Latest / How I Work / The two AI mistakes hurting your team's productivity, with Dom Price
Transcript
- 0:00Speaker 1: Like, if you have inefficient and ineffective processes and people's
- 0:04Speaker 1: systems and you layer in AI, you are doing stupid
- 0:07Speaker 1: things faster.
- 0:08Speaker 2: That was Dom Price, former Elassian work futurist and one
- 0:13Speaker 2: of the sharpest things I know on how organizations actually work.
- 0:18Speaker 2: Dom joined me off the back of some brand new
- 0:21Speaker 2: Australian research that found that ninety percent of Australians are
- 0:26Speaker 2: using AI at work, but only ten percent say it
- 0:29Speaker 2: is significantly improving organizational performance. That stat probably doesn't surprise
- 0:36Speaker 2: you either. I mean, we've all seen it AI rolled
- 0:41Speaker 2: out organizationwide heaps of enthusiasm, and somehow the way we
- 0:45Speaker 2: actually work looks exactly the same. Dom and I get
- 0:50Speaker 2: into why so many AI bets are landing flat two
- 0:55Speaker 2: new terms that I'm now obsessed with, bot sitting and
- 0:59Speaker 2: bot shitting, and why your human operating system might still
- 1:04Speaker 2: be running Windows ninety five. This is one of those
- 1:08Speaker 2: chats that I think makes you want to pause and
- 1:10Speaker 2: really rethink how your team works before you layer on
- 1:15Speaker 2: yet another AI tool. The research from Glean's Work Ai
- 1:41Speaker 2: Institute found that ninety percent of Australians are using AI work,
- 1:47Speaker 2: but then we've only got ten percent saying it's significantly
- 1:50Speaker 2: improving organizational performance, which is a stat that doesn't surprise me.
- 1:55Speaker 2: But Dom, can you shed light on what you think
- 1:57Speaker 2: is happening here?
- 1:58Speaker 1: Yeah, it's the thing. I think some of it's the
- 2:01Speaker 1: stage we're in. We've not quite realized what it is
- 2:04Speaker 1: we're meant to do with this mythical thing called day I.
- 2:07Speaker 1: So we're using it. You'll notice the research doesn't say
- 2:09Speaker 1: what we're using it for, footy tipping, World Cup predictions,
- 2:12Speaker 1: like we could be using it for anything. And I
- 2:14Speaker 1: think that's the reality. Is a lot of the organizations
- 2:16Speaker 1: that I'm working with have done this kind of I
- 2:18Speaker 1: call it the Woodstock theory, like book them and they
- 2:21Speaker 1: will come like put AI out there and people just
- 2:23Speaker 1: work out how to use it without really giving them
- 2:25Speaker 1: the full training and embedment and what does it mean
- 2:28Speaker 1: and how do we activate? So people are using it,
- 2:31Speaker 1: but are they getting the tangible benefit and even the
- 2:33Speaker 1: magic word that means you love the magical word of productivity,
- 2:37Speaker 1: like I feel more productive, but the organization isn't getting
- 2:40Speaker 1: any benefit. It's like me saying I watched the fitness video,
- 2:43Speaker 1: I feel fitter. I've not done any exercise, right, so
- 2:46Speaker 1: we've got to actually follow through on this stuff to
- 2:48Speaker 1: get the benefit. Otherwise it's just noise. And I think
- 2:51Speaker 1: that's the stage we're at. We've overcome the hump of
- 2:55Speaker 1: this is mythical and it's made up. We've overcome the
- 2:58Speaker 1: fear factor. We're now using it, but we're not using
- 3:01Speaker 1: it in earnest right. We're using it to do old things,
- 3:03Speaker 1: probably faster and maybe old things we shouldn't have been
- 3:06Speaker 1: doing anyway. So I feel like the rest of twenty
- 3:08Speaker 1: twenty six is a time for us to go pause. Right,
- 3:11Speaker 1: this thing exists now, it's mainstream. How do we change
- 3:15Speaker 1: how we work so that that ten percent becomes significantly high?
- 3:19Speaker 1: Because if we're not getting a benefit from it, then
- 3:21Speaker 1: what's the point? Like that's the extental question, meaning you
- 3:23Speaker 1: ask all the time, if you're doing something and it's
- 3:25Speaker 1: not adding incremental value short term, medium to long term,
- 3:29Speaker 1: why would you do it? And then a actually be
- 3:30Speaker 1: no different than any other investment decision we make.
- 3:33Speaker 2: So two terms that I had not heard before chatting
- 3:37Speaker 2: with you about this stuff dom is bot sitting and
- 3:41Speaker 2: bot shitting. Yes, can you tell me what these two
- 3:46Speaker 2: activities involve?
- 3:49Speaker 1: The bot sitting is a fascinating one. It's the hidden work.
- 3:52Speaker 1: It's every colleague that goes they say something, You're like, oh,
- 3:55Speaker 1: that look's amazing. Only topy five minutes. We feel like
- 3:58Speaker 1: we need to brag about how little time things take.
- 4:01Speaker 1: And you're like, did it take you five minutes? You're like, well,
- 4:03Speaker 1: the output top five minutes, but it's only seven hours
- 4:06Speaker 1: of programming, you know, going through the different lems, go
- 4:09Speaker 1: through different models, different prompts, different answers, going through feedback loops.
- 4:12Speaker 1: Eventually I got something out and by the way, I
- 4:15Speaker 1: have to now look after that, right, like a tamagotchi.
- 4:17Speaker 1: I have to nurture it so that that output stays good.
- 4:21Speaker 1: And that accounted for about six and a half hours
- 4:23Speaker 1: a week. So yes, we feel more productive, but that's
- 4:27Speaker 1: because we're doing one bit faster, but we've added another task.
- 4:30Speaker 2: But I mean, like, if you think about what that
- 4:32Speaker 2: looked like before AI, I mean, is that just me
- 4:36Speaker 2: as a human instead like doing that editing, reviewing work,
- 4:40Speaker 2: or like, how do we know that we weren't just
- 4:43Speaker 2: doing that as a human anyway, and that instead we're
- 4:45Speaker 2: just wasting time with AI or not wasting improving quality
- 4:48Speaker 2: hopefully That's kind.
- 4:49Speaker 1: Of the existential question that a lot of the leaders
- 4:51Speaker 1: I'm working with right now, we're asking, which is have
- 4:53Speaker 1: we just moved stuff around? Is it spending the same
- 4:56Speaker 1: total amount of time the same elapsed or duration? Is
- 4:59Speaker 1: this say, hopefully the quality is good up, but that's
- 5:02Speaker 1: questionable because we'll come on to bot shitting in a minute.
- 5:04Speaker 1: Sometimes the outputs and outcomes have gone down. But if
- 5:08Speaker 1: we're doing the same amount of time but we just
- 5:10Speaker 1: move the tasks around, are we getting any benefit? And
- 5:12Speaker 1: we've added in a holder cost? Right, AI is not free.
- 5:15Speaker 1: It's a high cost resource, both in terms of actual
- 5:18Speaker 1: monetary cost but also planetary costs. So let's do this
- 5:20Speaker 1: really carefully, I think, before we throw it away and
- 5:24Speaker 1: go it's the same thing, therefore it's not worth it.
- 5:26Speaker 1: I think this is the pause moment, And there's quite
- 5:29Speaker 1: a few leaders I'm working with now where it feels
- 5:31Speaker 1: like a measure twice cut once moment, Like just there's
- 5:35Speaker 1: no mad race to forget the phone moment. You're not
- 5:37Speaker 1: going to miss out right, just stop for a minute
- 5:39Speaker 1: and have a look at how your business operates and
- 5:42Speaker 1: ask yourself that there's an exercise I've been doing on
- 5:44Speaker 1: like minus one zero one, right, So minus one is
- 5:47Speaker 1: using AI to do stupid things faster. So let's imagine
- 5:50Speaker 1: we have thirty hours of meetings a week in Mantha
- 5:52Speaker 1: and they're awful, right, they destroy our soul, they destroy
- 5:55Speaker 1: our mojo. We cry at the end of the week.
- 5:57Speaker 1: But we have them because we have to. And we
- 5:59Speaker 1: get AI and we use AI to take transcripts of
- 6:02Speaker 1: the meetings we don't like. So now we have thirty
- 6:04Speaker 1: hours of meetings and thirty hours of transcripts to read. Right,
- 6:07Speaker 1: you're even more depressed. You've added no value. So what's
- 6:10Speaker 1: the point? Zero is saying? How do we change the
- 6:13Speaker 1: meetings we have? I cancel some add some new ones in,
- 6:17Speaker 1: change the format, the frequency, who attends, like, what happens
- 6:20Speaker 1: the gender?
- 6:20Speaker 3: Like?
- 6:20Speaker 1: How do we make our meetings better? And how do
- 6:23Speaker 1: we use AI to improve them? To take note, to
- 6:26Speaker 1: take actions, to take follow through so that the end
- 6:28Speaker 1: of one meeting is the start of the next. We're
- 6:30Speaker 1: holding each other accountable for the actions we're meant to take.
- 6:33Speaker 1: Our meetings improve, our outputs improved. That's a zero. You're
- 6:37Speaker 1: still doing the same thing, but more efficiently and effectively.
- 6:40Speaker 1: And the magic thing that everyone's after in the hype
- 6:42Speaker 1: cycle is the plus one. What's a new thing I
- 6:45Speaker 1: can do with AI? The thing is we're all at capacity,
- 6:49Speaker 1: so we only get to do the plus one if
- 6:51Speaker 1: we've taken out some of the minus one and the
- 6:53Speaker 1: zero's first, if we've got the efficiency gain. So there's
- 6:56Speaker 1: an element of pressure on leaders to go you've got
- 6:59Speaker 1: a clean house first, Like if you have inefficient and
- 7:02Speaker 1: ineffective processes and people's systems and you layer in AI,
- 7:06Speaker 1: you are doing stupid things faster, like please don't do that,
- 7:10Speaker 1: and I'm seeing that suddly way too much versus going,
- 7:12Speaker 1: how do you understand that human operating system first? Then
- 7:16Speaker 1: get the improvement, then lay on AI. And if you
- 7:18Speaker 1: do that, you should get way less bot sitting right
- 7:21Speaker 1: cause you should be doing the right things.
- 7:22Speaker 2: We'll get to bot shitting in a moment, but on botsitting,
- 7:25Speaker 2: what I have found, and I mean, you know, for
- 7:29Speaker 2: regular listeners of this podcast, they will know that I
- 7:31Speaker 2: always bang on about the importance of AI training, partly
- 7:34Speaker 2: because I run an AI training company. But what I
- 7:37Speaker 2: found when I think about my own processes and when
- 7:40Speaker 2: I hear about how other people are using AI and
- 7:42Speaker 2: when I think about how much time is spent botsitting,
- 7:44Speaker 2: so reworking and arguing with the AI, and do more
- 7:47Speaker 2: of this and do less of this, And like, the
- 7:49Speaker 2: longer you spend in a chat thread arguing with AI,
- 7:52Speaker 2: the least chance you've got to actually at the bottom
- 7:55Speaker 2: of that thread something really great coming out. And so
- 7:58Speaker 2: a lot of it is skilled that can be taught,
- 8:00Speaker 2: like and what I see one of the biggest mistakes
- 8:03Speaker 2: that I see people making is that they know what
- 8:05Speaker 2: the AI is doing wrong. They're like, I don't like that.
- 8:08Speaker 2: I don't like that. AI tell I don't like you know,
- 8:11Speaker 2: how you've structured this thing that I need to do
- 8:13Speaker 2: every week, like you know, case in point, creating the
- 8:15Speaker 2: show notes for how I work. And if I just
- 8:18Speaker 2: argued with the AI every time, I would spend so
- 8:21Speaker 2: much time bot sitting to do that particular task. And
- 8:24Speaker 2: I think just the very simple unlock is using the AI,
- 8:28Speaker 2: whether it's improving the skills that you've trained in in
- 8:31Speaker 2: claude or your GPTs or agents that you've created to
- 8:35Speaker 2: actually get them prompt them to unpack what have you
- 8:40Speaker 2: learned from this chat thread and update your system instructions
- 8:44Speaker 2: or update this skill and I know for me just
- 8:47Speaker 2: at a personal productivity level, let alone team and organization productivity,
- 8:52Speaker 2: that for me has.
- 8:53Speaker 3: Led to a lot less bot sitting. How do we
- 8:56Speaker 3: reduce bot sitting?
- 8:57Speaker 4: Like?
- 8:57Speaker 3: What have you found in your life?
- 8:58Speaker 4: Dom?
- 8:58Speaker 1: The first task is get it for purpose first in
- 9:01Speaker 1: a low fidelity, simplest like pen and paper. Right, So,
- 9:04Speaker 1: if I think I had a sparing session this morning
- 9:06Speaker 1: with one of my colleagues, it was pen and paper,
- 9:08Speaker 1: low fidelty sparring first. We then used AI to accelerate
- 9:12Speaker 1: our thinking, and we've then come back and done a
- 9:14Speaker 1: higher fidelity riff on that to refine it and make
- 9:17Speaker 1: sure it's exactly fit for purpose. The acceleration in the
- 9:20Speaker 1: middle that would have taken weeks the research and the
- 9:23Speaker 1: fidelty it helped us get to was in a record time.
- 9:25Speaker 1: We would have got there anyway, but it's accelerated our
- 9:28Speaker 1: path there so we've got quicker time to value. But
- 9:31Speaker 1: we were very specific about where we did a human
- 9:34Speaker 1: a premium thing, and this is the bit a lot
- 9:36Speaker 1: of people are missing. I think in business right now,
- 9:39Speaker 1: a lot of consumers are happy to pay a premium
- 9:41Speaker 1: for a human experience on certain things. That don't want
- 9:44Speaker 1: everything to be automated. So there's a large O i't
- 9:48Speaker 1: name them for obvious reasons, but a large mortgage provider.
- 9:51Speaker 1: They have a bricks and mortar store, they have an
- 9:53Speaker 1: internet service, and they have an agent surface. Right, because
- 9:56Speaker 1: there's lots of people that want to procure a mortgage
- 9:59Speaker 1: who will favor bricks and water store and they will
- 10:01Speaker 1: pay a premium for doing that. Don't exclude them because
- 10:04Speaker 1: the whole world's gone AI. So this is where we
- 10:07Speaker 1: need to just understand, like are these shifts real? And
- 10:10Speaker 1: where is our market? Do we need to be first
- 10:12Speaker 1: mover or is it okay to have a different distribution
- 10:15Speaker 1: channel based on what our end users or end customers needs.
- 10:18Speaker 1: And I don't think that takes a huge amount of
- 10:20Speaker 1: thinking for a leader. I just I want to empathize
- 10:23Speaker 1: with the most leaders that I work with right now.
- 10:25Speaker 1: They are so overwhelmed, right like that is AI this.
- 10:28Speaker 1: I've got multiple generations in the workforce awaken in the morning,
- 10:31Speaker 1: I look at the news a panic. Is the share
- 10:32Speaker 1: market going up? Is it going down? Or interest rates
- 10:34Speaker 1: going up? We're going down? And should to buy a
- 10:36Speaker 1: house for to sell house? Like there's so much overwhelming
- 10:39Speaker 1: cognitive load on these leaders. They haven't got that nth
- 10:42Speaker 1: cycle to go. I wonder if I reinvented how my
- 10:45Speaker 1: teams work, if I could get some tangible benefit, like
- 10:48Speaker 1: they know, intent wise, they should be doing that. They've
- 10:50Speaker 1: just not got the capacity. And that's where your training thing.
- 10:53Speaker 1: That's the un lot that that training is permissioned to
- 10:55Speaker 1: go and play. And when you play with this stuff,
- 10:58Speaker 1: you really go and accelerated cycle. If you do it
- 11:01Speaker 1: because someone told you to do it, you're just going
- 11:03Speaker 1: to layer on power onto old ways of working.
- 11:06Speaker 2: Let's talk about bot shitting. Now, tell me what is
- 11:09Speaker 2: bot shitting because I love the term.
- 11:11Speaker 1: If you think box sitting is as a human having,
- 11:14Speaker 1: as you explained, the elongated argument with the bot, which
- 11:17Speaker 1: for some reason you feel they need to win against
- 11:20Speaker 1: a token machine that is really predictively guessing what you
- 11:22Speaker 1: think instead of doing that. Bot shooting is where you
- 11:25Speaker 1: outsource everything to the bot because obviously it's better, and
- 11:28Speaker 1: you share all the answers without any context and without
- 11:31Speaker 1: any verification, so you take yourself completely out of the loop.
- 11:34Speaker 1: We've seen lots of press about this recently, different professional
- 11:38Speaker 1: services firms sharing reports with huge hallucinations with citing research
- 11:43Speaker 1: that doesn't exist. Right, So bot shitting is where you've
- 11:46Speaker 1: put so much faith in AI that you've taken the
- 11:48Speaker 1: human out of the loop, and actually you've lowered the
- 11:51Speaker 1: quality bar. And this is not just the slot that
- 11:54Speaker 1: we see regularly, right that we were like that was
- 11:56Speaker 1: clearly in BII. There's no thought, there's no compassion, there's
- 11:59Speaker 1: no real understand you just doing that to get clicks.
- 12:01Speaker 1: It's not just that, it's actual organizations shipping informative reports
- 12:06Speaker 1: with opinions where those opinions haven't been verified by a
- 12:10Speaker 1: human even their basic thing is the research that we
- 12:14Speaker 1: reference actually real. And when you see this someone, I've
- 12:17Speaker 1: seen the press recently with like fifty sixty seventy percent
- 12:20Speaker 1: of research cited was hallucinations. It didn't exist. Now, the
- 12:24Speaker 1: danger for that botshit thing I find mildly amusing because
- 12:28Speaker 1: it's just the way my mind works. It makes me laugh.
- 12:30Speaker 1: Then I have this little moment of fear where I'm like,
- 12:32Speaker 1: if you publish that, the average end user will believe it.
- 12:36Speaker 1: So we've created not just a myth, we've created an
- 12:39Speaker 1: untruth in market where we've cited something with high confidence
- 12:44Speaker 1: that is not true. So this is where we need
- 12:46Speaker 1: to look at our governance models, our entire human system
- 12:49Speaker 1: of how we think about what we share, how we
- 12:52Speaker 1: corroborate what's in any AI tool, and how we think
- 12:55Speaker 1: about our values right and our sort of morals as
- 12:58Speaker 1: an organization. How do we verify a level of trust
- 13:01Speaker 1: in the information we share and how do we do
- 13:04Speaker 1: the right level of verification? And without that, all we've
- 13:07Speaker 1: done is create noise. And I've not met a sing
- 13:09Speaker 1: good person yet. It's like, you know what, I'd like
- 13:11Speaker 1: more information? Like no one's something. I just I've not
- 13:14Speaker 1: got off emails on a daily basis. I'd like more emails,
- 13:17Speaker 1: maybe some more meetings, maybe some more meeting notes, that'd
- 13:20Speaker 1: be great, Like no one's after more. What we're craving
- 13:23Speaker 1: is context. And I think one of the things that
- 13:26Speaker 1: I've got out some of this recent research is we're
- 13:28Speaker 1: lacking context. And so when we're pouring fuel on content,
- 13:32Speaker 1: we're creating more content. But as and users were like,
- 13:35Speaker 1: I want what's the next action I should take? What
- 13:38Speaker 1: decision do I need to make? Which project needs more
- 13:40Speaker 1: help and which team's fine and needs to just leave
- 13:43Speaker 1: them alone? We want rich context but what we're getting
- 13:46Speaker 1: is noise, and box shitting is a sign that we're
- 13:48Speaker 1: creating more noise. The scary thing is we're creating noise.
- 13:52Speaker 1: That's very believable.
- 13:53Speaker 3: How do we overcome this huge problem of boch shitting?
- 13:56Speaker 2: Because I find myself certainly in conversation that I'm having
- 14:02Speaker 2: with our clients at inventingum AI and in keynotes and
- 14:05Speaker 2: things like that, is I talk about this AI slop avalanche,
- 14:09Speaker 2: and I get that there's a difference between bot shitting
- 14:11Speaker 2: and AI slop, and it's almost like the ramifications of
- 14:14Speaker 2: bot hitting are a lot bigger. They're also very public
- 14:17Speaker 2: facing as well. And what I feel like leaders haven't
- 14:22Speaker 2: got their head around is that volume no longer equals
- 14:27Speaker 2: quality Because in the past, like four years ago, if
- 14:30Speaker 2: someone were on my team were to come to me
- 14:33Speaker 2: with a fifty page competitor analysis and strategy plan for
- 14:38Speaker 2: the next financial year, I'd be like, oh, man, like,
- 14:41Speaker 2: so much work's going into this is probably really high quality.
- 14:43Speaker 2: They've probably spend a lot of time on it. It's
- 14:45Speaker 2: a really long report, and so therefore volume equals quality,
- 14:48Speaker 2: and that is what we've been taught to believe in
- 14:50Speaker 2: the workplace. But now volume does an equal quality, and
- 14:55Speaker 2: quite often there's an inverse relationship, particularly with those that
- 14:58Speaker 2: are not skilled at using AI. And then, and I know,
- 15:02Speaker 2: like from other research that I've read, managers are becoming
- 15:05Speaker 2: this huge bottleneck because all their staff for producing things
- 15:09Speaker 2: at lightning speed often quite a large amount of volume.
- 15:12Speaker 2: And now the managers like, oh my god, I've become
- 15:15Speaker 2: like a vigilance machine. I've become a document checker and
- 15:19Speaker 2: that's really sad. Where we've got documents that are created
- 15:22Speaker 2: by AI. And then if managers are just using the
- 15:24Speaker 2: I to check the documents, I mean, then there's like
- 15:26Speaker 2: a whole world of mess. So what are you seeing
- 15:29Speaker 2: as the solutions to botchhitting and AI slop or work
- 15:33Speaker 2: slop or whatever we want to call it.
- 15:35Speaker 1: I think whilst it's a laughable moment, it's the thing
- 15:37Speaker 1: that if we don't corral and control or govern in
- 15:40Speaker 1: a sensible fashion, we'll be looking back in a few
- 15:42Speaker 1: years going like what did we do here? Like we
- 15:45Speaker 1: will place really confident bets on the wrong thing because
- 15:48Speaker 1: we trusted the wrong insight. So part of this comes
- 15:51Speaker 1: back to again the pause moment of going what's the
- 15:53Speaker 1: total cost of ownership here, like, let's look at this
- 15:56Speaker 1: at a system level, whereas I think what we've been
- 15:58Speaker 1: doing is looking at a transactional level. I can use
- 16:01Speaker 1: AI to reduce my core volume and be like, what
- 16:04Speaker 1: if speaking to your customers was actually an opportunity to upsell?
- 16:07Speaker 1: So you save time, but you've missed the human moment,
- 16:10Speaker 1: or like I can use AI to do this, yeah,
- 16:12Speaker 1: also someone else can use it to do that, right,
- 16:14Speaker 1: I think we need to start internally and actually practice
- 16:16Speaker 1: on ourselves. Right. So there's an organization I was working
- 16:19Speaker 1: with recently. They have a structure about each sub team
- 16:23Speaker 1: writes a memo about what happened in their month. That
- 16:25Speaker 1: memo gets scaled up to the functional leader and they
- 16:28Speaker 1: sarve the memo goes up to it, and they're like,
- 16:30Speaker 1: let's use AI to do that. Now that there was
- 16:33Speaker 1: a person in the room who's like, hand up our
- 16:35Speaker 1: source data is terrible, and everyone's like, yeah, but it
- 16:39Speaker 1: used to take two weeks and now it's only going
- 16:41Speaker 1: to take two days. It's still terrible, and please don't
- 16:46Speaker 1: make any decisions off the back of that. So actually
- 16:48Speaker 1: their first step was to go, this isn't sexy work.
- 16:51Speaker 1: We need to get a data governance structure right of
- 16:54Speaker 1: what do we record, where do we record it, what's
- 16:56Speaker 1: our name? Culture like, it's very simple base business things.
- 17:00Speaker 1: But when they got that, they could then go as
- 17:03Speaker 1: a system, what does this look like? But the questions
- 17:05Speaker 1: one asked, like three hours into this workshop, this amazing
- 17:08Speaker 1: question which was imagine we've done all that and the
- 17:11Speaker 1: answer is six. Let's say we get the final document
- 17:15Speaker 1: out and the answer is six or maybe they or
- 17:17Speaker 1: maybe it's four. What are we going to do with
- 17:19Speaker 1: that number? And everyone looked at each room was like,
- 17:22Speaker 1: I don't know, and he's like, again, you're doing something
- 17:26Speaker 1: because you once did, versus going like, if I'm going
- 17:30Speaker 1: to get an insight from this, what action am I
- 17:32Speaker 1: going to take? So there's this is a pre AI example.
- 17:36Speaker 1: But there's a chief digital officer I was working with
- 17:38Speaker 1: a few years ago. They were new Enroll and the
- 17:40Speaker 1: very first thing that did they owned the reporting team
- 17:42Speaker 1: that did all the dashboards analytics right, And you know,
- 17:44Speaker 1: every business that we work with is like, I need
- 17:46Speaker 1: more dashboards, I need more analytics. You're like, you've not
- 17:50Speaker 1: made a decision in the entire time I've been here.
- 17:51Speaker 1: So I'm not sure any more information will help. But
- 17:54Speaker 1: what this guy did was he locked down every single
- 17:56Speaker 1: report and password protected it, and he went back to
- 18:00Speaker 1: the board meeting the following quarter and it's like, here's
- 18:03Speaker 1: the answer, like seventy six percent if you didn't open
- 18:05Speaker 1: your reports so you're not using the and like, no,
- 18:09Speaker 1: I used last quarter's report. It's like even worse. That
- 18:12Speaker 1: definitely had the wrong data in it. But you never
- 18:14Speaker 1: asked for the passwords, so I know you never accessed it,
- 18:17Speaker 1: but you've asked for more reports because we made reporting free.
- 18:21Speaker 1: I think one of the struggles we've got. I don't
- 18:22Speaker 1: know if it's like a uniquely human thing or just
- 18:24Speaker 1: the system we're in, of the change we've gone through
- 18:27Speaker 1: the last few years. A lot of the leaders that
- 18:29Speaker 1: I'm sort of coaching and mentoring that night, now I'm like,
- 18:31Speaker 1: if I gave you a magical AI tool and it
- 18:33Speaker 1: saved you ten hours a week, and a genuine ten
- 18:35Speaker 1: not a bot sitting ten hours, but genuine ten hours,
- 18:37Speaker 1: what's the next best task that you would do with
- 18:39Speaker 1: those ten hours. I think they've never genuinely thought of
- 18:44Speaker 1: having that much time free, so they don't know the
- 18:46Speaker 1: next best tasks to do so they fill the time
- 18:49Speaker 1: with lower value tasks.
- 18:50Speaker 2: That reminds me of something that is a question that
- 18:53Speaker 2: I will often ask leadership teams when I start working
- 18:55Speaker 2: with them around AI, and I'll ask just really simple question,
- 19:00Speaker 2: what's your why? Like why are you embracing AI? And
- 19:02Speaker 2: the answer is not because everyone else is like there
- 19:06Speaker 2: needs to be a legitimate reason.
- 19:07Speaker 3: Are you embracing it.
- 19:09Speaker 2: To get really big efficiencies so you can fire half
- 19:12Speaker 2: your workforce and increase your share price?
- 19:14Speaker 3: And if so, let's be honest about that.
- 19:15Speaker 2: Are you embracing it so that people can maybe work
- 19:18Speaker 2: a forty hour week rather than a sixty hour week?
- 19:21Speaker 2: Are you embracing it so that people can have more
- 19:23Speaker 2: time with customers face to face or at least synchronous
- 19:26Speaker 2: And so few organizations can identify this, like are you
- 19:31Speaker 2: finding this as well?
- 19:32Speaker 1: I do a version of that game. The only thing
- 19:34Speaker 1: I like the other slight difference is what would you
- 19:37Speaker 1: stop so now you know the power of this magical,
- 19:40Speaker 1: mythical AI unicorn thing, what are the things that you
- 19:42Speaker 1: no longer need to do? So companies I just find
- 19:44Speaker 1: they have a massive buyas for addition, and we're terrible
- 19:47Speaker 1: at removal. Basically, all we've done is add barnacles to
- 19:50Speaker 1: a ship, and then we said to the team, can
- 19:52Speaker 1: you make the ship go faster? And I keep on
- 19:55Speaker 1: buying faster motives, but the ship won't go faster because
- 19:57Speaker 1: it is literally wedded with barnacles. Maybe the job is
- 20:00Speaker 1: to remove some barnacles.
- 20:01Speaker 4: Right.
- 20:02Speaker 1: So the subtraction game with leaders is fascinating. The other
- 20:05Speaker 1: one is And I did this with a team recently.
- 20:07Speaker 1: I was like when you had your wonderful loss site
- 20:10Speaker 1: before the end of the year, and you went to
- 20:11Speaker 1: the vineyard and you got drunk and you came up
- 20:13Speaker 1: with a strategy. I imagine there was a massive argument
- 20:16Speaker 1: about all the things you wish you could have done
- 20:18Speaker 1: that you couldn't do because you were capacity constrained. Where
- 20:21Speaker 1: are they? Where are the other opportunities that you thought
- 20:24Speaker 1: were market makers but you just couldn't do it? Maybe
- 20:28Speaker 1: now you can. There was a company recently. Their CEO
- 20:31Speaker 1: took their previous two years business cases that they said
- 20:34Speaker 1: no to and they re evaluated them all using AI,
- 20:38Speaker 1: not putting them through a by saying through the lens
- 20:40Speaker 1: of AI, are there any of these no's that would
- 20:43Speaker 1: become a yes, and a couple of them became a yes.
- 20:46Speaker 1: Not as a final launch product in market, but the
- 20:49Speaker 1: chance to do a proof of concept and a pilot
- 20:51Speaker 1: was infinitely cheaper using AI capabilities than using their old
- 20:55Speaker 1: human capability. So to your point, before that research paper
- 20:58Speaker 1: that used to take months take a team of two,
- 21:00Speaker 1: you can now do in ten minutes. You're like, Okay,
- 21:02Speaker 1: maybe we can get to an answer quicker. It's not
- 21:04Speaker 1: that we're approving the business case to go to production.
- 21:07Speaker 1: We're approving the learning curve of can we get more
- 21:10Speaker 1: informed decision making? That was golden Now a couple of
- 21:13Speaker 1: them made it through and they found new addressable markets brilliant.
- 21:17Speaker 1: The ones that didn't, they were more confident in their
- 21:19Speaker 1: decision making of why they said no because they did
- 21:21Speaker 1: approve the concept. They tested it and it became a null.
- 21:25Speaker 1: But it costs ten thousand dollars, not not ten million dollars,
- 21:28Speaker 1: and so our cost of experimentation goes down. That's a
- 21:31Speaker 1: great upside for companies. It just does you don't get
- 21:33Speaker 1: the benefit of that if you haven't got a cultural
- 21:35Speaker 1: experimentation coming up. Next.
- 21:38Speaker 2: Dormer shares the heuristic that he uses to decide when
- 21:42Speaker 2: to bring AI in and when to keep it out.
- 21:45Speaker 2: He also shares why he built a board of directors
- 21:48Speaker 2: inside Claude to challenge his own blind spots, and the
- 21:51Speaker 2: question he asks every leadership team that almost no one
- 21:56Speaker 2: can answer. If you're looking for more tips to improve
- 22:02Speaker 2: the way you work, can live. I write a short
- 22:05Speaker 2: weekly newsletter that contains tactics I've discovered that have helped
- 22:08Speaker 2: me personally.
- 22:10Speaker 4: You can sign up for that at Amantha dot com.
- 22:13Speaker 4: That's Amantha dot com now dom. I want to kind
- 22:21Speaker 4: of switch gears and get into.
- 22:23Speaker 3: How you use AI, because.
- 22:25Speaker 2: I don't think we've ever talked about this, and quite frankly,
- 22:27Speaker 2: I am fascinated. So to start with, I feel like
- 22:30Speaker 2: you're someone that has a really interesting decision making heuristics
- 22:34Speaker 2: for things, and so walk me through how do you
- 22:37Speaker 2: personally decide whether to use AI for something or not?
- 22:42Speaker 1: I default too a little herostic I of do loa
- 22:45Speaker 1: for deelty first, So the loaf of Deelty I always
- 22:48Speaker 1: use as a baseline of going before I ask anyone
- 22:50Speaker 1: else for input, what do I think directionally? The answer
- 22:53Speaker 1: might smell, look and feel like right very loose guardrails.
- 22:56Speaker 1: I then want to very quickly go into sparring. Now
- 22:58Speaker 1: I sometimes do that with humans because it's just I
- 23:00Speaker 1: want a diversity of opinion. Sometimes I'll do it with AI.
- 23:03Speaker 1: I want a quick output. It's like we know each
- 23:05Speaker 1: other quite well. Now I've trained my models quite well,
- 23:08Speaker 1: so you can get me to an answer quite quickly.
- 23:09Speaker 1: And now that's lovely, of course I like it. I've
- 23:12Speaker 1: trained the model, right, So it's very self satisfying. But
- 23:14Speaker 1: that is a speed of execution on something that's often
- 23:18Speaker 1: relatively low risk. Right, It's been done numerous times, and
- 23:21Speaker 1: so it is semi formulaic. Where I get more human
- 23:25Speaker 1: in the loop. Is this is a net new thing.
- 23:27Speaker 1: We've not done this before, So it's when do we
- 23:30Speaker 1: use human and then how do we use so in
- 23:31Speaker 1: the example this morning, right, me and a colleague on
- 23:34Speaker 1: a video conference, just chatting through it on a live
- 23:36Speaker 1: screen together, is it this? Yes? Now more that like
- 23:39Speaker 1: and just getting the flavor I directionally. Once we had
- 23:41Speaker 1: the flavor, he's going to have done a huge amount
- 23:43Speaker 1: of research super fast that is validated and questioning that
- 23:47Speaker 1: gives us more assumptions. And we did another round, right,
- 23:49Speaker 1: So it's human AI to accelerate human again, AI to accelerate,
- 23:53Speaker 1: and we have a meeting this afternoon with humans right
- 23:56Speaker 1: to finalize that content. The final refinement is a human
- 24:00Speaker 1: piece of content that we've used AI to accelerate. In
- 24:04Speaker 1: my fact, I know you do a lot of public
- 24:05Speaker 1: speaking as well. I mean one of my biggest ones
- 24:07Speaker 1: is in public speaking. So taking a briefing call being
- 24:11Speaker 1: very articulate with the action is this. The focal points
- 24:14Speaker 1: are that it is not this, this is the audience, right,
- 24:16Speaker 1: And it's not just getting that as a transcript. It's
- 24:19Speaker 1: using that to go into any of the other ms.
- 24:21Speaker 1: Giving me a first round of slides, give me a
- 24:22Speaker 1: first front of this, give me a brief, give me
- 24:24Speaker 1: an abstract right so I can use it throughout that
- 24:26Speaker 1: entire loop. And it just means I'm putting less pressure
- 24:30Speaker 1: on my notebook or my memory to go did they
- 24:33Speaker 1: want leadership or something else like? And it just reduces
- 24:36Speaker 1: my error rate. Again, that's quite a manual thing. The
- 24:39Speaker 1: biggest fun I'm having right now, and this is like
- 24:41Speaker 1: completely basic, but I just I geek out on basic
- 24:44Speaker 1: stuff sometimes. I had a colleague ping me yesterday. The
- 24:47Speaker 1: booked a meeting in a calendar. It had the topic
- 24:50Speaker 1: that the client name the piece of work. I knew
- 24:52Speaker 1: they were very busy before the meeting, so they didn't
- 24:54Speaker 1: have time do some good information in the meeting notes.
- 24:57Speaker 1: We've got an ask B luminous thing in claud so
- 25:00Speaker 1: the companies could right after it and I just what
- 25:02Speaker 1: context can you give me on this meeting before the
- 25:05Speaker 1: meeting boom? Right, So we started the meeting where they're like, hey,
- 25:09Speaker 1: I don't know what you know about this, and I'm like,
- 25:11Speaker 1: here's the three things I know. Here's what I know
- 25:13Speaker 1: about history, Here's what I know about weere're at right now.
- 25:15Speaker 1: It meant that the half an hour we had together
- 25:18Speaker 1: down that was focused. It wasn't giving each of the
- 25:21Speaker 1: context we started with. The context probably took us three minutes, right,
- 25:25Speaker 1: and then we had twenty seven minutes together to go
- 25:27Speaker 1: what are the decisions we need to make? How do
- 25:29Speaker 1: we action this or the relationships we need to build?
- 25:31Speaker 1: It is on the stuff that was really human and
- 25:33Speaker 1: so again using AI to go instead of waiting for
- 25:37Speaker 1: that meeting, get the context and do this through hand
- 25:38Speaker 1: to hand combat. How can I use the fact that
- 25:41Speaker 1: we share a lot of our information internally to get
- 25:43Speaker 1: that context early to make the most use of our
- 25:46Speaker 1: human time together. And that's paying a huge dividend in
- 25:50Speaker 1: how we work across our teams.
- 25:51Speaker 2: With that example there, I love that. So what are
- 25:53Speaker 2: the inputs that the AI is drawing from to get
- 25:57Speaker 2: that rich context.
- 25:59Speaker 1: So we've got a very clear structure of how we
- 26:01Speaker 1: store all of our information. We have naming guidelines of
- 26:03Speaker 1: how we store our client work right and all that that,
- 26:06Speaker 1: like permission guidelines are all in there. And so what
- 26:09Speaker 1: it means is that is that as we store and
- 26:11Speaker 1: save it, if we're responsible and we make sure we
- 26:14Speaker 1: use Gemini to record our meeting winnis we use this
- 26:16Speaker 1: for that and all that is stored in the right way,
- 26:19Speaker 1: and we just give claud access to that, we can
- 26:22Speaker 1: start to run those queries, right, we can get that
- 26:23Speaker 1: level of refinement. The first time we run it absolutely
- 26:26Speaker 1: awful because we all store things with different names, different
- 26:29Speaker 1: name and conventions in different places, and it was it
- 26:31Speaker 1: wasn't hallucinating, it was bringing back our own depravity. Right.
- 26:35Speaker 1: It's playing back to us our terrible way of storing
- 26:38Speaker 1: data and information, which is simple act of cleaning that
- 26:41Speaker 1: up has got us a way higher for deuelty answer.
- 26:43Speaker 1: And now again I don't always fully trust the answer.
- 26:45Speaker 1: It gets me seventy eight percent of the way there,
- 26:48Speaker 1: but it's a great start point to go Okay, where's
- 26:50Speaker 1: this going. I just want to give you one more example.
- 26:53Speaker 1: There's a little bit left field. I think you know
- 26:55Speaker 1: Barrier Riley Barry is a friend of mine, great friend author.
- 26:59Speaker 1: He took me to an excise years ago called board
- 27:01Speaker 1: of directors because he was a work for himself. And
- 27:03Speaker 1: he's like, I haven't got a board of directors. So
- 27:05Speaker 1: once a year he made up a board of directors
- 27:07Speaker 1: and he wrote them a letter with his strategy. And
- 27:10Speaker 1: so I created my own board of directors in claude,
- 27:13Speaker 1: and I use this for when I'm going down the
- 27:15Speaker 1: dom price rabbit hole, officially trademarked where I'm looking for
- 27:18Speaker 1: lots of validation and I'm like, ooh, validation's good for
- 27:22Speaker 1: the ego, it's terrible for the output. So I'm like,
- 27:24Speaker 1: board of directors, critique me, challenge me. So I have
- 27:27Speaker 1: selected personas. I've created personas based off real life people,
- 27:31Speaker 1: but hybrids of people. One of my personas could be
- 27:34Speaker 1: half a Mantha and half Brene Brown right to fuse
- 27:36Speaker 1: them together. You're one human being, and you're called Amantha
- 27:39Speaker 1: Brown from now on, and you are on my board
- 27:42Speaker 1: of directors, and your job is to critique me on
- 27:44Speaker 1: these things because this is what I respect you for.
- 27:46Speaker 1: And so when I'm coming up with net new ideas
- 27:48Speaker 1: or challenges that really helped me find blind spots quicker
- 27:52Speaker 1: than I can anywhere else. Again, I don't think that's
- 27:55Speaker 1: a technology thing. I think that's a willingness to do
- 27:58Speaker 1: it thing that technology. And if I wasn't going to
- 28:01Speaker 1: do it, the technology is not going to do it
- 28:02Speaker 1: for me. It's not going to yell, dumb, You've got biases, right,
- 28:06Speaker 1: None of the tools are designed to do that. However,
- 28:09Speaker 1: if you acknowledge that you have, you can really ease
- 28:11Speaker 1: them to challenge those biases.
- 28:13Speaker 3: Okay, so many thoughts on all of that.
- 28:15Speaker 2: Firstly, I don't know how your organization is getting you
- 28:18Speaker 2: to stick to file naming conventions that I'd like. I'd
- 28:21Speaker 2: love to know how how you are sticking to that
- 28:23Speaker 2: dumb knowing what I know about you, Yeah, I.
- 28:27Speaker 1: Have people around me to help. I'm awful, and I
- 28:29Speaker 1: fully acknowledge. I'mlike, this is not more full.
- 28:31Speaker 2: But I think what that brings up is that I
- 28:33Speaker 2: find there's this thing that people who have not sort
- 28:37Speaker 2: of built the AI fluency think that AI is a
- 28:40Speaker 2: little bit magic and it can turn really crappy data
- 28:43Speaker 2: into really high quality data, and it can find outputs
- 28:47Speaker 2: or like you know, draw things to bring into a
- 28:50Speaker 2: chat that just any human would have no idea what
- 28:53Speaker 2: they're about because of poor naming conventions or you know,
- 28:56Speaker 2: anything like that, and it's like, a, I can't do that.
- 28:58Speaker 2: So I just think that's such an important point. I
- 29:00Speaker 2: love the Board of Directors example, we do something very
- 29:04Speaker 2: similar and inventing. We call it building your advisory board
- 29:07Speaker 2: with different personas.
- 29:09Speaker 3: And then sometimes what we do when.
- 29:11Speaker 2: We're working, particularly with executive leadership teams, is we've got
- 29:14Speaker 2: a template that is essentially like the system instructions are
- 29:17Speaker 2: a system prompt to put into an agent or a GPT,
- 29:21Speaker 2: and it's a template that people can fill out on
- 29:23Speaker 2: their actual Board of directors so you know them better
- 29:27Speaker 2: than us, and so you know, executive teams have to
- 29:29Speaker 2: present to the board many times a year in different
- 29:32Speaker 2: way shape or forms, or prepare papers, and we actually
- 29:35Speaker 2: help them build an AI version of their actual irl
- 29:39Speaker 2: board of directors so valuable.
- 29:41Speaker 1: We've done exec board, the footboard of directors and management team,
- 29:46Speaker 1: and one layer down they'll like it equipment like a
- 29:48Speaker 1: disk type profile, personality test. They'd or done some of
- 29:51Speaker 1: the ones. So we just lead at the min and
- 29:52Speaker 1: where oh dom it. If you're working with Amantha a
- 29:55Speaker 1: month is more detail oriented like this, like brevity bred on,
- 29:57Speaker 1: you're more of the basa. Actually think about her style
- 29:59Speaker 1: and and therefore, if you're presenting to a Mantha, here's
- 30:01Speaker 1: the things to think about. We just built that into
- 30:04Speaker 1: a It was all existing data, so it wasn't hard
- 30:06Speaker 1: to build that in. But again, just that insight for
- 30:08Speaker 1: them was like, oh, I now know how to meet
- 30:10Speaker 1: you halfway. It's not I'm going to just my style
- 30:13Speaker 1: and I'm an actor, it's how do I meet you
- 30:15Speaker 1: where you are and communicating the way that you can
- 30:17Speaker 1: consume because the whole point in me telling this story
- 30:19Speaker 1: is to access you in some way. It's not just
- 30:21Speaker 1: a broadcast and say it out loud. It's for you
- 30:23Speaker 1: to consume it, and so it does build a little
- 30:26Speaker 1: bit that self awareness.
- 30:27Speaker 2: I would love to know, don you're someone that creates
- 30:30Speaker 2: a lot of content puts a lot of content out
- 30:32Speaker 2: into the world, whether that be from LinkedIn to keynote,
- 30:37Speaker 2: speaking and everything in between. What are your rules of
- 30:42Speaker 2: thumb or even workflows in terms of you working with
- 30:46Speaker 2: AI to create that content.
- 30:48Speaker 1: I will use it to create visuals because it doesn't
- 30:51Speaker 1: matter whether it's canver, Google slides, PowerPoint. Okay, now I
- 30:55Speaker 1: am awful at creating visuals. I can have a wonderful
- 30:57Speaker 1: image in my head, but the USB stick does not
- 30:59Speaker 1: fit into my and it does not produce a one
- 31:01Speaker 1: ful image when I try and create it. I know
- 31:02Speaker 1: my limitations, but I can use I've trained enough of
- 31:05Speaker 1: the models now on existing content, brand guidlines, et cetera.
- 31:09Speaker 1: For very simple visuals. I'll use it for that. I
- 31:11Speaker 1: use it to stress test, where I'll put content that
- 31:14Speaker 1: I've written through and go again, rip this apart, tell
- 31:17Speaker 1: me where this is good? Where's it got to the bose?
- 31:20Speaker 1: Whether I not made a point. I did one recently.
- 31:21Speaker 1: I was writing for a large newspaper and ophead and
- 31:24Speaker 1: I put it through and it's like, you're missing a
- 31:26Speaker 1: problem statement, You've gone straight to a solution. I was like,
- 31:29Speaker 1: have ye fair good points? I added that line in
- 31:32Speaker 1: as a human because for me, there's high fidelty interactions
- 31:37Speaker 1: where it's more important. I don't mind showing my work
- 31:40Speaker 1: because it's showing my work that actually gets you to
- 31:42Speaker 1: understand the method in the madness. And I think, what
- 31:44Speaker 1: we're missing And you kind of made this point before
- 31:46Speaker 1: if we race to the output, but we don't know
- 31:49Speaker 1: the messiness of how we got there, did we really
- 31:51Speaker 1: learn it? And I'm old enough to remember that you
- 31:53Speaker 1: used to get marks for showing your working because that
- 31:57Speaker 1: was the logic, right, That the whole point that the
- 32:00Speaker 1: marking your mass paper ons not did you understand the logic?
- 32:02Speaker 1: Not did you get the answer right? You've got a
- 32:04Speaker 1: guest and got it right, but did you understand the logic?
- 32:06Speaker 1: And so often when I'm trying to work through a
- 32:08Speaker 1: thought process, I'll use it there. I use it very
- 32:11Speaker 1: little for my creative writing, and that's just a personal preference.
- 32:15Speaker 1: I use it to prompt me or to critique me,
- 32:18Speaker 1: but not to write on my behalf. I just find
- 32:20Speaker 1: that for me is such a personal thing where I'm like,
- 32:23Speaker 1: if I've got the privilege, and I acknowledge a large
- 32:25Speaker 1: amounts of my job are a privilege to either stand
- 32:28Speaker 1: on stage and present to an audience, or write an
- 32:30Speaker 1: article that goes in a magazine that people are going
- 32:32Speaker 1: to take their time out to read. That should have
- 32:34Speaker 1: come from my heart, my soul, my institutional knowledge, my
- 32:37Speaker 1: lived experience, my empathy, my scars, my smiles, like it
- 32:40Speaker 1: should come from that. And yes, I might be able
- 32:42Speaker 1: to train claud or anyone to do a version of that,
- 32:46Speaker 1: But I don't want a digital twin writing my thought
- 32:49Speaker 1: provoking articles for me, because they're mine. It's my IP,
- 32:52Speaker 1: and so I keep those very close. I do use,
- 32:54Speaker 1: like I say, AI to critique them, but not to
- 32:56Speaker 1: write them on my behalf. And maybe that will change.
- 32:58Speaker 1: Maybe that's my limitation and I'll loosen that over time.
- 33:01Speaker 1: But I've always got this little sort of axioum in
- 33:03Speaker 1: my head, when is this a human conversation. One of
- 33:06Speaker 1: the things I love in life outside of work is
- 33:09Speaker 1: hosting dinner parties. I have a four meter dining table
- 33:12Speaker 1: downstairs with enough room for twelve people, right. I have
- 33:15Speaker 1: a massive kitchen and I love cooking and hosting, and
- 33:17Speaker 1: whenever I do that, there's this little thing that goes off,
- 33:20Speaker 1: which is there's no technology here like the fidelity of
- 33:23Speaker 1: that experience. The things I love about it the incidental conversation,
- 33:27Speaker 1: the interruption, is the nuance, the tone, the different centers
- 33:30Speaker 1: you get for seeing people's moods throughout the night right,
- 33:33Speaker 1: all the laughter, the tears, the happiness. All that is
- 33:36Speaker 1: this rich tapestry of life. Please don't AI that away
- 33:39Speaker 1: from me, right, I want more of that, And I
- 33:41Speaker 1: think sometimes we get carried away. Hey, I can do everything.
- 33:43Speaker 1: I'm like sod Off, I don't want you to do everything.
- 33:45Speaker 1: I want to do more human things. And for me
- 33:48Speaker 1: for the time being, I acknowledge that whatever it is
- 33:51Speaker 1: my positional power or my history or states whatever that
- 33:54Speaker 1: enables me to write and for people to read, that
- 33:56Speaker 1: that is such a privilege. That's a human thing. For
- 33:58Speaker 1: now that I use AI timplified, but I will not replace.
- 34:02Speaker 3: I am also a fan of dinner parties dom.
- 34:04Speaker 2: I had this interesting experience at the beginning of the year. I,
- 34:08Speaker 2: my husband and I were invited to go to a
- 34:11Speaker 2: Shabbat dinner. And I'm not particularly religious, but it was
- 34:15Speaker 2: one of the best nights of my year. So at
- 34:17Speaker 2: a Shabbat dinner, there's no technology like you're literally it's
- 34:22Speaker 2: just human to human and we were so inspired by
- 34:25Speaker 2: this that we're like, well, let's let's do our own.
- 34:27Speaker 2: And we hosted it a couple of weeks ago, and
- 34:31Speaker 2: really the one rule, aside from one of my beautiful
- 34:34Speaker 2: friends cooked some horlour bread, which is tradition delicious. It's
- 34:37Speaker 2: kind of like sugar plus bread if you've never tasted it,
- 34:40Speaker 2: and pretty much like that, the only rule or ritual
- 34:44Speaker 2: we kind of stuck to was just no mobile phones.
- 34:47Speaker 2: And it was so interesting what that did. I remember
- 34:50Speaker 2: someone was trying to remember something. It might have been
- 34:53Speaker 2: a song or some cultural moment, and they couldn't quite
- 34:55Speaker 2: remember it. They wanted to show it, and instead we
- 34:58Speaker 2: just kind of had to problem solve our way through that.
- 35:00Speaker 3: As humans.
- 35:00Speaker 2: We couldn't just look it up on Google or perplexity
- 35:03Speaker 2: or chat JPT. And there's something really beautiful about that.
- 35:06Speaker 1: But you're also probably reconnecting the synapses in their brain, right,
- 35:08Speaker 1: You're activating parts of the brain that are meant to
- 35:11Speaker 1: be active. Like I talk about the value of like
- 35:13Speaker 1: personal experience and lived experience, So we've got to work
- 35:16Speaker 1: out what are the experiences that just make us feel
- 35:18Speaker 1: fulfilled but fill our cup. Let's just do more of
- 35:21Speaker 1: them right, And I think sometimes we over engineer it
- 35:24Speaker 1: that a I can do. I can do all my
- 35:25Speaker 1: shopping for me. Agentic commerce is gonna be worth X
- 35:28Speaker 1: billion dollars and we'll never have to go to the
- 35:29Speaker 1: shops again. I have twin boys at hand. I like
- 35:33Speaker 1: going to the shops right because I said we're out
- 35:36Speaker 1: of mills. She's like, we had a lot of love
- 35:37Speaker 1: to put it down and saying I'm going to the
- 35:38Speaker 1: shops and I wander around, and I sometimes I wonder
- 35:41Speaker 1: raniously and I pick ingredients and I get inspired and
- 35:44Speaker 1: I work out what I'm going to cook. And I
- 35:46Speaker 1: get by the kids when on unpack and they want
- 35:47Speaker 1: to do chop shop with me, Like that's part of
- 35:49Speaker 1: the experience. I don't need an agent to do that
- 35:52Speaker 1: for me. I don't need a fridge center that tells
- 35:54Speaker 1: me amount of garlic. I know I'm out of garlic.
- 35:56Speaker 1: I used it all yesterday. So I think we need
- 35:58Speaker 1: to be really careful about if we use AI and
- 36:00Speaker 1: technology to automate away everything carte blanche, and we don't
- 36:03Speaker 1: do the thinking about what are the human things and
- 36:06Speaker 1: things that add to our experience, the things that make
- 36:08Speaker 1: us more fulfilled, or in your case, that the societal
- 36:11Speaker 1: connection that you got from both of those dinners, from
- 36:13Speaker 1: the one you were telling the one you hosted. Oh ah,
- 36:16Speaker 1: that a way for me, right, So it's not that
- 36:17Speaker 1: it's bad, it's let's just use it where it's it's
- 36:20Speaker 1: fit for purpose, and then how do we invest our
- 36:22Speaker 1: time in doing the things that genuinely make us feel fulfilled.
- 36:25Speaker 1: I don't know if it's just because I'm in my
- 36:26Speaker 1: leg forties and I've got kids, I've suddenly tapped into this,
- 36:28Speaker 1: but I'm like, there must be more to life than
- 36:30Speaker 1: chasing AI. I don't need an AI robot in the
- 36:33Speaker 1: house doing all the things for me.
- 36:35Speaker 2: So in closing, I mean, I could just talk to
- 36:37Speaker 2: you for hours on this, but if we have limited time,
- 36:40Speaker 2: what does healthy AI use look like for us individually
- 36:45Speaker 2: and also in organizations?
- 36:47Speaker 1: My view is to not think about AI, is to
- 36:49Speaker 1: think about what is the business problem you're solving. You
- 36:51Speaker 1: kind of mentioned it before, like it's the what so
- 36:54Speaker 1: what now?
- 36:55Speaker 4: What? Like?
- 36:55Speaker 1: What are you solving? So? What what are you solving
- 36:57Speaker 1: it now? What what are we going to do? AI
- 36:59Speaker 1: should come into the now what It's always in the
- 37:02Speaker 1: third step, but it shouldn't be in the first two,
- 37:04Speaker 1: right that don't do this for the sake of AI.
- 37:07Speaker 1: Do it because there is a tangible benefit we see
- 37:09Speaker 1: in that new customer opportunity. We can connect with our people.
- 37:12Speaker 1: We have more empathy for them. They feel more connected
- 37:14Speaker 1: that they feel more heard and understood, and they're more
- 37:16Speaker 1: engaging in their happier people So they're spending the same
- 37:18Speaker 1: amount of time in work, but they're more fulfilled. That's
- 37:20Speaker 1: a win. That's not a bad thing. Right, You're probably
- 37:23Speaker 1: going to get discretionary effort from me in that world
- 37:25Speaker 1: and not have to pay me anymore. Like take that
- 37:27Speaker 1: right now. Maybe it's a cost saving, Maybe it's a
- 37:29Speaker 1: market thing. I don't know, whatever it is, Like, that's
- 37:31Speaker 1: the business opportunity and the step to get there is
- 37:35Speaker 1: always a component of your workplace, your physical environment, how
- 37:39Speaker 1: you work, your human operating system of your people and
- 37:41Speaker 1: your teams, and your technology. Right, but we're so over
- 37:45Speaker 1: indexed on the tech piece, we're forgetting about the other two,
- 37:48Speaker 1: the physical environment we work in. As me and you
- 37:50Speaker 1: have studied a lot through what happened with work from home,
- 37:53Speaker 1: through the pandemic and our hybrid and that argument is
- 37:55Speaker 1: still going on. But also like our human operating system
- 37:58Speaker 1: most of the businesses I work with in the ASEX,
- 38:01Speaker 1: they're human operating systems Windows ninety five. So you might
- 38:03Speaker 1: have clawed five dot nine. Well done right, you're using
- 38:07Speaker 1: Ferrari style horsepowering your technology. But the way your humans
- 38:10Speaker 1: and teams work and meet and make decisions, the way
- 38:13Speaker 1: you think about purpose and values, connecting strategy to goals
- 38:16Speaker 1: to execution, all those things are Windows ninety five and
- 38:18Speaker 1: you've got the blue screen to death. So focus on
- 38:21Speaker 1: that bit upgrade that make it fit for purpose and
- 38:24Speaker 1: then work out how you augment and layer in AI
- 38:27Speaker 1: as a component part, not as the reason to do anything.
- 38:30Speaker 1: So the example I give you is a company I'm
- 38:32Speaker 1: working with recently like we're going to do an AI hackathon.
- 38:34Speaker 1: I was like why, I like, without being an idiot,
- 38:36Speaker 1: I don't understand why do a business hackathon to improve
- 38:39Speaker 1: your business and use AI as the amplifier, but don't
- 38:42Speaker 1: do it for the sake of AI because you come
- 38:44Speaker 1: up with a whole lot of tools and toys and
- 38:47Speaker 1: paraphernaviia that no one needs or wants to use.
- 38:49Speaker 3: Dom thank you so much for coming back on how
- 38:51Speaker 3: I work.
- 38:52Speaker 2: Every time you're on I learned something new and you
- 38:55Speaker 2: just have such a way of making things that can
- 38:58Speaker 2: feel really complex and over whelming feel just so simple
- 39:02Speaker 2: and easy to understand.
- 39:03Speaker 3: So thank you for your time, don thanks.
- 39:05Speaker 1: For having me, and also Manthea, thanks for all the
- 39:06Speaker 1: work you do, because this sharing you do on the
- 39:08Speaker 1: practical stuff you see inspire so many people, Like there's
- 39:12Speaker 1: so many people hungry out there for better ways of
- 39:13Speaker 1: doing stuff and you always share that. So yeah, thank
- 39:15Speaker 1: you return from all your audience.
- 39:17Speaker 2: So my biggest takeaway from this episode is that companies
- 39:22Speaker 2: need to fix the human operating system before they layer
- 39:26Speaker 2: in the technology. So next time someone in your organization
- 39:31Speaker 2: proposes a new AI initiative, ask what business problem it's
- 39:36Speaker 2: actually solving, and if the answer starts with because everyone
- 39:40Speaker 2: else is doing AI, that is your signal too.
- 39:43Speaker 4: Pause.
- 39:44Speaker 2: Now, if you want to keep going down this AI route,
- 39:48Speaker 2: then you will definitely love. The episodes that we release
- 39:52Speaker 2: on the How I Work feed every Monday are how
- 39:55Speaker 2: IAI episodes which contain very practical tips on how to
- 39:59Speaker 2: get or out of AI. And you might also enjoy
- 40:03Speaker 2: my recent chat with Professor Scott Anthony on how AI
- 40:09Speaker 2: has changed how he approaches problem solving and how it's
- 40:13Speaker 2: changed several of his workflows. You can find a link
- 40:16Speaker 2: to that in the show notes. And if you enjoy
- 40:19Speaker 2: this episode, please hit subscribe or follow so you don't
- 40:22Speaker 2: miss what's coming out.
- 40:23Speaker 3: Next on How I Work. If you like.
- 40:25Speaker 4: Today's show, make sure you hit follow on your podcast
- 40:29Speaker 4: app to be alerted when new episodes drop. How I
- 40:32Speaker 4: Work was recorded on the traditional land of the Warrangery People,
- 40:35Speaker 4: part of the Cooler Nation