signed up for your webinars. And we can't like I signed up for your executive trigger, but I'm not sure I can make it. Well, they get emailed to us.
I'll be in the slack group. So the recordings have been put in the slack group for each one. Okay. Now you'll be able to watch both this morning's and this evenings, I should have triggered does not require an NDA to go to the webinar. It doesn't anymore, because we're not at the level we expect it to be with the partner software company, we won't be able till next week. only talking about that in outlines, it makes it a bit easier to be honest. All right. Sounds good. And that was partly because of though I wouldn't need to talk about it. I knew I couldn't resist talking about it. So hence the need for them do. So today, the NDA. It's where it's kind of like, I'm really excited by it. I'll talk briefly about what we're doing in the exhibition. But it's actually really exciting. It's the first link between sense maker in NLP tools at scale.
A lot of user
now on the gin,
we've seen you drink a lot of gin lately.
That's because North America gets me in the evening. All right. Thank you my first sip of 1801 1801
in any time zone you choose is that now you
have to be on your own time zone. I mean, it's okay. The sun is not above the yard anymore. The sun year has actually gone below the horizon. So I think that's that's considered acceptable. Yeah. Yeah. Is we are getting longer days, I've actually booked two weeks in the Lake District at the end of April. Nice. Well, it's a it's high risk. All right. According to our beloved buffoon, your so called Prime Minister will be able to do this certification from the 11th of April. So I booked it from the 18th. With an agreement, I can postpone it, but I really need to get into the hills.
Okay, give it a few more minutes.
Hang up. So just for your research project, I'm doing this all of these calls entirely from memory without notes. notes you asked about memory, that's what I'm doing.
And you're muted a rare occurrence with you. So I'm quite pleased to have witnessed it for last. Sorry, couldn't resist.
Okay, let's get started.
How many people were on the first session on this and just checking. Yeah, if you. Okay, so normal caveat. Sorry, I reserve the right to be modelling consistent between sessions, right. entangled triggers is, to my mind, one of the three or four most important things we've created because it links it actually embodies complexity principles. So in a complex system, the most important thing is interactions, not things. So entangled fears is designed for that. It's also designed to build informal network connectivity, but in such a way that it can actually link to the formal system in real time at a micro level. Now the reason for that is really important. informal networks I shall be slightly controversial here but I'm prepared to defend this mark is on the call. informal networks are not subject to the second law of thermodynamics. Be They take me there open systems, they just take energy, it doesn't cost you anything to build an informal network, because it's going to be there anyway, it's natural for humans. Whereas a formal system, yeah, it takes a huge amount of energy to create a maintain. So if you go back to the one of the original Canadian articles, complex acts of knowing, you'll find that looks at the ratio between formal and informal networks and IBM and identifies it as 160. And that's only the groups that actually manage things in virtual environments. And there was zero cost to IBM apart from the infrastructure in allowing that to happen. And it's a huge source of energy ideas, informal networks. Yeah, kind of like form around interest groups having worked together, there's all sorts of reasons they created, and they take their own energy. Now, the issue and this is one of the big focuses has never been with how to harness those and how to move stuff into the formal system on a just in time basis, not on a permanent basis. port and principle of knowledge management, yeah. If you ask people for help, they'll nearly always help you. If you ask them to codify what they know in advance of the request, they won't do it. So one of the functions of informal networks is to quickly connect with people when you need their knowledge, but not to assume that that knowledge can be codifies in advance. And that was one of the areas where Kevin started. And we develop the whole set of methods around that, including social network stimulation. And the way SMS works, it was kind of like a sort of opposite to social network analysis. As I said, Why are we doing these analytical tools and trying to change something? Why don't we stimulate its formation, so we don't need to analyse it seemed to me fairly self evident. But Rob Krause never agreed with me on that. Alright. So the way SNS worked is very simple, is you have rules for Team formation. And the rules effectively prevent people from working with people they already know, because you've got those relationships in the bag. And they forced diversity into the system. So I've used rules, for example, like somebody in your team must have joined the company within the last six months. That's actually interesting, because it assimilates new people much faster. And they have a naive perspective. So they ask difficult questions. I wish I'd thought about that before I devised the rule. as me being retrospectively coherent, yeah, I've done things like somebody in your team has to do have a degree in philosophy or anthropology. But I've checked the HR database beforehand to check there's enough of those around. Yeah. So you create rules for Team formation, then you let people's cell phone teams, provided the teams comply with those rules. So this is if you wanted this is what in complexity is called an enabling constraint. You don't allow people just to work with whoever they want. You want create, you create heuristics, which create diversity. And the way we get people to form teams, by the ways we use daikon agency software. So you put what are the heuristics you satisfy, and then you search for people, you take them out for a date, if you want to work with them, then they're on your team. So it's a you know, teams that sell form, have more power than teams, which are allocated by personnel. But the constraints ensure diversity. And if you don't fancy that speed, dating works quite well as well, right.
So that's kind of like team formation, we then have a series of rewards or things that you can earn. And we've used things like three months sabbatical for every winning member of the team. Now that can be quite motivating them to trivial things like in one of the big farmers in Basel, we have the right to have your journals in paper form, not electronic. And the fact that every single chemist competed for that heavily kind of like told him something, it's people wanted their physical journals, so they could write on them and put markers in and stuff like that. So it the rewards range. And then you have an intractable problem, something that you think you would love a solution to, but you don't know what the solution is. And whether the solution is being provided is explicitly measurable. It's not down to judgement. So what you do is you then announced a programme which says we got these problems each of them is linked with these rewards. If you can sell form a team based on these rules, and you achieve this goal, you get this reward. That's the first time we ever did it was with a big bank in Australia. So the goal was $20 million of added value benefit to the client. This was working with IBM teams. The prize was six months sabbatical which is in a consultancy group of your prize to kill your best friend for And basically people competed in rules for Team formation, you know, two thirds x Westpac, 1/3, X, IBM, this was an outsourcing deal far enough in the past, I can name the client. And what we got is three, I think three teams achieve the results. Yeah. And about five teams got close, but didn't make it. So if you think about it, $20 million of added value benefit team size seven people. Yeah. So you've got three and a half man years. That's actually a bloody good deal. Yeah, so in economic terms, it's actually an extremely successful method at solving problems. But its real purpose is to get everybody within three degrees of separation of everybody else. Yeah, we've actually run it actually, I made this comes back to our conversation. This is also a way of finding startups. If you can't put together a diverse team and solve a problem within a defined time frame, where what why should I invest in you? It's kind of I'm
a co founder, why should I invest? Yeah.
So you've got options there, right. So. So that was SNS, right? And that was valuable. And then we started to use trios for problem solving. So the first time I use this was with agile as a pre Scrum technique. So put together a bright young coder with an experienced systems architect. So the young and bright and naive with old and tired and cynical, which is always a good combination. Yeah. And put them in a trio with a user trained to talk to it, people. And that was where I've actually I'm going to put this in the wiki shortly, because I've taken a book which says, so your child has got Asperger's Syndrome, how do you communicate with them? And that's what I'm repurposing. Yeah, because actually, well come on, it's very NASCAR. Most people in it are at that end of the spectrum, don't introduce new data, do things in a structured way, it's a lot easier to train people than the other way around. Alright, so we will throw 15 triggers at a problem and see what they came up with, before you moved into the development process. And then in South Wales, we put we said to young kids, if you can, if you want to be trained in the software, we did a hackathon. This was a narrative based project, then you have to bring along somebody from your grandparents generation to work with you as a pair programming team. And if you come up with good ideas, we'll put you into a trio with somebody from the government who can make them happen. So you get the sort of principle. Yeah, by putting people together from diverse backgrounds and throwing them to the problem in parallel, we found completely novel solutions. We've also started the experiments in this in Grampian region in Scotland recently, where we put teacher with pupil with parents, that's a COVID support mechanism. And of course, a teacher is connected with old with lots of parents. So each node in that network can connect with other nodes in the distributed system. And what we're currently doing is creating a version of sense maker for those trios to keep diaries. And of course, if three people are there, there's a social obligation to complete the diary. Yeah, and yeah, the way I often explain this is if you go out to dinner with a single stranger, there's a huge stress on you. Because you've got to carry the whole burden of the conversation. If you go out for dinner with two strangers, there's no stress. Yeah, there's less burden within the system. Yeah. And I won't tell you the story about the African or the Welshman in the Englishman over dinner in Stellenbosch. And somebody, as somebody asked me later, right, and I'll tell you now, if you're English, you may get offended by it. So you may or may not want to hear it. That means Gary is bound to ask me for it. But we'll see. So either way, the concept of trios was important. So what we then start to do, so why don't we entangle these? So whereas social network stimulation, yeah, in fact, we LinkedIn connects people. So it builds an informal network, the informal network is still independent to the formal system. So what we started to do, and there's a big trial on this going on in Florida at the moment, is to identify roles within the field. So for example, in social work, you might have social worker, health worker, community policeman, you might have teacher, parent child, you might have teacher, community policeman, social worker, so you can think about it we've, we've mapped out about 30 or 40 roles in the Florida community. And what we then said is that these are pairs so we start with pairs. So we know that these two roles will see value in working together. So we make that really easy, and we recruit them into the programme. The incentive, by the way is there is money available for interventions you come up with as a trio Without going to pay you, but if you come up with a good idea, there's money available to take your idea forwards. But you've got to keep your diary up today. And then we have optional trios. So all of the other roles you can choose one of the other roles to be on your trio provided is not one of you, and that that variation is deliberate, because it allows variety to evolve within the network. And then there are conditional trios like the one we did in South Wales. So if your pair comes up with a good idea, we'll put you in a trio with somebody in government who can make your idea work. So the rows are critical, because the roles are part of the formal system. So everybody's keeping diaries and answering questions. So you've got a shared narrative database across all of the participants. So this is deliberate entanglement.
So rather than let the Bramble bushes grow in the thicket, we're actually deciding roughly what the constraints are on how the Bramble bushes should grow. So we this is an ecosystem designed in that sense. And what that means is if the network picks up early signs of mental stress, then there are roles that will see it and can intervene early, so it doesn't have to go into the formal system. So it's basically if you look at it, informal networks, spot stress better than formal systems. Yeah, or this is a weak signal detection device. Right. So that basically is in tangled trails. Yeah, and I've now got it down to 60 minutes. So I'm getting better by the day on this, right. It's written up in a blog post, it's linked with a key metaphor, which is that a micro cansia founder, so can like the formal system of the trees, and then the fungal roots connect the tree roots. But they're underground, in invisible, but they're essential to the health of the ecosystem. And they extend interest in the 500 times as far as the tree roots. Yeah, and they're far more resilient, and they provide nutrients to the tree in return for sacrifice. It's a symbiotic type relationship. So I'm also trying to induce symbiosis between formal and informal. And on a big call with the NHS about this later on this week, because we see this as a critical way in which we can reduce mental health pressure on the formal system. Because the mental health plague, which is about to hit us is major. So if we can increase network density, and one things we're doing with a diary, as you can see, I need to see more serious stories like mine. The fact you can hear other stories of people in similar or worse positions is actually hugely useful. Because it reduces that sense of isolation. I'm responsible for this, I shouldn't be feeling like this. I'm gonna minute other people feel about it. I wrote a blog post last year, which I didn't publish it on social media got picked out, which is I've almost had enough. Alright, is, and I would probably wrote it because I had almost had enough. But I thought that, you know, people like me with that. So the network needed to declare we were at that point as well. So other people could see it wasn't just them. Alright, so this is a sort of representation of that type of capability. Yeah. So entangled trios can work at a citizen level, it's a key one at a company level, it reduces the dependency of the formal system, you can radically reduce the levels of management. And you can reduce the investment in the formal system, because the informal system is doing a lot of the work for you. And of course, the informal system will help you find the people who should go into formal management physicians anyway. Because they're already validated by the way they interact with the system. So at a high level, that syntactical trees
so Dave, I've just put a an assumption into the chat, which is that if I've heard that correctly, that entangled trios are a subsystem of the human sensor networks,
is that they're a type of human sensor network. But yeah, they're not as and the point is, because the nodes are keeping diaries. They're actually a part of your sensor network, because you can ask them questions with the same software. So yeah, they does that as well.
So that's their main aim to be part of the human sensor network. Is that is that
what No, I mean, the thing I think everybody's got to understand about complexity interventions is you normally have about five or six targets benefits. Yeah, yeah, it's not it's not single intervention, single problem. It's single intervention, multiple problems. That's that's the great power of this. So yeah, it creates a network but it also actually creates a peer to peer information flow and it creates More resilience in the system independently of the sensor network. So for example, I've been designing one of these and COVID stop this, but I'll go back to it shortly, I was working with West Virginia on the COVID problem. So we were looking to create multiple entanglements between families, please social workers, educational institutes, to create an early detection that work on on opioid abuse. And which had correction capability within the network without the formal system having to get involved. And the long meetings in West Virginia with the local police, and they were really keen on this. Because what they want is community police officers to intervene early before it becomes a crime. Yeah, it is. And that's actually quite important is that ability to actually move in so that there's lots of uses for this as a technique. Your mute polar,
is there an approach to analysing the diaries? Or are we expecting that be entangled trio experts in that trio will see the weak weak signals?
I think there's two things, because it sounds vague, because it's permission based because it's got high obstruction metadata, then you will be able to look at patterns in the overall narrative from all the nodes. So one of things we're talking about with the NHS at the moment is nobody will fill out wellness surveys. But they will do lessons learned in a crisis. So lessons learned in trios is more valuable. Yeah, medical students, senior surgeon, nurse, for example. Yeah. But what we can do build into the signifies is early indications of mental breakdown. So we'll get indications not of individuals, but of areas where we're getting into sort of post traumatic stress and things like that, we'll get that in real time. But also within the system, people will be able to see those weak signals themselves and actually move to help people. And you've also got the more stories like mine stories, which disrupt mine type analysis. Yeah, in the system. So we got both. So if I can follow up the
what are the interactions between the three? So let's say you had 40 rolls, you pick the three years? And they are formed? How do they interact between themselves versus putting their diaries into the sense maker database? So they need to focus? Alright.
So the focus we're currently using, but it's not the only one is lessons learned. So what are we learning during the crisis? And the idea is they do that collaboratively and enter data together as well as keeping diaries? Yeah. So it's that mutual dependency. What we did on the teacher pupil parent is homeschooling. Yeah, what have I learned? What are the problems? Where am I going? We then find that was an interesting subset of that in the children of health workers. were the best people that capturing stories from health workers at the end of shift. Because people will come back, and they needed to tell the story of what they lived through to their children, but they didn't want to disturb them. But if the child was a journalist, all of that worked really well. So kind of like it had a cathartic sort of process debriefing. Yeah. So part of the method has got to be to find a purpose that people find useful. Yeah. And yeah, this is one of the key principles on human sensor networks anyway, is people need to see utility in contributing not because you want it for something else. But because it's something that they can do locally. But it also appeals to human curiosity, the ability to look at more stories like mine, human beings are interested in that. So they need the three of them, the three are meats together and talks about the purpose. And then they make an entry into the database here, or they review the entries they put in that day and talk about where they're going to focus to get some rest of the day, you created some generalised instructions. And again, we need that in the wiki, and then you just allow diversity in the system and see what comes in. And of course, if things start to work, then you auto suggest it to the other trails. Did you know these trails did this and look what they got, oh, we might want to copy that. So you create that image as if I mean, this is a this is what we call messy coherence. And the thing grows, and the connections grows, and some things work. And sometimes they break and sometimes they don't. Yeah. So I guess,
on that basis, and are you envisaging the connective tissue between the trio's being VSS maker or kind of Yeah.
So it could be through zoom conversation. I mean, sensemaking makes it a distributable asset that you could get It is conversations. Okay. Now, if there was sensitivity, but remember, one of the big benefits of sense maker is that people, you know, we can put permission on front end. So you can basically say nobody can see my story, only people in my trio can see the story, only researchers can see the story, or anybody can see the story. But the signification is shared, which is what we're really after. Because the signification tells us the overall patterns. Yeah, the narrative is just an explain them's
one question. So this issue of diversity, you have the chance to do it intuitive or by random? Are I'm sure you know, the concept of the Gallup Strengths Finder. You really put it down to talent. And BSA? Well, when you have three identical personalities, perhaps it's boring. But when you have the right mix of talents, you have a complimentary nationbuilder. It's a wider grant. Political does it fit together?
Yeah, we're focused on roles, not talents. Okay. So so that's the primary unit, because the role is the interface with the formal system. Yeah. Now, roles almost by definition, have different talents anyway. Yeah, in terms of the way it works, it's also as opposed to things like the strengthsfinder, what we're actually doing is let letting the system sample itself within heuristics.
So that's my belief, you can do it explicit, with a tool like this, or most of the times intuition, intuition, and normal human interaction, and that's fine. Most of the time, it will
happen anyway. But then it's not an accessible network. So what it is, is wild brambles in a thicket, which can't be navigated. What we're trying to do here is to create an easier channel for what people will do naturally, by connecting them to a wider network. So that that's why this is a deliberative process. So the point is to connect people and make that connection visible in terms of the learning from it, rather than just let it happen. Okay. And it also means you can actually create linkages wouldn't happen in normal conversation. So by getting you know, put into roles who would not normally network together, you get novelty in the system.
So and then you have a common task. And that brings people together have a common purpose and focus. Okay, thank you.
Hey, if you do the scan of the field, and you find out 20 plus roles,
how do you find the three years and it's actually in the method? Alright, so first of all, you form pairs. So the decision criteria for that at the moment is these two roles will see purpose in spending time together. Now, and that's kind of like key, and then you give them options on the third. So you can say, you know, you're in this pair, you can now choose from these roles, and then you might create different rules on that. Yeah. But then they choose which role they want on the trio, and you then link and connect. So that's where we create some diversity in it. And it then becomes quite interesting who people pick. Now, Steve, I had a question. I'm not very familiar with the sense maker software. But I'm just wondering why we've chosen the slack to do our feedback rather than sense maker and whether there's whether this could have been a an experiment that encompass both of those. Because Slack, you're dealing with explicit stuff, which has to be structured and slacks better for that for a short term group. Yeah. I am looking at the moment about potentially making a COVID lessons learn free individual use version of sense maker available so that people can play with it. And then provided they share with other people have because utility, right, but not to allow them to see group results. I mean, mass results. So we're playing with ideas on that. But the key thing about sense maker is you need volume and abstraction. The slack group is working with low volumes with a very structured purpose. So it's more explicit. If
so, I'm bringing this in startups. So as you mentioned, I'm looking at this from two angles. for single would be like, correct me if I'm mistaken, but it would appear that accelerators and incubators like Y Combinator and the valley actually follow a very similar model to what you described in making sure that the startups can share the lessons learned. between them. So it would make sense to me if if looking at some of the incubators that I've been involved in, that haven't had the same results as Y Combinator, obviously, would actually benefit from something like this, if they did apply this diary model with sharing.
Yeah. We could put out put API's into those systems as well. So we can add that in. But I think one of the things I've noticed in the valley is, the social networks are actually quite biassed. So they don't have sufficient diversity in them. So it's kind of like who do I know. And if I know somebody, I'll get introduced, and then I'll get to the right cafe and have the right conversation. This is a method of breaking that. Because part of the problem that you got at the moment, and this is a big problem on VC allocation, is people tend to allocate memory to things which are similar to things which have worked before, which is actually our priority, not what VC is of stuff is about. All right, but it is what people do, because they're comfortable. Yeah. This is one of the examples of incubation we talked about earlier, which will come on to the next session, right? It's about creating a mechanism by which people who don't fit the comfortable profile of people I want to connect with would become visible.
Yeah, absolutely. The other the other aspect I'd look at this is weak signal detection. Yeah. So to me, I look at weak signal detection from a startup point of view as a story of Reddit. So for Reddit, what they were doing basically is they were scouring the database to see exactly what types of conversations were happening while they were building the network. That basically led them to the idea of the sub reddits. And the entire data structure in the way they presented it down the road as a social network. So in this case, would the weak signal protection through trios, can this be automated? Or does it have to have a human element
it can. And what I would do in that context is I would actually define a limited number of roles, and actually almost require an entrepreneur to create a network of trios, and get data from it as a proof of concept. Because if you can't create a basic network and get people involved in doing something together, well, you're not gonna be able to do this anyway. Alright, so and then we'll look at the results and what we can then do. I mean, this is one of the big things we did on counterterrorism. And this is weak signal detection, is we built cases of fragmented observations prior to the terrorist attack. And use those as training datasets to trigger alerts to the plausibility of another terrorist attack. And that's where the high obstruction data came in. So one of the things I'm interested in doing, we talked about, let me talk about in the films, but it's the same thing is, this is the what I call a Little Miss Sunshine project. So I don't know whether you ever seen Little Miss Sunshine can be seen that? Yeah, it's brilliant. I mean, it's just hysterically funny. I mean, how anybody can say it's anti feminist. I do not know, it's just brilliant, right. But it was a low budget movie. There wasn't a blockbuster, the same with, you know, the Full Monty, for example, in the UK. So one of these we're interested in is how do we find the low budget films that will become blockbusters, because the studio's tend to go for the things which look like, you know, Avengers three, or whatever, right, in terms of what works. So what we can do is, um, there are two things we can do. And that one is we can with sensemaking, get the test audiences on films to actually interpret each scene rather than interpret the whole film. So in the in the pre showing of the film, that's potentially far more valuable. Yeah, and that's something I really want to work on, because we will we work the whole thing out. But we couldn't get, we couldn't get one of the studios interested. So it sort of it's in suspension. But we now have to do that. Because what you can then do is that data then also gives you and then you can look at the data in the build up to the production of a film. And we could of course, in simulation, we can create that data with young executives who don't know the outcome. So that creates training data sets, which can actually tell you look at this one more closely. So this may be the one idea in 500, which is actually going to come through so you can see how I can move that across into startup. Oh, yeah. Yeah, I can use I can and it's, I mean, this is a lot of sense maker was all about, it's about training datasets. It's how do I create training datasets to spot weak signals. That's how we started.
So like, if we look at the If I take it back one notch, I'm going to use this analogy to reflect it back on startups. So if we say that it let's take the the, the the terror prediction. So if in theory, what the trio's with the entangled trees would be able to achieve by by observing the story generated by people who are under, for instance, Al Qaeda, you'd be able to, theoretically be able to make a reasonable prediction to ISIS.
Yeah. And what we one of the things that this was devised, for instance, make was devised for was interagency collaboration. Because you, I mean, I still remember when, when the congressional reporter 911 came out, I mean, Dennis gone in myself and find x, and we're just screaming at people on the hill, you've got the wrong solution. You're putting everything into one big ministry with one big database. So you're reducing the requisite variety of discounting? Yeah, what you need to do is to create this keep separate stuff, what you need is something that says something odd is happening here. So we think and we ran through a full simulation on this, we could have got the flying schools, one, we could have brought that to the attention of an NSA. And that was the key signal, which everybody missed on the build up to 90,000 people who've been trained to take off and fly but not land. Yeah, now it was reported. But it was one of trillions of signals coming in. So the problem Poindexter gave me was how do I find the signals I need to pay attention to? And that's what these sort of methods do.
Yeah, brilliant.
The way you build prior knowledge in is not through rigid processes is through training datasets to trigger alerts. Yeah, and that's what sense maker indexing was designed for. Because without the abstract signification, you don't get sufficient ambiguity. So keyword searches don't work. Because they they lack requisite, requisite variety.
If you can just we've got Collins had his hand up for a little while too, but I know I'm in.
I'm in his I tend to dive into interest in discussions. We're doing that all the time. So I apologise. colleague go.
You're muted. Call it.
We can't hear you.
Oh, no.
Can I ask you something thing? Is it 301 diaries?
Sorry.
I'm just saying free or one diaries. mean you have an intake of real? And let's take the teacher parent.
Yeah. And the child's would you have free diaries? Would you have one? Yeah, you'd have three diaries and three lessons learned that conversations about is an agreement and review. Yeah. Even in a childcare example, the stuff we're doing with old people's homes in Netherlands at the moment. We're creating pap groups of relatives and old people in nursing staff, only sharing the stories within those groups. If anybody's interested in that, by the way, we're now ready to generalise that. Because that's the key thing in nursing homes where you can't meet people.
So if you had a data set, it would look like one of those novels where you have a lot of people's perspective on the same situation.
Yeah,
same timeline. Yeah, it is similar. And there is actually a brilliant IBM System which they they've taken off their website, which I love to write. And it was I kept finding things in IBM Research and nobody would let me commercialise them because the researchers just wanted to do research. But what they'd actually done is they taken the entire corpus of work of Philip Glass. I left it in class, I mean, and yeah, I'm proud of this. I've actually sat through Einstein on the beach without going for a total break that takes considerable training and preparation, because it's a five and a half hour opera, right. And it's beer. I mean, minimalism is beautiful. So they took all of his music. And they created a thing which identify common patterns, and then LinkedIn connected them and represented them. And as you played with the console, you can actually hear the music. So you discovered aspects of glasswork that you would no otherwise never have discovered? Yeah, that was one of the metaphors I use when we were building the counterterrorism thing. It's the ability to see complex patterns through abstraction. And that's what anything any NLP programme will never do, because it's looking for repetition, not novelty. So I'm going to talk about this on the other ones. we're doing is we're introducing an abstraction into the selection of the NLP criteria so that we can actually find novelty. But you don't do it if you're explicit. Currently, you are able to talk now, if I asked you to unmute.
Let's see if you can hear me now. Yay, we can go. Excellent. So that's my bluetooth wasn't working. So I was wondering, I work for you said, by this message, you do not need so much formal management anymore. Now I work in a large organisation of 30,000 people and their love for management is really hierarchical structured. And I use Canada to create value and they see that but at the same time, that seems to be no interest. How would you interest a formal organisation Oh,
acid, that's actually quite easy. You go to senior executives, and you say, we all know that the real work in his company has done over the over the over the watercooler conversations, and who you know matters more than where you are. So why don't we try to manage it? Okay, talk to the CEO, okay, he told the CEO kind of like I mean, SNS is a brilliant thing for CEOs. Because they say I'm you know, I trust my staff here is the programme or any group of staff that conform which can solve this problem, you get this remote, or the CEOs point of view, it finds the people who middle management screening. I can tell you, you go to the Board of a company and say, middle management to disguising people you need to talk to and they immediately jump on you and say yes, can we do something about it?
Yeah. I totally agree. And this is exactly what happens. And we call it a clay layer in our company. Don't worry, I live I live in chalk downs and clay or I know what a clay layer me. I was walking with him yesterday. Actually CO is approachable. I do speak to him from time to time. But before I get to talk to that, man, I need to go through the clay layer and align what I'm going to say that makes me makes it difficult. But once you actually have all the managers had to go to the toilet, there was a moment to talk about something else.
No, I definitely did. I don't want to believe you introduced a chemical into the tea caddy in order to induce that process. I have known people do that sort of thing. I think the field guide will help you. Yes, I'll go with that and say look, it says we should do this. I know the guys we can do something. Like the rest. Yeah. That's worth it. Because I will certainly do that. That's good idea. Yeah.
Thanks. Okay.
Who's next?
up very sorry, Paul, that was about to say I was choosing between you and Cindy as to who I called out for being too quiet. So you volunteered go for.
So somebody asked the question. What did the CEO does not believe it can be done.
And then basically, as a late adopter move companies, sorry. I mean, this is this is one a one year market lifecycle selling. And we used to train people on this. If you've got a late adopter company and late adopter buyer why you says trying to sell them something new. They won't buy it. Yeah. I mean, people did different people buy different things at different types of lifecycle, you need to the other thing, I mean, just this is just general guidance, you do not want to sell to an early adopter. Because basically anything you tell them, they'll tell you how they've already thought of it on what they thought is better than you anyway. And they're trying to use you to endorse some bloody stupid idea, which isn't anything to do with you. I get really fed up with these people. Yeah, what you want to sell is to early majority. Yeah, early majority one things that will make a difference, but they don't expect lots of references because they want first mover advantage. They don't want the pain of creating it. And I've got a lot of work in over in the years as to how you sit in this is the most him how you sell on the other side of a chasm. As you don't know, Moore's Crossing the Chasm go and read it is really important for anybody who wants to do innovation here. Because we are removed, we sat and looked at it when I was in charge of strategy. I said we can't afford to go through the belly chasm for new products. So we got to introduce novelty to people who actually have budgets rather than people who are wasting our time we call them tire kickers, if you know that friends. Yeah, and if you do novelty, you may a hell of a lot of tire kickers right. So a lot of the early methods I've developed were actually about selling on the other side of the chasm. And that's what you try and do. So find an intractable problem and create a risk free experiment. Yeah, one of the fascinating things I'll tell you the most effective way we've ever found this whole sense maker, is that this is an interesting set of ideas. Why don't we spend Two days training you and you do five pilot projects. And then we spend two days helping you do the analysis. And people will buy that. And they'll do some serious projects in it. If you enter them and said, We want to do one project, and it's really serious with this novel method, they turned it down. But if you say this is an experimental process to test ideas, we'll see what comes into it. And that works. It's a different psychology. We have standard templates for that, by the way, if you want to sell it. It's multiple sensing instead of trying to sell sensemaking. For one thing, you say we'll spend two days working together, then we'll create up to five sense maker projects, and then we'll work together on them, then we'll analyse them together. And it's really experimental and people feel safe about it.
Because it's an experiment.
Yeah, the crazy thing is they do some serious projects around what we did with the UN right un abusers for 10 years since we did this. One of really fat, two fascinating ones. One was to understand roamer attitudes in Eastern Europe, the other is post Chernobyl recovery. Right? So they there was some serious projects, which came into the process, but because senior management and licenced experiment, people came to the experiment with really interesting projects with the fact that we authority to do things which they wouldn't have had otherwise. So.
As Heidegger famously says, That man things is a master of language, language is the master of man. The way you describe things determines whether people will accept them or not. I, Cindy, I should provoke you. Yeah.
Yeah. Thanks. Um, based on that answer? How do you find the right rewards to make this work? Oh, well,
if you're talking about in tangled trails, which we're talking about, then actually, the peer to peer flow is actually we think enough, we don't know. But you might want to introduce some other stuff. The ability to actually have money for localised projects. So if a trio comes up with a project, you can have money. That's been very powerful. So when we did the stuff in South Wales, it was you got to pay you then become a trio. If the third degrees, then there's a budget. So the budget is predetermined. So it's in your interest to do this, because you might get investment in your local community. So those sort of things work.
And I'm stuck in my public sector mode. Again, though, when I'm going to unionised environments, you can't do that.
I hate tennis, we've done this with government. Okay, if you say there is this fun, and there is this control process for allocating the money, it's okay. Okay, yeah. I mean, if you want help on this, I mean, Chris, in the well shielded office, and I work this all the time.
I'm true. Yes. Okay. Thank you.
Just a question, obviously, entangled three years means we're talking about three people. But would it be thinkable to have more than three people like five or seven years? It might be and I think that's worth exploring, right? It's worth an experiment, there is a cognitive issue here on the limit. So I think five you could just about get away with but it will be difficult. Because this is Miller's number seven plus or minus two. So the cognitive load of handling more than five relationships is too high. Yeah, three is safe. Right. So I think, yeah, that there's that sort of principle coming. Yeah. Now, we said that roles might allow you to expand a bit more, but I, you know, is like, let's get the basics right first, and then build from that. But yeah, it might might be the case.
Can I assume you wouldn't use for because you might get split scary. Now
again, never have even numbers Always. Always have a is actually quite interested in asymmetry. I mean, human beings are asymmetric. Every every aspect of is asymmetric. We don't like symmetry.
It's interesting. If you want to take a picture of yourself, and take the left side of the face and mirror, it actually, it's scary. It's uncanny
valley, though, isn't it? That's when you start getting into the slightly
uncanny valley processes.
Dave, can I ask you, how would you apply this to leadership development? I've got a real question. Actually, today from the leadership team. They initially want to use sense maker to do assessment. I say Hold on there. It's not about
that's actually quite good. You can do that. I use entangled trios, which basically He says leader plus new employee plus somebody from an admin function. So basically a couple potential leader with brand new employee, and then have a series of roles that they can create on their trail. Yeah. And then let them run and see what comes out to your you could do what we did with Chanel, which is leadership group, still my favourite client of all time, because Chanel look after you, right? is basically leadership journaling for trainee leaders. Yeah. And then you give them tasks every week to interact and gather material from other people. Yeah, and of course, because they're on a leadership development programme, they can't like, do whatever you tell them anyway, I mean, that they can't not volunteer. Yeah. It's like new employees, new employees are the best interrogators I know. So in the first three months, you basically say, you've got to keep a daily journal, you're a new employee. And every week, we'll give you a different senior member of staff to go into. You want to capture knowledge from people about to retire, that works. Yeah, because yeah, there's an obligation to teach young people to join, there's not an obligation to teach your successor. In fact, you if you're honest, you really want your successor to fail. I mean, you we need to be honest about this sort of stuff for it. It's called the grandparent is called grandparents revenge, grandparents tell things to grandchildren, they won't tell two children and vice versa. So we just use that same phenomenon.
What about if it's a group of all pretty senior executives? And they are actually in the, in the cohort of leadership development programme?
Right. So okay, so a, you could do journaling, but B, you could do okay, we're going to pair each of you with somebody who joined the company in the last three months. And you got to work together as a pair. Because if you can't work with brand new joiners, who are very young, you're not a leader. And then for your trio, you can choose one of these groups. So you create a set of roles. Yeah. So it might be a middle manager, it might be somebody from HR, you can create those and see what they pick. And you got a short term project, I would do that short term, like you got, you know, three weeks in which you've got to intensively gather data and build it. Now the other advantage then as you start to create connections, which otherwise wouldn't exist. And the nice thing about new employees is they give people a completely different impression, because they haven't yet learned the politics. Or you can do even more clever, right? I've done this before, senior leaders with university students who want to be considered for employment. I mean, it's actually quite interesting. If you take a master student in a say an MBA into work with a really senior leader. They can ask them questions at the senior leader hasn't answered before, but they are zero threat. All right, then builds connections with the local college. And is it actually that's God is a classic convexity, but is there a purpose? Yeah, it's actually a recruitment process.
So that's the payer, right. So the leader and the student and then you will want an N West a trio, then
you give people a series of roles and the pair choose which of the roles they want to go with. That's how you choose for it. So you work out what those roles are, and they choose, you give them a degree of freedom on that, because who they choose will be relevant.
What happens if conflict breaks out in your tree? Oh,
that's fine. That tree, it doesn't work, but you got 20 or 30 of them working in parallel.
If an individual tree in a forest gets struck by lightning in the forest doesn't worry about it. I'm just wondering if we're going from Bonnie's question, and you wanted to pair the senior leaders with them, young person asking them question like asking them questions about the dirty laundry in the company?
Well, first of all, they're not going to answer that. And that's stupid. All right, but they are going to answer questions. And to be honest, if a senior leader can't form a cohesive group with a young with a young potential employee, I wouldn't put them as a leader anyway. I'd use it as a test. I actually do it on Legion Gen process if you can't form this trio. Oh, you're not gonna survive as a leader anyway, you're an Arrogant Bastard.
Hey, what's, what's the duration? you have in mind?
I dreamed that for a month, month maximum,
you know, resolve conflict to because you said the sun staff.
Yeah, I did it for an intensive month because also that would work with a college but you can imagine for example, you know, Barney, you go to Cambridge University. I mean, right, you're right. Next door, you're on the Science Park and you say you go to the judge Institute and say we got we need Yeah, 50 MBA students at the judge or MSC students who will pair with our senior leaders, and it's a golden opportunity, I mean, God, you're gonna get you're gonna be swamped by volunteers. And they're the detail work. But that's the way you get the knowledge out of those guys. And then you have different different trios, the trios could be external. Yeah, that you could say, one of things you can do is get somebody in the ministry of health or get somebody as a doctor in local, you could actually do the distance in the disintermediation trick on that, and see which pairs can find people, the whole end user clients, to work with them. This is a classic complexity technique. It's a really simple process, which generates multiple types of unexpected results. And that's what you want in convexity, if we all we're doing is defining the linkages. So we're using the
problems that I mean, you define the linkage is up to them, to them to find solutions. So
you're but also what you actually find is they'll find solutions to problems you haven't defined. Yeah, that's another side benefit.
Can I jump in on diaries? So, so if we, because we touched upon the fact that a trio is going to be he is going to each potential, you're going to be writing a diary, we got three levels of privacy, my cell phone in my tree on public, but if we want to standardise and we want to do for instance, we want to run an algorithm to basically go over 1000s of diaries, with a format for the diary being required recommended
me to do that was sense maker, because the way the the way they interpret this story is actually not confidential. That's the big breakthrough medical research with sense maker is you can keep the story confidential or the picture confidential, but you share the metadata. So the metadata is already done at the point of entry. And that gives you all the analysis you want.
Even in the open ended questions, like even in a like, formative, open ended question we Yeah, that's flexible, or should it be formatted?
That's how sensemaking works. It works on high obstruction metadata. Yeah. And yeah, basically, you can withhold the content, but you share your interpretation, because that's neutral. Yeah. And that's actually what you want. That's how we want data, it means we get huge volumes of quantitative data. And this is my physics background coming through. It's kind of like, you know, say there's loads of notional scientists, advocates NetFlow the data, they can't form a conclusion. So what I'm doing is I want a lot of data. And that's what this does.
So if you don't share the actual stories is suggesting that so let's say the leaders to the journal, then they see the patterns, right? How they do the signifier,
and you see they can ask for the story. So you can build that in. So if I was interested in is people will disclose a narrative if somebody asks for it, and they say why they need it?
And is this to encourage them to have a discussion around the patterns that they discuss the
investigation, sharing discovery, novelty, that's what you're building? You're actually building something which is a scalable version of watercooler conversations. Okay, guys, it's coming up to the hour. Yeah. So we're now into wiki production. I will be issuing an email tomorrow on this because, I mean, there's some great discussions going on in slack but kind of like we need to shift from exploring the subject into documenting the methods. Alright, so that's the big tasks that we're going to start to move into. So cathead is I'll be in contact with you guys as well, because now we need to drive it. Okay, I may see some of you on the next one. Yeah, my last one before dinner. Okay, so you want to the minute some of you say bye bye.