've got government and you've got, you know, academia or whoever in the room, civil society.

00:39:15:20 - 00:39:30:10
Unknown
But also that so participatory and that is so kind of bottom up. It's, it's like, let's not just look for top down solutions that's actually see what is happening on, on the ground. What are the good ideas coming through and how do we kind of harness this and how do we get inspired by each other. And so

00:39:30:12 - 00:39:32:18
Unknown
we started in 2023.

00:39:32:20 - 00:39:36:01
Unknown
we've probably run like probably about 12 of these things.

00:39:36:03 - 00:39:39:08
Unknown
they did used to be every quarter. But I think we're on track for about eight this year. Now.

00:39:39:10 - 00:39:45:04
Unknown
we've looked at things like innovation, creativity, education, health

00:39:45:04 - 00:39:45:24
Unknown
The next one we're doing,

00:39:45:24 - 00:39:47:19
Unknown
is on investor due diligence.

00:39:47:19 - 00:40:07:02
Unknown
it's a, it's what should investors be looking at when they think about AI. It's a it's a it's a deeper, more practical workshop than we usually do. But we've been we're lucky to partner with tech UK and with, with a couple of the investment bodies and NCSA to look at what has been learned elsewhere through AI procurement, AI assurance, AI insurance that investors can learn from.

00:40:07:04 - 00:40:16:11
Unknown
And then the one after that is gonna be on AI and children. So what do young people feel about where I is going? And we're delighted that we've got a young person on the panel, speaking.

00:40:16:13 - 00:40:22:04
Unknown
we try to make sure that the conversation is led by people who bring expertise that crosses sectors, and that

00:40:22:05 - 00:40:22:22
Unknown
disciplines.

00:40:22:23 - 00:40:26:22
Unknown
There's usually a tech specialist and an ethical specialist and a subject matter expert

00:40:27:01 - 00:40:28:16
Unknown
But what really makes them different,

00:40:28:16 - 00:40:42:09
Unknown
is the pot of participatory element. So once we've heard from the panel, we break out into into groups and everyone discusses it amongst themselves, and then we all come back together for what ends up being a such a lively, rich discussion.

00:40:42:11 - 00:40:44:12
Unknown
it's quite hard to bring it to a close.

00:40:44:12 - 00:40:48:22
Unknown
all the speakers end up coming away saying that these visuals, conversation we should be having.

00:40:48:22 - 00:40:53:22
Unknown
And then what I what I really got out of it is the enthusiasm to be in the conversation.

00:40:53:24 - 00:40:55:11
Unknown
And it's one of the reasons why,

00:40:55:11 - 00:41:18:20
Unknown
I, I would highly recommend to company leaderships not to think this is something you've got to solve for your staff, but to involve your staff in solving it together because that there is such an appetite to to have this conversation. People can feel this enormous thing is happening, and not having any agency or voice in this huge thing that's happening is part of what makes it so destabilizing, and part of what feeds that distrust.

00:41:18:22 - 00:41:29:09
Unknown
Again, thinking about the ideas of humility and curiosity and collective intelligence actually tapping into what people are already thinking, you know, surfaces so much, that we haven't seen.

00:41:29:10 - 00:42:00:12
Unknown
So, yeah, I mean, we love doing them. And we've been really and we've been really struck by the, by the appetite for them. And I just think, and hopefully we would like to lend some of this to, to, to movements towards a also kind of national conversation on AI, which is another thing think I think we should probably be having I really think not only the more minds in that sense of like, you know, the wisdom of crowds, but also the sense of actually everybody being able to feel that they have some agency and some ownership over what's happening, what's happening to them and what's happening to us in this, in this extraordinary

00:42:00:12 - 00:42:01:03
Unknown
moment in time.

00:42:01:05 - 00:42:31:13
Unknown
So, yeah, there's a survey that we keep referring to is then, because it really shows the dichotomy of what we're dealing with. And what's executives, as well as employees on the ground are dealing with. So writer and the AI consultancy, did this survey on AI adoption for the second year running. And what came out of it, like the key information that I keep in mind because it's so telling, are that 29% of employee, sabotaging actively the AI, the implementation of AI in their company.

00:42:31:15 - 00:42:51:18
Unknown
And on the other hand, 75% of executive don't really think their AI strategy they think is for sure. So for PR, for investor relations, they're not really for internal guidance. And then 61% of CEO fear that they're going to lose their job if they can't leave that company through that AI transformation. So you have those two groups of people not talking to each

00:42:51:18 - 00:42:52:13
Unknown
Yeah, exactly.

00:42:52:18 - 00:42:54:13
Unknown
also not really knowing what they're doing.

00:42:54:13 - 00:43:08:24
Unknown
And it's very clear that nobody is an expert. But it takes courage and honesty and humility to actually lead through that. So it's everything that we've

00:43:09:01 - 00:43:12:07
Unknown
That's exactly.

00:43:12:09 - 00:43:34:12
Unknown
expert? How do you admit, you know, how do you say that to your employees? And I feel that the five seed, that you have seemed to be such an easy answer to what needs to happen, you know, and I just want to put all those people, put them in a room, tell them, please be honest and humble about the difficulties, and let's all have a tough conversation together.

00:43:34:14 - 00:44:07:21
Unknown
And of course, that's easy to say, but it's quite it's much more difficult when there is no psychological safety, because are you going to be laid off in the next wave because you've spoken up? You know, it's really difficult with the current climate to have this conversation. And I believe that and although this is a famous company, I do believe that the people who choose to have those difficult conversation leaders who choose to be open and to build trust with their employees, will see much better outcome.

00:44:07:23 - 00:44:25:00
Unknown
To choose courage. Exactly. And choose to do these things. We'll do will do better. Because if you build that trust, even if it involves taking some risk upfront because you're going to have some uncomfortable conversations, and maybe it's going to take a little bit longer, but you're more likely to get the better adoption, and you're more likely to build trust.

00:44:25:06 - 00:44:47:12
Unknown
And if there's psychological safety, there is a very clear link between that and innovation. You don't get innovation if it's all to say I think we should do things differently. You don't get innovation without that. And at the moment we have the opposite of psychological safety. So you will get, I think, economic benefits of adoption, trust building, reskilling, rather than constantly having to hunt to fire and rehire and innovation.

00:44:47:13 - 00:45:05:22
Unknown
the economic argument that is there for doing it, it just might take 12, 18, 12 months, 18 months, two years instead of what you're hoping will be 3 or 6 months. But guess what? You're going to get three, six months anyway. Because one of the things we learned from that transformation playbook failure, is that telling people what to do doesn't result in them doing it.

00:45:05:24 - 00:45:10:20
Unknown
It's just, what they can also see there. It's not something if you know.

00:45:10:22 - 00:45:16:13
Unknown
yeah, it's a bit like parenting. You can tell your kids to do something that they will do as you do, not as you say.

00:45:16:15 - 00:45:31:11
Unknown
Yeah, exactly. Exactly. So role model it. Role model it. I don't know either, you know, but but look, I guess, you know, I don't know either, but our best way of getting through this is to do it together to harness our understanding of who we are, what we're trying to achieve. We know our customers. We know our business.

00:45:31:17 - 00:45:43:23
Unknown
What is the thing that we can do? What are the pain points? That this is a huge opportunity for us to get rid of? And if we can do this together successfully, we may not need to lose anyone. So but you know, we don't know. We'll go on that journey together. You know, it's a different model of leadership.

00:45:43:23 - 00:45:48:07
Unknown
But, and it's scary, I get that, but I honestly think it will lead to better results.

00:45:48:09 - 00:46:11:20
Unknown
one final question for you, fear. And, but we we've talked about a lot of the challenges that we face that this is an impossible brief, and we don't ever want to just look through rose tinted glasses. We need to be realistic. But what is there that we've got to be optimistic about our future when it comes to when it comes to AI.

00:46:11:22 - 00:46:35:17
Unknown
Well, you know, despite everything that we've said, I actually I'm really quite optimistic. I'm more optimistic now than at any point in the last five years, because I think at least now we're having the conversation. And I also think thanks to social media, where where it's taken 15 years for us to see the harms that have occurred and those harms on inevitable to social media, they are the result of design choices, the result of the economic model that is at play.

00:46:35:20 - 00:46:51:12
Unknown
And some of that stuff's really hard to fix now because it's baked into the design. But we're at that point in the window with AI where, you know, multiple futures are possible and there is public awareness because of the experience of social media that actually maybe we need to be a bit more intentional about this. I think it's the other reason why you're seeing the resistance.

00:46:51:14 - 00:47:00:19
Unknown
People are saying slow down of actually, these are quite, you know, the big, big stakes. But I think if we get these decisions right, we could have a hugely exciting, future.

00:47:00:21 - 00:47:09:20
Unknown
there's this great quote from the biologist E.O. Wilson, which is, we've got Paleolithic, brains, medieval institutions and nearly godlike power through technology.

00:47:09:20 - 00:47:29:15
Unknown
And, you know, we're we're living with those growing pains, the the possibility of being able to, revamp, and rethink how our institutions work so that they're more effective. Is fantastic. The, the idea it might be possible to have a much more deliberative democracy where more people have a say in how our countries are run. That's fantastic.

00:47:29:18 - 00:47:53:12
Unknown
Our workplaces are designed from the industrial. You know, we carried a lot from the industrial, era over. So we treat people as if they're machines. When they're not. We could redesign processes that actually run with the grain of how we are as human beings, rather than against it. And we could have so much of a happier, more fulfilling, working environment in which, you know, we are more effective and and much happier with it.

00:47:53:16 - 00:48:13:13
Unknown
So the possibility of redesigning for a future that is much more effective, much more exciting and, and is not only economically prosperous, but also you know, really good for us at a, at a human level, too. I think that's, that's that's within reach. It's possible depending on the choices we make. So I think, I think there's a lot to play for.