1
00:00:00,120 --> 00:00:03,160
Architects shouldn't try to be 
the smartest people, but they 

2
00:00:03,160 --> 00:00:04,720
should make everybody else 
smarter. 

3
00:00:04,720 --> 00:00:07,400
That's an amplifier. 
You don't want to be a kind of 

4
00:00:07,640 --> 00:00:10,520
Oracle where people come with 
the questions and look for magic

5
00:00:10,520 --> 00:00:12,000
answers. 
The good architects are usually 

6
00:00:12,000 --> 00:00:14,600
the ones where magically 
everything goes well and nobody 

7
00:00:14,600 --> 00:00:17,840
knows exactly why. 
Too many architects try to find 

8
00:00:17,920 --> 00:00:20,120
answers when they don't have a 
question. 

9
00:00:20,120 --> 00:00:22,400
But here's the thing that 
answers all possible questions. 

10
00:00:22,440 --> 00:00:25,960
That would be nice. 
That doesn't work anymore. 

11
00:00:26,080 --> 00:00:29,960
We've gotten so in love with 
complexity that if we actually 

12
00:00:30,040 --> 00:00:33,640
cut through the complexity, we 
sometimes doubt ourselves. 

13
00:00:33,960 --> 00:00:35,360
Don't stumble on the finish 
line. 

14
00:00:35,520 --> 00:00:38,880
If you made sense out of this 
and suddenly it seems obvious, 

15
00:00:39,040 --> 00:00:42,800
you've done a fantastic job. 
Today I'm joined by Gregor Hope,

16
00:00:42,920 --> 00:00:45,360
retired from Big Tech, used to 
work in Silicon Valley in 

17
00:00:45,360 --> 00:00:48,440
enterprise and even have vendors
which gives them a unique 

18
00:00:48,440 --> 00:00:50,200
perspective to talk about 
architecture. 

19
00:00:50,680 --> 00:00:53,560
We discuss how to go from 
software engineer to architect, 

20
00:00:53,840 --> 00:00:56,640
what great behaviour looks like 
and how you can spot bad 

21
00:00:56,640 --> 00:00:59,080
architects and what are some 
common traps. 

22
00:00:59,080 --> 00:01:02,400
You shouldn't fall into a 
software engineer, some of which

23
00:01:02,400 --> 00:01:05,239
you won't expect. 
Fantastic takeaways in this one,

24
00:01:05,239 --> 00:01:11,080
so enjoy. 
What have you seen in how 

25
00:01:11,080 --> 00:01:13,880
architects execute that makes 
them really good architects for 

26
00:01:13,880 --> 00:01:15,640
the people that actually go on 
that career path? 

27
00:01:15,920 --> 00:01:18,400
Yeah, I always say, I mean, 
first comment would be it's not 

28
00:01:18,560 --> 00:01:22,480
a sort of binary choice or I 
don't think that being an 

29
00:01:22,480 --> 00:01:25,880
architect is determined by 
what's on your business card. 

30
00:01:26,200 --> 00:01:29,480
So I think you can be a 
technical product person with a 

31
00:01:29,560 --> 00:01:33,320
very architecture mindset, right
for me. 

32
00:01:33,640 --> 00:01:35,560
Well, let me start the other way
around. 

33
00:01:35,560 --> 00:01:38,280
The the bad architects are 
easier to spot. 

34
00:01:38,280 --> 00:01:40,480
The good architects are usually 
the ones where magically 

35
00:01:40,480 --> 00:01:43,920
everything goes well and nobody 
knows exactly why the bad 

36
00:01:43,920 --> 00:01:46,920
architects are easier to spot. 
And it's like people who spew 

37
00:01:46,920 --> 00:01:49,920
out a lot of buzzwords, you 
know, it's like, oh, everything 

38
00:01:49,920 --> 00:01:51,840
must be cloud native or loosely 
coupled. 

39
00:01:51,840 --> 00:01:54,360
Now say I don't need an 
architect to tell me that. 

40
00:01:54,360 --> 00:01:56,000
I can just put a poster on the 
wall. 

41
00:01:56,120 --> 00:01:58,320
And then why, why do I need 
this? 

42
00:01:58,520 --> 00:02:01,560
This person or people who 
believe they should have all the

43
00:02:01,560 --> 00:02:04,080
decision power. 
It's like, oh developers, I'm 

44
00:02:04,080 --> 00:02:06,360
here to guide you. 
You know, should make 3 

45
00:02:06,360 --> 00:02:08,720
components, right? 
No more, no less, right? 

46
00:02:08,720 --> 00:02:12,280
Those I find the not so stellar 
architects. 

47
00:02:13,360 --> 00:02:16,600
The mantra that we have in the 
workshop and that I that I very 

48
00:02:16,600 --> 00:02:21,120
much follow is that architects 
shouldn't try to be the smartest

49
00:02:21,120 --> 00:02:23,760
people, but they should make 
everybody else smarter. 

50
00:02:24,000 --> 00:02:27,160
All right, that's an amplifier. 
You don't want to be a kind of 

51
00:02:27,320 --> 00:02:30,200
Oracle when people come with the
questions and look for magic 

52
00:02:30,200 --> 00:02:32,960
answers, Oh, what should we do? 
People come to me, right? 

53
00:02:32,960 --> 00:02:36,160
And I'm like, well, that's kind 
of charming, but how, how do I 

54
00:02:36,160 --> 00:02:38,320
know it's your project? 
It's your application, right? 

55
00:02:38,320 --> 00:02:40,800
How can I make all the decisions
for you? 

56
00:02:41,120 --> 00:02:44,320
But what I can do is help you 
make a, a better decision. 

57
00:02:44,680 --> 00:02:48,000
And I think to me, that is one 
of the characteristics of a, of 

58
00:02:48,000 --> 00:02:51,000
a great architect. 
You can absorb the context, 

59
00:02:51,000 --> 00:02:54,800
things that people explain to 
you, but you can uncover blind 

60
00:02:54,800 --> 00:02:58,440
spots or maybe help people see 
different points of view, 

61
00:02:58,440 --> 00:03:02,640
different angles, distilled 
trade-offs that they might be 

62
00:03:02,640 --> 00:03:06,720
implicitly making, but they're 
not aware of those kind of 

63
00:03:06,720 --> 00:03:08,560
examples. 
I think that makes a great 

64
00:03:08,560 --> 00:03:12,080
architect and it makes a nice 
architect if you wish, because 

65
00:03:12,320 --> 00:03:15,400
you're not trying to be the the 
ivory tower big boss, but he 

66
00:03:15,400 --> 00:03:17,920
actually an amplifier to the 
team. 

67
00:03:18,200 --> 00:03:21,400
Yeah, From my position, both in 
product and in software, I have 

68
00:03:21,400 --> 00:03:24,280
seen architects that are very 
much, OK, it's my way or the 

69
00:03:24,280 --> 00:03:26,920
highway and then they are a 
check stop that I need to 

70
00:03:26,920 --> 00:03:28,600
actually pass. 
So I have to convince an 

71
00:03:28,600 --> 00:03:31,680
architect and actually have a 
conversation with them. 

72
00:03:31,680 --> 00:03:33,880
And from their position, there's
a lot of power in that position.

73
00:03:33,880 --> 00:03:36,680
In some organizations, 
especially in more traditional 

74
00:03:36,800 --> 00:03:40,040
bigger enterprise organizations,
bank for example, especially 

75
00:03:40,040 --> 00:03:43,960
where risk is very much 
mitigated, people don't want to 

76
00:03:43,960 --> 00:03:45,840
afford to take risks. 
There's no need to afford to 

77
00:03:45,840 --> 00:03:48,120
take risks because everything 
will just be operational. 

78
00:03:48,440 --> 00:03:50,480
I feel like that's where you 
have architects, which are kind 

79
00:03:50,480 --> 00:03:52,960
of a stop gap between a lot of 
innovating and solutioning. 

80
00:03:53,320 --> 00:03:57,240
Yeah, I definitely see 
architects as folks who lower 

81
00:03:57,240 --> 00:04:00,040
risk. 
And I often tell people, you 

82
00:04:00,040 --> 00:04:02,560
know, we actually done this 
exercise where we pictures like,

83
00:04:02,560 --> 00:04:04,280
hey, what's your value 
proposition? 

84
00:04:04,560 --> 00:04:07,800
And there's an architect. 
Often I can say I actually lower

85
00:04:07,800 --> 00:04:10,200
your risk. 
You might be fine without me. 

86
00:04:10,280 --> 00:04:13,400
You can just cobble together 
some application and maybe it 

87
00:04:13,400 --> 00:04:14,840
works, maybe everything will be 
fine. 

88
00:04:14,840 --> 00:04:17,720
Maybe it scales, maybe it 
doesn't have security exploits, 

89
00:04:17,720 --> 00:04:19,399
right. 
But that would definitely be a 

90
00:04:19,399 --> 00:04:22,440
risky proposition. 
So I think in a, in a, in a good

91
00:04:22,440 --> 00:04:27,160
sense, architects actually 
anticipate risks and mitigate 

92
00:04:27,480 --> 00:04:30,000
risks. 
Like if scalability is a need 

93
00:04:30,000 --> 00:04:33,160
that you have, right, If I call 
this an option that you want to 

94
00:04:33,160 --> 00:04:36,080
have, then we can make the sure 
make we can make sure that the 

95
00:04:36,080 --> 00:04:38,320
system has those 
characteristics. 

96
00:04:38,520 --> 00:04:43,120
So it's really a risk management
function and we all know that we

97
00:04:43,120 --> 00:04:45,400
do risk. 
It has value, right? 

98
00:04:45,400 --> 00:04:47,760
That's money in the bank. 
I used to work for insurance. 

99
00:04:47,760 --> 00:04:49,720
So we know that risk equates the
cost. 

100
00:04:49,720 --> 00:04:51,520
So lower risk is equals lower 
cost. 

101
00:04:51,840 --> 00:04:53,960
But it's probably in the 
different way than many 

102
00:04:53,960 --> 00:04:56,600
traditional architects think. 
So, right. 

103
00:04:56,600 --> 00:05:00,520
So for example, lowering risk 
doesn't just mean working to 

104
00:05:00,520 --> 00:05:04,880
plan or doing everything upfront
than having sort of 1 golden 

105
00:05:04,880 --> 00:05:07,920
design or 1 golden reference 
architectures. 

106
00:05:08,480 --> 00:05:11,240
I think the banks sometimes 
conflate that. 

107
00:05:11,240 --> 00:05:15,520
Like basically they think if 
your plan is perfect, then the 

108
00:05:15,520 --> 00:05:20,320
risk is low, but they focus 
purely on execution risk. 

109
00:05:20,320 --> 00:05:22,880
Like, do we build what we said 
we're going to build? 

110
00:05:23,240 --> 00:05:25,200
But software has very different 
risk. 

111
00:05:25,200 --> 00:05:27,600
Like, will the users like it, 
right? 

112
00:05:27,600 --> 00:05:29,400
Will it actually do its business
function? 

113
00:05:29,400 --> 00:05:31,880
Does it actually move the 
needle, right? 

114
00:05:31,880 --> 00:05:33,640
Is it making revenue for us? 
Is it going to make the 

115
00:05:33,640 --> 00:05:36,040
customers happy? 
Does it grow market share, 

116
00:05:36,040 --> 00:05:38,000
right? 
And I think that's a whole 

117
00:05:38,000 --> 00:05:39,560
different, different set of 
risks. 

118
00:05:39,800 --> 00:05:42,240
And I see different 
organizations have very 

119
00:05:42,240 --> 00:05:45,960
different point of views on what
kind of risks they they focus 

120
00:05:45,960 --> 00:05:48,080
on. 
And that reflects on what the 

121
00:05:48,240 --> 00:05:51,480
how the architects behave. 
I'm wondering what your thought 

122
00:05:51,480 --> 00:05:54,040
then is next to risk about 
complexity. 

123
00:05:54,120 --> 00:05:57,080
I I recently had a conversation 
with a senior software engineer,

124
00:05:57,400 --> 00:05:58,640
Bassem. 
He's from GitHub. 

125
00:05:58,880 --> 00:06:02,120
He works on GitHub Actions, so 
massive scale, very cool what 

126
00:06:02,120 --> 00:06:04,320
they do. 
And his take, some people 

127
00:06:04,320 --> 00:06:06,080
actually thought it was 
controversial, he says. 

128
00:06:06,280 --> 00:06:09,960
I like in the way we design 
systems, especially at GitHub to

129
00:06:09,960 --> 00:06:12,480
keep things simple, because 
simple becomes complex when 

130
00:06:12,480 --> 00:06:15,480
you're talking about scale. 
And then especially in this 

131
00:06:15,480 --> 00:06:19,160
episode in the Commons, people 
that haven't operated at scale, 

132
00:06:19,320 --> 00:06:20,560
they don't necessarily 
understand it. 

133
00:06:20,560 --> 00:06:23,880
They think complexity is a big, 
big part of the job and you 

134
00:06:23,880 --> 00:06:26,560
should design for a rocket ship 
because you cannot afford 

135
00:06:26,560 --> 00:06:28,720
necessarily to migrate or to 
evolve software. 

136
00:06:29,040 --> 00:06:30,200
And I'm curious how you see 
that? 

137
00:06:30,400 --> 00:06:32,400
Well, I don't see it as 
controversial at all. 

138
00:06:32,440 --> 00:06:35,840
I would say simplicity is one of
the biggest strengths that a 

139
00:06:35,840 --> 00:06:39,160
good design can have, right? 
There's always, of course, 

140
00:06:39,160 --> 00:06:42,000
there's a, there's an optimum in
everything, right? 

141
00:06:42,000 --> 00:06:45,200
So, and this comes more of 
building platforms because I 

142
00:06:45,200 --> 00:06:49,160
wrote a whole book about 
platform strategy and many 

143
00:06:49,160 --> 00:06:51,480
domains have an inherent 
complexity. 

144
00:06:51,920 --> 00:06:54,920
Let's say you're building a 
cloud deployment like an 

145
00:06:54,920 --> 00:06:57,280
internal developer platform. 
Let's say you build distributed 

146
00:06:57,280 --> 00:06:59,360
systems. 
Distributed systems have an 

147
00:06:59,360 --> 00:07:02,480
inherent complexity. 
You need to deal with retries 

148
00:07:02,480 --> 00:07:07,160
and timeouts and item potency, 
back pressure, retry storms, 

149
00:07:07,160 --> 00:07:09,400
right? 
This is just stuff, this loss of

150
00:07:09,400 --> 00:07:11,520
physics. 
There's no way you can sort of 

151
00:07:11,520 --> 00:07:16,480
cheat your way out of that. 
So that complexity is inherent. 

152
00:07:17,480 --> 00:07:21,000
And we had a lot of debates when
I worked at AWS for this where 

153
00:07:21,280 --> 00:07:24,400
the product designers always 
said we need to make it simpler.

154
00:07:24,920 --> 00:07:29,320
And like, well, we can only make
it so simple because things have

155
00:07:29,320 --> 00:07:33,080
an inherent complexity because 
we build distributed serverless 

156
00:07:33,080 --> 00:07:35,960
systems. 
So that is the the one end of 

157
00:07:35,960 --> 00:07:37,680
the spectrum. 
To be careful, though, you 

158
00:07:37,680 --> 00:07:41,320
should make it, well, as simple 
as possible, but no simplest. 

159
00:07:41,320 --> 00:07:43,200
The smart. 
Yeah, it's sort of the smart 

160
00:07:43,200 --> 00:07:44,920
version of it. 
But I think it's good to 

161
00:07:44,920 --> 00:07:48,920
understand what is the inherent 
complexity, especially for 

162
00:07:48,920 --> 00:07:51,920
platform builders. 
And then my guideline would be, 

163
00:07:51,920 --> 00:07:56,520
don't try to make it simpler 
than that, but make it intuitive

164
00:07:56,800 --> 00:07:58,800
to deal with the inherent 
complexity. 

165
00:07:59,080 --> 00:08:02,840
Yeah, don't pretend the 
complexity doesn't exist, but 

166
00:08:02,840 --> 00:08:06,640
make it easier for people to 
deal with it, make it intuitive.

167
00:08:07,080 --> 00:08:10,880
And I think that is a good 
guideline, but you try to aim 

168
00:08:10,960 --> 00:08:14,600
towards the minimum, just don't 
overshoot. 

169
00:08:15,120 --> 00:08:19,400
I think too much complexity is 
one of the biggest problems we 

170
00:08:19,400 --> 00:08:24,960
invite into our technology lives
and technology has become more 

171
00:08:24,960 --> 00:08:27,480
complex. 
Anyhow, I always say writing a 

172
00:08:27,480 --> 00:08:30,880
monolith in Java and sticking 
that I want server was a lot 

173
00:08:30,880 --> 00:08:32,960
simpler than the software I 
today. 

174
00:08:33,679 --> 00:08:37,679
The software I today of course 
has many nice characteristics, 

175
00:08:38,000 --> 00:08:40,559
right? 
Is auto scaling, self healing, 

176
00:08:40,559 --> 00:08:43,840
integral created, distributed. 
It has many attributes that we 

177
00:08:43,840 --> 00:08:46,400
wish for, but it's not actually 
simpler, no. 

178
00:08:46,400 --> 00:08:51,520
So a good architect's job is 
really to conquer the complexity

179
00:08:51,520 --> 00:08:55,280
that is there, to break down the
complexity, to abstract away the

180
00:08:55,280 --> 00:08:59,760
complexity, because otherwise we
become the the limiting element.

181
00:09:00,080 --> 00:09:02,720
If things are too complex, 
people will make mistakes. 

182
00:09:03,080 --> 00:09:06,480
Cognitive load is an important 
term these days, right? 

183
00:09:06,480 --> 00:09:08,560
If it's too complex, the 
cognitive load will go up, 

184
00:09:08,800 --> 00:09:12,400
People will make mistakes, 
people will be slower or worst, 

185
00:09:12,640 --> 00:09:14,800
people will be hesitant to make 
a change. 

186
00:09:15,080 --> 00:09:18,560
If something is too complex, 
what will happen is you make a 

187
00:09:18,560 --> 00:09:20,840
change over here, something over
there will break. 

188
00:09:20,840 --> 00:09:23,000
Nobody knows why. 
You don't want to touch it 

189
00:09:23,000 --> 00:09:24,360
exactly. 
And then you don't want to touch

190
00:09:24,360 --> 00:09:25,920
it. 
And that's the worst case in a, 

191
00:09:26,240 --> 00:09:29,360
in a ever changing world, having
a piece of software that you're 

192
00:09:29,360 --> 00:09:31,520
afraid to touch, well, that's 
called legacy. 

193
00:09:31,760 --> 00:09:34,920
We have plenty of that. 
So we don't really want to build

194
00:09:34,920 --> 00:09:37,400
more of that. 
I'm wondering if you've ever 

195
00:09:37,400 --> 00:09:41,000
been in a position where you 
were of a certain opinion and 

196
00:09:41,000 --> 00:09:43,520
architects make other people 
smarter, but people don't 

197
00:09:43,520 --> 00:09:44,880
necessarily have to agree with 
you. 

198
00:09:44,880 --> 00:09:48,240
So then you are kind of this 
person that people either listen

199
00:09:48,240 --> 00:09:50,280
to or they disagree with, but 
you still have to work. 

200
00:09:50,640 --> 00:09:53,480
I think a big part of it is 
working with either people or 

201
00:09:53,480 --> 00:09:57,000
bigger aspects of teams. 
Have you ever found yourself in 

202
00:09:57,000 --> 00:09:59,560
a position where people disagree
with what you say and then how 

203
00:09:59,560 --> 00:10:01,520
do you help them? 
Of course not, because I'm 

204
00:10:01,520 --> 00:10:03,880
always right. 
Yeah, that makes a lot of things

205
00:10:03,960 --> 00:10:05,160
easy. 
That's why I became an 

206
00:10:05,160 --> 00:10:07,800
architect, right? 
I'm always right now, of course,

207
00:10:07,800 --> 00:10:09,840
right. 
And actually, the workshop today

208
00:10:09,840 --> 00:10:13,320
even went into some of this, 
right, where we're saying is 

209
00:10:13,680 --> 00:10:18,600
rather than just giving an 
answer, you frame the solution 

210
00:10:18,600 --> 00:10:21,440
space, you try to expand the 
solution space. 

211
00:10:21,680 --> 00:10:25,400
I have so many stories where 
people talk left versus right. 

212
00:10:25,920 --> 00:10:28,200
We have a picture that we use. 
You might remember it's the 

213
00:10:28,200 --> 00:10:32,040
circle versus the triangle. 
People look at a cylinder, but 

214
00:10:32,040 --> 00:10:35,040
one is like sort of round, long 
cylinder and one person looks 

215
00:10:35,040 --> 00:10:37,760
from the front and basically 
says The thing is a circle. 

216
00:10:37,960 --> 00:10:39,920
And the other person looks from 
the side and says, well, I have 

217
00:10:39,920 --> 00:10:43,120
no idea what you're looking at, 
but that is a rectangle, right? 

218
00:10:43,120 --> 00:10:45,800
And they will never come to any 
conclusion. 

219
00:10:46,000 --> 00:10:48,120
And that's where people sort of 
start fighting. 

220
00:10:48,120 --> 00:10:50,920
It's like, oh, Jason is better 
than YAML and Copenandes is 

221
00:10:50,920 --> 00:10:53,520
better than anything else. 
And it must be in a container 

222
00:10:53,520 --> 00:10:56,560
versus that, right? 
People associate themselves with

223
00:10:56,560 --> 00:11:00,400
a certain technology choice 
without being clear on, well, 

224
00:11:00,480 --> 00:11:02,760
what's really the solution 
space? 

225
00:11:02,800 --> 00:11:06,240
Like what are the dimensions? 
What is sort of the map that we 

226
00:11:06,240 --> 00:11:09,320
are discussing? 
We might be discussing different

227
00:11:09,320 --> 00:11:12,960
paths, but at least in our mind,
we should have the same map. 

228
00:11:13,120 --> 00:11:16,360
Yeah, because I see this a lot. 
And this can have two flavors. 

229
00:11:16,360 --> 00:11:20,360
Either people never agree or 
they agree but they have a 

230
00:11:20,360 --> 00:11:23,400
different map in the head so 
they think they agreed but they 

231
00:11:23,400 --> 00:11:25,640
actually walk out thinking very 
different things. 

232
00:11:25,640 --> 00:11:30,360
So what I advise to people is 
have a common framing so at 

233
00:11:30,360 --> 00:11:33,600
least you understand the 
solution space and that 

234
00:11:33,600 --> 00:11:38,040
shouldn't be up for debate. 
And then based on that framing, 

235
00:11:38,280 --> 00:11:41,160
then you can have the 
discussions. 

236
00:11:41,360 --> 00:11:44,960
So let's take simple example. 
People debate micro services. 

237
00:11:45,760 --> 00:11:49,520
The first thing I will do is 
break this down into, well, what

238
00:11:49,520 --> 00:11:53,080
does that actually mean? 
That means modularity, but 

239
00:11:53,080 --> 00:11:56,560
there's two kinds of modularity.
There's design time modularity 

240
00:11:56,560 --> 00:12:00,840
and there's runtime modularity. 
So you really have 4 choices. 

241
00:12:01,200 --> 00:12:04,720
You can have a monolithic 
design, little bit spaghetti, or

242
00:12:04,720 --> 00:12:08,080
you can have a modular design, 
you know, well structured, and 

243
00:12:08,080 --> 00:12:12,920
you can deploy A monolith or you
can deploy in more small pieces.

244
00:12:13,200 --> 00:12:16,760
So now what you've achieved as 
an architect, you double the 

245
00:12:16,760 --> 00:12:19,520
solution space. 
It's no longer monolith versus 

246
00:12:19,520 --> 00:12:23,320
micro service. 
Now you have 4 quadrants. 

247
00:12:23,520 --> 00:12:26,480
This is for example where the 
model or monolith comes from. 

248
00:12:26,880 --> 00:12:29,880
Alright, so let's say we want 
software that's maintainable 

249
00:12:29,880 --> 00:12:32,800
overtime and needs to deal with 
a complex domain. 

250
00:12:33,160 --> 00:12:37,080
We wanted to be well structured 
at design time because otherwise

251
00:12:37,080 --> 00:12:40,560
we can't represent that domain. 
But maybe we don't have the 

252
00:12:40,560 --> 00:12:43,840
scalability needs to make this 
like 20 different micro 

253
00:12:43,840 --> 00:12:47,560
services, an independent auto 
healing, asynchronous, loosely 

254
00:12:47,560 --> 00:12:50,280
coupled, event driven. 
Maybe we don't actually have 

255
00:12:50,280 --> 00:12:53,440
those needs. 
And then now you have a solution

256
00:12:53,440 --> 00:12:55,840
for this, right? 
And we call that modular 

257
00:12:55,840 --> 00:12:59,840
monolith, but I call that 
mapping the map, like building 

258
00:12:59,840 --> 00:13:03,640
the frame on which we discuss 
and then we can and discuss 

259
00:13:03,640 --> 00:13:08,400
which quadrant is useful for us.
But making that a 2 step process

260
00:13:08,600 --> 00:13:11,280
makes for a much more 
constructive discussion because 

261
00:13:11,280 --> 00:13:14,400
it's no longer sort of you 
versus me or my thing is better 

262
00:13:14,400 --> 00:13:17,080
than your thing kind of thing, 
which some engineers like to do.

263
00:13:17,360 --> 00:13:20,680
But it's more like, here's our 
world view and now let's discuss

264
00:13:20,680 --> 00:13:23,480
where in which quadrant we are 
and and why. 

265
00:13:23,720 --> 00:13:25,920
And that ends up being a lot 
easier thing. 

266
00:13:25,920 --> 00:13:30,080
So you disagree on some parts, 
but you have a common framing 

267
00:13:30,080 --> 00:13:32,840
for your disagreement. 
So the likelihood that you come 

268
00:13:32,840 --> 00:13:36,920
to a solution is much higher. 
Yeah, whenever I see a lot of 

269
00:13:36,920 --> 00:13:39,400
people discussing this 
asynchronously, typically it's 

270
00:13:39,400 --> 00:13:40,920
in Slack. 
And whenever there's been a 

271
00:13:40,920 --> 00:13:45,240
Slack message of a thread over a
40 messages, I'm like, people 

272
00:13:45,240 --> 00:13:46,640
talk to each other and jump in a
call. 

273
00:13:47,080 --> 00:13:48,800
But I know you're a big visual 
thinker. 

274
00:13:49,160 --> 00:13:51,720
How quickly do you actually say,
OK, we need to park this and we 

275
00:13:51,720 --> 00:13:54,240
need to Draw Something out 
visually so we can actually see 

276
00:13:54,440 --> 00:13:56,400
if we're speaking the same 
language and talking about the 

277
00:13:56,400 --> 00:13:58,440
same thing? 
Yeah, I, as you, as you hinted 

278
00:13:58,440 --> 00:14:01,640
me, I like pen and paper. 
I like analogue visual kind of 

279
00:14:01,640 --> 00:14:03,080
things, white boards, flip 
charts. 

280
00:14:03,080 --> 00:14:06,920
So the workshop is actually full
of flip charts taped on the 

281
00:14:06,920 --> 00:14:11,360
wall. 
What I like to do is not use so 

282
00:14:11,360 --> 00:14:16,080
much the standard notations. 
They're useful in like 3-4 UML, 

283
00:14:16,080 --> 00:14:18,840
etcetera. 
They're useful for communicating

284
00:14:18,840 --> 00:14:21,760
what has been done. 
But I like to use sort of one of

285
00:14:21,760 --> 00:14:25,880
visual models to tease out the 
nuances, right? 

286
00:14:25,880 --> 00:14:29,080
Is it like this or is it like 
that in words? 

287
00:14:29,080 --> 00:14:32,720
It's much easier to contradict 
yourself or be fuzzy. 

288
00:14:33,480 --> 00:14:36,400
Then in the diagram, yeah, you 
make like 2 boxes and either 

289
00:14:36,400 --> 00:14:38,480
there's a line or there's no 
line, right? 

290
00:14:38,480 --> 00:14:42,360
And words, it's like these 
things have some relationship 

291
00:14:42,360 --> 00:14:44,920
and they correlate it use like 
1,000,000 words, right? 

292
00:14:45,160 --> 00:14:46,560
But you're not sure, is it two 
boxes? 

293
00:14:46,560 --> 00:14:48,000
Is it three? 
Are they connected? 

294
00:14:48,000 --> 00:14:51,000
Are they not connected? 
So a picture doesn't have that 

295
00:14:51,000 --> 00:14:53,800
fuzziness. 
So I'm a big fan of that. 

296
00:14:54,960 --> 00:14:59,200
The reason I think I jump so 
quickly to pictures is because I

297
00:14:59,200 --> 00:15:02,680
only come in at these points. 
I only generally come in when 

298
00:15:02,680 --> 00:15:06,920
people either sort of their 
debate has ground to a halt or 

299
00:15:06,920 --> 00:15:10,000
there's a high profile decision 
to be made or people can come to

300
00:15:10,000 --> 00:15:12,840
an agreement. 
So generally I arrive sort of a 

301
00:15:12,840 --> 00:15:16,800
pen in hand kind of thing 
because if they come to a 

302
00:15:16,800 --> 00:15:20,760
conclusion without me, then I 
don't need to do anything, which

303
00:15:20,760 --> 00:15:23,800
is also fine. 
But usually what I find is 

304
00:15:23,800 --> 00:15:28,080
people have different world 
views are not clear on the 

305
00:15:28,080 --> 00:15:32,000
constraints or the trade-offs or
what I call the coordinate 

306
00:15:32,000 --> 00:15:33,480
system. 
They don't know what their world

307
00:15:33,480 --> 00:15:35,840
map looks like. 
And if you don't know what your 

308
00:15:35,840 --> 00:15:38,600
world map looks like, it's very 
difficult to discuss a path 

309
00:15:39,120 --> 00:15:41,080
because you don't even know 
which sort of planet or which 

310
00:15:41,080 --> 00:15:43,880
coordinate system you're on. 
And that's where the visual 

311
00:15:43,880 --> 00:15:47,160
models tend to help a lot. 
They, they look like simple 

312
00:15:47,480 --> 00:15:50,600
sketches. 
But in the end, what I find in 

313
00:15:50,600 --> 00:15:54,920
my workshops is simple sketches 
have much more depth than people

314
00:15:55,320 --> 00:15:58,320
believe. 
We we sometimes do the exercise 

315
00:15:58,360 --> 00:16:03,240
of listing all the dimensions 
that you can put into a sketch. 

316
00:16:03,600 --> 00:16:08,840
So basically you can give things
a size, a shape, a shading, you 

317
00:16:08,840 --> 00:16:12,960
can label things in order 1234, 
you can put a legend on, you can

318
00:16:12,960 --> 00:16:17,040
nest things, you can position 
things next to each other, 

319
00:16:17,040 --> 00:16:19,840
etcetera. 
And we come to about 20 

320
00:16:19,840 --> 00:16:24,040
dimensions that you can express 
with just two colour pens and a 

321
00:16:24,040 --> 00:16:27,280
piece of paper. 
And that's the richness that you

322
00:16:27,280 --> 00:16:31,440
want because that way you tease 
out where the disconnect is or 

323
00:16:31,440 --> 00:16:33,960
where the big system design 
trade-offs are. 

324
00:16:34,120 --> 00:16:37,200
So that's why I still like pen 
and paper a lot, because the 

325
00:16:37,200 --> 00:16:40,520
tool doesn't get in your way. 
Yeah, I've seen you do this. 

326
00:16:40,880 --> 00:16:43,640
And the first time I saw this, I
was like, well, this is like 

327
00:16:43,640 --> 00:16:45,880
very effortless. 
And I feel like it's an 

328
00:16:45,880 --> 00:16:48,680
incredibly valuable skill. 
So then for me to acquire a 

329
00:16:48,680 --> 00:16:52,800
similar skill or that exact 
skill would be to just do it to 

330
00:16:52,800 --> 00:16:55,320
try and figure out which 
dimension fits with which 

331
00:16:55,320 --> 00:16:58,400
situation and to actually get 
hands on practice and 

332
00:16:58,400 --> 00:17:01,280
visualising things and then 
having a dialogue with people to

333
00:17:01,280 --> 00:17:03,800
see if it's accurate or not. 
Is that also it from your 

334
00:17:03,800 --> 00:17:05,800
perspective? 
And I would say the most 

335
00:17:05,800 --> 00:17:08,280
important thing is you don't 
need to be a gifted artist. 

336
00:17:08,400 --> 00:17:10,040
Well, ain't nobody was born an 
artist. 

337
00:17:10,040 --> 00:17:12,200
So I I took design sketching 
classes. 

338
00:17:12,240 --> 00:17:13,440
It doesn't. 
That makes it easy it. 

339
00:17:13,760 --> 00:17:15,800
Doesn't it doesn't actually 
really show it was. 

340
00:17:16,480 --> 00:17:19,560
And the reason it doesn't show 
because it's all practice. 

341
00:17:19,880 --> 00:17:22,280
It's all muscle memory. 
The people who draw the the 

342
00:17:22,280 --> 00:17:24,160
beautiful things, It's just 
muscle memory. 

343
00:17:24,160 --> 00:17:27,720
It's just pure repetition 
because I'm busy slash lazy, 

344
00:17:27,720 --> 00:17:29,520
depending on how you want to put
it. 

345
00:17:29,760 --> 00:17:32,440
I don't do it enough. 
So it's not about being a great 

346
00:17:32,440 --> 00:17:35,200
artist. 
I think what really helps in 

347
00:17:35,200 --> 00:17:38,920
this is, well, the best way to 
learn it is to do it with 

348
00:17:38,920 --> 00:17:42,280
somebody who does it right. 
Pair up, gang up, have a mentor.

349
00:17:42,520 --> 00:17:44,800
That's why I always recommend 
works so much better. 

350
00:17:45,000 --> 00:17:47,560
If you see somebody do it, 
somebody gives you feedback. 

351
00:17:47,840 --> 00:17:51,160
You learn so much faster than 
sitting in your quiet chamber 

352
00:17:51,160 --> 00:17:54,320
kind of trying to read a book 
kind kind of thing. 

353
00:17:54,520 --> 00:17:56,320
So that is, is is the first 
part. 

354
00:17:56,640 --> 00:18:00,680
The second part I think that 
really helps in this is the 

355
00:18:00,680 --> 00:18:03,800
pictures we draw the models, 
right? 

356
00:18:03,800 --> 00:18:05,840
They have meaning, They have 
semantics, right? 

357
00:18:06,320 --> 00:18:09,240
Multiplicity is one of those 
dimensions you throw, you draw 

358
00:18:09,240 --> 00:18:12,400
three stack boxes. 
Well, actually probably doesn't 

359
00:18:12,400 --> 00:18:15,520
mean three, it means multiple. 
Or if it actually supposed to 

360
00:18:15,520 --> 00:18:18,080
mean three, you probably need to
tell these people, right? 

361
00:18:18,080 --> 00:18:20,200
So there's strong semantics 
behind it. 

362
00:18:20,520 --> 00:18:26,360
So the diagrams, we draw a weird
combination of left brain and 

363
00:18:26,360 --> 00:18:28,600
right brain. 
On one hand, they're artistic, 

364
00:18:28,600 --> 00:18:31,240
they're scribbles, they're fun. 
It's a very right brain kind of 

365
00:18:31,240 --> 00:18:33,640
creative activity. 
On the other hand, they're very 

366
00:18:33,640 --> 00:18:36,400
structured because they're very 
logical models, right? 

367
00:18:36,400 --> 00:18:38,600
You draw an arrow. 
Is that the data flow? 

368
00:18:38,600 --> 00:18:40,800
Is that the control flow? 
Is the synchronous or 

369
00:18:40,800 --> 00:18:43,760
asynchronous? 
So being able to switch back and

370
00:18:43,760 --> 00:18:46,920
forth between this left brain 
and right brain, I call this the

371
00:18:46,920 --> 00:18:49,240
ping pong between left and right
brain. 

372
00:18:49,440 --> 00:18:51,520
I think that's what makes this 
work. 

373
00:18:51,520 --> 00:18:55,520
So you use your structured 
engineering mind to get 

374
00:18:55,520 --> 00:18:59,520
something down, but then you use
your creative right brain mind 

375
00:18:59,520 --> 00:19:02,240
to like, oh, is there missing 
dimension? 

376
00:19:02,240 --> 00:19:03,960
Is there another way to express 
this? 

377
00:19:03,960 --> 00:19:05,720
Is there something we're not 
seeing? 

378
00:19:06,000 --> 00:19:10,360
And then you iterate that. 
It's taken me out so a long time

379
00:19:10,360 --> 00:19:13,800
to get to that inside. 
But I still find that is the the

380
00:19:13,800 --> 00:19:17,520
best technique to flip between 
your structured logical thinking

381
00:19:17,520 --> 00:19:21,600
and you sort of creative kind of
artistic kind of thinking. 

382
00:19:21,720 --> 00:19:24,800
And I think that's actually you 
don't need to be a genius at 

383
00:19:24,800 --> 00:19:28,160
either one, but if you can flip 
back and forth, that makes a 

384
00:19:28,160 --> 00:19:30,240
world of difference. 
Yeah, and the best 

385
00:19:30,240 --> 00:19:34,360
visualizations backed by the 
story are what people are going 

386
00:19:34,360 --> 00:19:37,520
to remember. 
I feel like especially with when

387
00:19:37,520 --> 00:19:40,160
the stakes are high and you have
both in your pocket, you can 

388
00:19:40,160 --> 00:19:41,840
land and make impact. 
Right. 

389
00:19:41,840 --> 00:19:44,200
And the story is like the right 
brain, right brain, right, 

390
00:19:44,200 --> 00:19:46,720
That's what people latch onto. 
And we just did this in the 

391
00:19:46,720 --> 00:19:49,480
workshop how we looked at and we
did this like the different IT 

392
00:19:49,480 --> 00:19:52,080
strategies for different 
business models, like very left 

393
00:19:52,080 --> 00:19:55,400
brain, very architecture and we 
have a certain business strategy

394
00:19:55,400 --> 00:19:57,520
in mind. 
And what is the matching IT 

395
00:19:57,520 --> 00:19:59,920
strategy? 
It's very classic, high level. 

396
00:20:00,360 --> 00:20:03,960
Architect kind of thing. 
But then we make a matrix out of

397
00:20:03,960 --> 00:20:05,840
this. 
Well, the matrix also pretty 

398
00:20:05,840 --> 00:20:07,800
left brain, right? 
It's a very structured way of 

399
00:20:07,800 --> 00:20:10,040
looking at it. 
But then we start looking at the

400
00:20:10,040 --> 00:20:13,080
matrix and we're looking for 
patterns and for shapes. 

401
00:20:13,400 --> 00:20:16,520
Now that is very right brain. 
That's very sort of recognition.

402
00:20:16,520 --> 00:20:19,640
Well, do we see something 
interesting and out of the 

403
00:20:19,640 --> 00:20:23,200
there's something interesting. 
The story comes out of that. 

404
00:20:23,640 --> 00:20:26,680
And then to me, that is the 
winning streak of what I call 

405
00:20:26,680 --> 00:20:30,520
the architect elevator. 
If you can go into a decision 

406
00:20:30,520 --> 00:20:32,440
meeting or what we call like the
penthouse, right? 

407
00:20:32,440 --> 00:20:35,320
The important people, right? 
And you can go in there and you 

408
00:20:35,320 --> 00:20:41,360
have a catchy story or a catchy 
visual, but you can back it up 

409
00:20:41,360 --> 00:20:44,200
with your technical skill. 
If somebody pokes and says, why 

410
00:20:44,200 --> 00:20:45,960
is this? 
Why do you have an X there? 

411
00:20:45,960 --> 00:20:48,680
Why is this linked to this? 
You're like, well, because XYZ 

412
00:20:49,040 --> 00:20:51,120
to me, that is the winning 
streak. 

413
00:20:51,480 --> 00:20:56,960
The catch a little bit is you 
must have both skills to do 

414
00:20:56,960 --> 00:20:59,120
that. 
Like I have found that you 

415
00:20:59,120 --> 00:21:00,760
cannot split that into two 
parts. 

416
00:21:00,760 --> 00:21:03,720
You cannot say, oh, you do the 
technical work and then you give

417
00:21:03,720 --> 00:21:07,000
it to sort of a graphics person 
to make it prettier because they

418
00:21:07,000 --> 00:21:10,040
don't know the model underneath.
They don't know what can be 

419
00:21:10,040 --> 00:21:13,600
changed or what can be adjusted.
They don't know the semantics. 

420
00:21:13,800 --> 00:21:16,720
So basically you need to get 2 
parts in one head. 

421
00:21:16,720 --> 00:21:19,760
You need to be able to play that
ping pong with yourself, right? 

422
00:21:20,080 --> 00:21:21,440
It's like, what pattern do I 
see? 

423
00:21:21,440 --> 00:21:24,680
What story do I have? 
But is this defensible from the 

424
00:21:24,680 --> 00:21:27,240
technical underlying 
perspective? 

425
00:21:27,240 --> 00:21:30,400
So you cannot split that into 
two heads. 

426
00:21:30,680 --> 00:21:35,560
You must get this into one head.
But once you do that, that is 

427
00:21:35,560 --> 00:21:38,760
excessively valuable, all right,
because it sticks. 

428
00:21:38,760 --> 00:21:41,320
It makes sense, it has a logical
foundation. 

429
00:21:41,320 --> 00:21:44,640
It's now hand waving and 
buzzword dropping, right is real

430
00:21:44,640 --> 00:21:48,520
engineering, but it kind of 
doesn't look like engineering. 

431
00:21:48,520 --> 00:21:52,080
It's not some giant blueprint 
that nobody can understand, but 

432
00:21:52,080 --> 00:21:56,040
it's actually a catchy storyline
or a catchy visual or what we 

433
00:21:56,040 --> 00:21:58,400
tried to do. 
We actually give them a name. 

434
00:21:58,800 --> 00:22:01,840
So today we just named the DIT 
Strategy Ladder. 

435
00:22:02,040 --> 00:22:04,240
And it's a ladder because it has
this shape. 

436
00:22:04,240 --> 00:22:07,240
So immediately it's sticky in 
people's heads. 

437
00:22:07,560 --> 00:22:12,080
And I think there's a little bit
of I'm always shy with recipes 

438
00:22:12,080 --> 00:22:16,640
because I always say like the 
recipe book and the really good 

439
00:22:16,640 --> 00:22:18,640
cooking school are two very 
different things. 

440
00:22:18,960 --> 00:22:21,720
The recipe book is like, well, 
if you can't feed yourself and 

441
00:22:21,720 --> 00:22:24,000
you need to make something, but 
follow the recipe. 

442
00:22:24,160 --> 00:22:26,760
But if you go to like cotton 
Bleu or somewhere like a real 

443
00:22:26,760 --> 00:22:28,600
cooking school, it's not about 
recipes. 

444
00:22:28,880 --> 00:22:33,120
You learn how things work and 
how they function and how to 

445
00:22:33,120 --> 00:22:35,360
make the recipes. 
So that's why I always careful 

446
00:22:35,360 --> 00:22:37,440
about recipes. 
But coming back to our 

447
00:22:37,560 --> 00:22:40,960
visualization, I think you're 
doing this kind of ping pong, 

448
00:22:41,320 --> 00:22:42,680
right? 
Seeing, do I have the 

449
00:22:42,680 --> 00:22:45,680
foundation? 
Can I visualize this OK? 

450
00:22:45,680 --> 00:22:47,880
Do I need more detail? 
Can I visualize that? 

451
00:22:48,080 --> 00:22:51,800
That is probably the best kind 
of recipe that that I would 

452
00:22:51,800 --> 00:22:53,680
suggest. 
There might be a lot of people 

453
00:22:53,680 --> 00:22:57,560
listening and they are in an icy
role, so very much hands on 

454
00:22:57,560 --> 00:23:00,760
responsible as software 
engineers and they could think 

455
00:23:01,000 --> 00:23:03,120
this is really cool. 
Like I want to experiment with 

456
00:23:03,120 --> 00:23:05,040
this. 
I feel like on a smaller scale 

457
00:23:05,040 --> 00:23:07,320
within your team you can. 
But when we're talking about 

458
00:23:07,320 --> 00:23:10,520
being responsible for software 
architecture or system design, 

459
00:23:10,520 --> 00:23:13,640
system architecture and then 
enterprise architect, what from 

460
00:23:13,640 --> 00:23:16,160
your perspective are the skills 
that they need from the hard 

461
00:23:16,160 --> 00:23:18,040
skill side? 
Let's start with that and then 

462
00:23:18,040 --> 00:23:19,680
eventually from the soft skill 
side as well. 

463
00:23:20,240 --> 00:23:22,760
Yeah, You think the hard skills 
side, you just got to understand

464
00:23:22,760 --> 00:23:25,160
your tech, you've got to be a 
good developer, understand the 

465
00:23:25,160 --> 00:23:27,200
trade-offs. 
And I would say the biggest 

466
00:23:27,200 --> 00:23:30,880
mistake I see on the hard skills
side is the people who used to 

467
00:23:30,880 --> 00:23:33,760
be technical, you know, they 
have the hard skills, but from 

468
00:23:33,760 --> 00:23:36,360
like five years ago or will be 
10 years ago. 

469
00:23:36,680 --> 00:23:39,920
That is what makes our field 
exciting, but it's also what 

470
00:23:39,920 --> 00:23:41,520
makes our field dangerous, 
right? 

471
00:23:41,520 --> 00:23:44,360
Because having had the hard 
skills might lead you to wrong 

472
00:23:44,600 --> 00:23:47,360
assumptions because your 
decision or your trade off might

473
00:23:47,360 --> 00:23:51,160
have been a good one five years 
ago or 8 or 10 years ago, but it

474
00:23:51,160 --> 00:23:53,600
no longer is. 
So the main thing is you need to

475
00:23:53,600 --> 00:23:56,360
keep your hard skills, your 
system design skills, understand

476
00:23:56,360 --> 00:24:00,760
trade-offs, operational aspects,
observability, domain driven 

477
00:24:00,760 --> 00:24:03,120
design, right? 
These are all things you, you 

478
00:24:03,120 --> 00:24:05,840
need to do. 
Then you know the, the skills 

479
00:24:05,840 --> 00:24:07,600
that you need to add and the 
mix. 

480
00:24:07,600 --> 00:24:11,120
And as I said, it's not just 
about adding more skills, it's 

481
00:24:11,120 --> 00:24:13,720
about finding the intersection 
between the skills, right? 

482
00:24:13,720 --> 00:24:17,000
I call this the the multiplier. 
You want the multiplier fact. 

483
00:24:17,000 --> 00:24:19,720
It's not just, oh, you're a good
techie person and you're a good 

484
00:24:19,720 --> 00:24:22,000
communicator. 
You're a good technical 

485
00:24:22,000 --> 00:24:24,400
communicator. 
And it's the combination of 

486
00:24:24,400 --> 00:24:27,640
both. 
That's, to me, the mental model 

487
00:24:27,640 --> 00:24:32,680
that you should have. 
The good test is do people come 

488
00:24:32,680 --> 00:24:36,040
to you to make them smarter? 
Do they come to you with a 

489
00:24:36,040 --> 00:24:40,320
problem and you know, not asking
you to solve the problem for 

490
00:24:40,320 --> 00:24:44,000
them, but they consider you a 
good sounding board like a 

491
00:24:44,520 --> 00:24:46,360
rubber ducky. 
We call this like the rubber 

492
00:24:46,360 --> 00:24:48,760
ducking. 
Are you a popular rubber duck 

493
00:24:48,760 --> 00:24:52,120
where people explain the problem
and you ask maybe a few 

494
00:24:52,120 --> 00:24:55,000
questions, maybe give a few 
hints and people go back like, 

495
00:24:55,320 --> 00:24:57,600
oh, that's an interesting way to
think about it. 

496
00:24:57,600 --> 00:25:01,080
Or you make a little sketch and 
you say, hey, would this kind of

497
00:25:01,080 --> 00:25:02,120
help you? 
And they're like, oh, and they 

498
00:25:02,120 --> 00:25:04,600
walk off with your sketch and 
they go do something. 

499
00:25:04,800 --> 00:25:08,520
I think if you can pull that 
off, that people see you as a, 

500
00:25:08,880 --> 00:25:12,600
as a source to kind of give them
a little mental push or a little

501
00:25:12,600 --> 00:25:17,640
bit of a mental boost, then for 
me that is a fantastic positive 

502
00:25:17,640 --> 00:25:19,400
feedback. 
And it means you're you're 

503
00:25:19,400 --> 00:25:21,560
actually on to something. 
Yeah, I love that. 

504
00:25:21,680 --> 00:25:23,200
That's a fantastic signal, 
actually. 

505
00:25:23,200 --> 00:25:26,160
If people come to you and you 
give them perspective or you 

506
00:25:26,160 --> 00:25:28,920
help them along the way, help 
them figure out what they 

507
00:25:28,920 --> 00:25:31,600
already know based on how you 
how much you know about them. 

508
00:25:31,680 --> 00:25:32,960
I think that's incredibly 
powerful. 

509
00:25:33,160 --> 00:25:36,800
Yeah, and I have the in the 
book, I have the model of the 

510
00:25:36,800 --> 00:25:39,120
phantom sketch artist, right? 
It's one of the I like 

511
00:25:39,120 --> 00:25:40,680
metaphors. 
So it's one of the many 

512
00:25:41,000 --> 00:25:43,120
metaphors, right. 
But what the phantom sketch 

513
00:25:43,120 --> 00:25:47,480
artist metaphor underliances 
that knowing something and being

514
00:25:47,480 --> 00:25:50,560
able to express something at two
different skills. 

515
00:25:50,960 --> 00:25:53,440
And the story is based in the 
old days when we didn't have CC 

516
00:25:53,440 --> 00:25:55,320
to camera. 
Camera are like every two 

517
00:25:55,320 --> 00:25:58,840
meters, right when the bank got 
dropped, somebody would come 

518
00:25:58,840 --> 00:26:01,480
talk to the witnesses and say 
well, what did the perpetrator 

519
00:26:01,480 --> 00:26:04,480
look like right? 
And you would think, oh, you 

520
00:26:04,480 --> 00:26:06,840
just give people a pen and a 
paper and they're going to draw 

521
00:26:06,840 --> 00:26:09,440
a picture of the bank robber. 
Well, you're going to get a 

522
00:26:09,440 --> 00:26:12,160
stick figure of the guy with the
money bag and a gun or something

523
00:26:12,160 --> 00:26:16,120
like ridiculous, right? 
So people know they have seen 

524
00:26:16,120 --> 00:26:19,480
what the person looks like, but 
they don't have the skill to 

525
00:26:19,480 --> 00:26:23,480
express that versus the phantom 
sketch artist doesn't know what 

526
00:26:23,480 --> 00:26:26,440
the person looks like, but they 
have the skill of like 

527
00:26:26,440 --> 00:26:29,520
articulating and expressing and 
the combination. 

528
00:26:29,720 --> 00:26:32,320
That's what makes the the good, 
good sketch. 

529
00:26:32,320 --> 00:26:35,520
And often as an architect, I 
tend to be the phantom sketch 

530
00:26:35,520 --> 00:26:37,400
artist. 
I don't have any more knowledge 

531
00:26:37,400 --> 00:26:39,240
than the other people. 
They know everything. 

532
00:26:39,240 --> 00:26:41,760
It's their application. 
They know what the bank robber 

533
00:26:41,760 --> 00:26:44,520
looks like. 
But I can help them express 

534
00:26:44,520 --> 00:26:47,800
that, put that in a different 
format or different way. 

535
00:26:47,800 --> 00:26:50,800
We're like when people say, Oh 
yeah, just like this, right? 

536
00:26:51,000 --> 00:26:54,880
And to me, that is a really good
collaboration because neither 

537
00:26:54,880 --> 00:26:58,520
one is telling the other person 
what to do or try to be smarter 

538
00:26:58,520 --> 00:27:01,320
than the other person, right. 
The phantom sketch artist is a 

539
00:27:01,320 --> 00:27:04,280
best, probably a better drawer 
than you are, but they can't do 

540
00:27:04,280 --> 00:27:06,280
anything without your input, 
right. 

541
00:27:06,280 --> 00:27:10,360
So it's a nice example of you 
need the combination of the two 

542
00:27:10,360 --> 00:27:12,400
skills. 
And that is generally 

543
00:27:12,400 --> 00:27:14,800
appreciated by people. 
Like, I come to them and say, 

544
00:27:14,800 --> 00:27:16,440
hey, tell me about your system, 
right. 

545
00:27:16,760 --> 00:27:20,120
And I start drawing something. 
And my favorite comment is, oh, 

546
00:27:20,120 --> 00:27:21,880
that's wrong. 
Excellent. 

547
00:27:21,960 --> 00:27:23,320
Yeah. 
Because, you know, you're 

548
00:27:23,320 --> 00:27:25,640
getting more out of people. 
It's like, oh, I drew a line 

549
00:27:25,640 --> 00:27:26,440
here. 
No, no, no, no. 

550
00:27:26,440 --> 00:27:28,560
This is not how it works. 
I'm like, well, excellent. 

551
00:27:28,680 --> 00:27:29,920
Well, how does it look? 
Right. 

552
00:27:30,080 --> 00:27:34,480
Once they see you draw back to 
what you understood, they can 

553
00:27:34,480 --> 00:27:36,640
easily see things. 
Oh, that's not what I meant. 

554
00:27:36,640 --> 00:27:38,720
Or there's something else. 
And then you're in a 

555
00:27:38,720 --> 00:27:41,120
constructive dialogue. 
And that's exactly like the 

556
00:27:41,120 --> 00:27:43,520
Phantom sketch artist, right? 
Did they have a wider nose, a 

557
00:27:43,520 --> 00:27:47,160
bigger eyebrows or different 
eyes or kind of the face of the,

558
00:27:47,160 --> 00:27:49,760
the shape of their face, Right. 
You're in a dialogue mode. 

559
00:27:49,760 --> 00:27:51,800
And in the end, they come out 
with a thing where people say 

560
00:27:52,200 --> 00:27:54,920
exactly like that, but they 
could have never made that 

561
00:27:54,960 --> 00:27:58,120
themselves. 
And I think if you do that, you 

562
00:27:58,120 --> 00:28:00,280
will become this go to person, 
right? 

563
00:28:00,280 --> 00:28:04,280
People walk away with something 
that's valuable to them, but you

564
00:28:04,280 --> 00:28:07,480
didn't lecture them or like 
constrain them or tell them, oh,

565
00:28:07,480 --> 00:28:09,920
you have to do it. 
Like this way is basically 

566
00:28:09,920 --> 00:28:13,280
extracted knowledge they already
had, but you played it back to 

567
00:28:13,280 --> 00:28:17,280
them in a format that's easily 
consumed and everybody wins. 

568
00:28:17,360 --> 00:28:19,920
I learn a lot, right? 
And they have a better 

569
00:28:20,240 --> 00:28:23,920
understanding of the trade-offs 
or the the system design. 

570
00:28:24,160 --> 00:28:27,280
There's there's one in this 
metaphor. 

571
00:28:27,320 --> 00:28:29,160
I'd always like playing with 
these metaphors. 

572
00:28:29,160 --> 00:28:33,080
There's one more quality that 
makes this very much related to 

573
00:28:33,720 --> 00:28:37,440
architecture, and that is in 
order to become a phantom sketch

574
00:28:37,520 --> 00:28:40,360
artist, kind of rare today, but 
when people wanted to become, 

575
00:28:40,680 --> 00:28:43,240
you need to study human anatomy.
Yeah. 

576
00:28:43,240 --> 00:28:45,720
You're not a painter. 
You're not a landscape painter, 

577
00:28:45,720 --> 00:28:47,920
random person. 
You draw faces. 

578
00:28:47,920 --> 00:28:51,640
And in order to draw a good 
depiction of a face, you need to

579
00:28:51,640 --> 00:28:55,600
understand the architecture of 
the body underneath, like 

580
00:28:55,600 --> 00:28:59,520
people, posture, face structure,
the bones and kind of things. 

581
00:28:59,800 --> 00:29:03,320
So the Phantom sketch art is in 
a way is a form of architect 

582
00:29:03,640 --> 00:29:07,040
because they learn basically, 
yeah, the the human bone 

583
00:29:07,040 --> 00:29:11,120
structures and anatomy because 
otherwise they cannot draw a 

584
00:29:11,120 --> 00:29:13,600
good picture. 
And that same is true for us. 

585
00:29:13,800 --> 00:29:17,240
It's not about drawing the most 
beautiful rectangle, but it's 

586
00:29:17,240 --> 00:29:20,480
understanding the anatomy, 
understanding what people say 

587
00:29:20,480 --> 00:29:24,920
about the relationship between 
the system components, and then 

588
00:29:24,920 --> 00:29:27,800
you draw the rectangles that 
represent that. 

589
00:29:27,800 --> 00:29:30,400
But you got to understand the 
subject matter. 

590
00:29:30,640 --> 00:29:34,240
Yeah, I I was thinking when you 
were speaking about hard skills,

591
00:29:34,720 --> 00:29:37,680
how much of enterprise 
architecture or being an 

592
00:29:37,680 --> 00:29:40,280
architect in essence has changed
nowadays? 

593
00:29:40,480 --> 00:29:42,800
Because I feel like I get 
bombarded with tools. 

594
00:29:43,080 --> 00:29:46,120
There's a lot more possible, 
specifically with large language

595
00:29:46,120 --> 00:29:49,440
models, but in essence, like the
Phantom Sketch artist, 

596
00:29:50,000 --> 00:29:52,240
everything underneath kind of 
still holds true. 

597
00:29:52,800 --> 00:29:55,280
From your perspective. 
Does architecture evolve 

598
00:29:55,360 --> 00:29:57,640
quickly? 
Because I feel like tooling does

599
00:29:57,640 --> 00:30:00,600
at a more rapid pace. 
Definitely the environment in 

600
00:30:00,600 --> 00:30:03,640
which you operate has changed. 
So there's a couple of 

601
00:30:04,080 --> 00:30:08,080
assumptions that architects used
to be able to make to make their

602
00:30:08,080 --> 00:30:11,200
lives easier that no longer hold
right. 

603
00:30:11,200 --> 00:30:14,040
And the one is like 
snapshotting, like the world is 

604
00:30:14,040 --> 00:30:16,600
moving too fast to be able to 
pretend. 

605
00:30:16,600 --> 00:30:18,960
It's like, OK, let's just assume
nothing changes. 

606
00:30:19,360 --> 00:30:22,400
Well, that turns out to be a 
very poor assumption in today's 

607
00:30:22,400 --> 00:30:25,000
environment. 
So that, for example, means 

608
00:30:25,000 --> 00:30:30,800
exercises like trying to draw a 
giant ID landscape, you know, or

609
00:30:30,960 --> 00:30:34,200
cataloguing all the applications
you have and putting that into 

610
00:30:34,200 --> 00:30:35,880
some tools. 
Taking a few months for that. 

611
00:30:35,920 --> 00:30:37,280
Yeah. 
And then a few months, by the 

612
00:30:37,280 --> 00:30:39,560
time you're done, this has 
already changed. 

613
00:30:39,560 --> 00:30:43,480
So for me, the assumption of 
being able to snapshot and 

614
00:30:43,480 --> 00:30:46,520
being, I call this the 
cartographer, being the person 

615
00:30:46,520 --> 00:30:50,720
who makes all the great 
landscapes, that is very useful,

616
00:30:50,720 --> 00:30:53,480
but it's no longer practical 
because stuff changes too 

617
00:30:53,480 --> 00:30:55,720
quickly. 
So one of the metaphors I use 

618
00:30:55,720 --> 00:30:59,360
which captures the change in 
role for Enterprise Architects 

619
00:30:59,360 --> 00:31:02,760
is rather than trying to be the 
cartographer who walks around 

620
00:31:02,760 --> 00:31:06,480
with a giant map of all the 
pieces, you need to be much more

621
00:31:06,480 --> 00:31:10,440
like a scout. 
You have a objective in mind. 

622
00:31:10,680 --> 00:31:13,360
You want to cross the river or 
beat the enemy or like move 

623
00:31:13,360 --> 00:31:14,840
somewhere or like whatever, 
right? 

624
00:31:14,840 --> 00:31:17,800
You have some objective, you 
send some scouts, right? 

625
00:31:17,800 --> 00:31:21,040
They're gonna go look at things.
They figure out what works, what

626
00:31:21,040 --> 00:31:23,640
won't work, right? 
And they come back to you with a

627
00:31:23,640 --> 00:31:28,080
very simple map, but a very 
timely and very expressive map, 

628
00:31:28,480 --> 00:31:29,640
right. 
So you want to cross the earth 

629
00:31:29,640 --> 00:31:31,720
like, oh, here's the mountain, 
here's the enemy, the bridge is 

630
00:31:31,720 --> 00:31:34,400
gone, but here's not so deep we 
can cross there, we can go 

631
00:31:34,400 --> 00:31:38,000
around whatever it is, right But
they have a very situational 

632
00:31:38,200 --> 00:31:41,880
they have a sort of purpose 
driven map The map is not trying

633
00:31:41,880 --> 00:31:46,040
to be like the perfect map of 
all the exact landscape. 

634
00:31:46,280 --> 00:31:49,240
They only depict what's it's 
relevant to your move. 

635
00:31:49,840 --> 00:31:54,840
And I think that is becoming a 
more viable metaphor of what 

636
00:31:54,840 --> 00:31:59,720
enterprise architects should do.
It's no longer plausible to like

637
00:31:59,880 --> 00:32:04,000
maintain this giant map where 
everything is on and everything 

638
00:32:04,000 --> 00:32:06,920
is up to date. 
Rather, you need to have a 

639
00:32:06,920 --> 00:32:08,680
purpose. 
You need to have a direction in 

640
00:32:08,680 --> 00:32:10,200
mind. 
Like LLMS Gen. 

641
00:32:10,200 --> 00:32:11,760
EI is a great example, right? 
OK. 

642
00:32:11,760 --> 00:32:14,680
De enterprise architects, what's
our strategy for this? 

643
00:32:14,920 --> 00:32:17,040
Where should we find the best 
first use cases? 

644
00:32:17,040 --> 00:32:19,560
How do we integrate this with 
the rest of our systems? 

645
00:32:19,840 --> 00:32:23,640
We know LMS have super high rate
of evolution, right? 

646
00:32:23,640 --> 00:32:27,120
So how do we keep that change 
out of other systems, right? 

647
00:32:27,600 --> 00:32:30,600
What are we doing with with 
Agents and Agentec, right? 

648
00:32:30,720 --> 00:32:33,680
Is that just another form of 
workflow integration or is that 

649
00:32:33,680 --> 00:32:36,080
something novel, right? 
Give me a point of view. 

650
00:32:36,280 --> 00:32:40,560
But it's very purpose driven. 
And as you will quickly see, you

651
00:32:40,560 --> 00:32:43,200
can only do that if you 
understand the subject matter, 

652
00:32:43,600 --> 00:32:45,920
right? 
You can't sort of arm wave your 

653
00:32:45,920 --> 00:32:49,240
way through that, right? 
You need to have, you know, the 

654
00:32:49,240 --> 00:32:52,640
hard tech skills as well. 
But for me, it's very purpose 

655
00:32:52,640 --> 00:32:54,240
driven. 
What should we do? 

656
00:32:54,240 --> 00:32:56,160
Do we have a point of view, 
right? 

657
00:32:56,360 --> 00:33:00,240
Do we have a question in mind? 
Too many architects try to find 

658
00:33:00,280 --> 00:33:02,520
answers when they don't have a 
question. 

659
00:33:02,520 --> 00:33:05,480
So it's like, oh, here's the 
thing, that answers all possible

660
00:33:05,480 --> 00:33:07,880
questions. 
I'm like, that would be nice. 

661
00:33:08,960 --> 00:33:11,280
Wouldn't that be, wouldn't that 
be easy? 

662
00:33:11,520 --> 00:33:15,600
That doesn't work anymore. 
You got to start with a question

663
00:33:15,600 --> 00:33:18,120
in your mind. 
And then like the scout, I was 

664
00:33:18,120 --> 00:33:19,480
like, whatever, can we cross the
river? 

665
00:33:19,480 --> 00:33:21,840
I was OK, come back with the 
answer to this. 

666
00:33:21,840 --> 00:33:24,600
It's got to be situational and 
timely. 

667
00:33:24,880 --> 00:33:28,360
Otherwise it becomes hard work. 
And we might like our hard work,

668
00:33:28,520 --> 00:33:29,840
but I always say we're not a 
museum. 

669
00:33:30,080 --> 00:33:32,600
We're the IT. 
You can't be a historian, Yeah. 

670
00:33:32,720 --> 00:33:34,000
Exactly. 
We can't be a historian. 

671
00:33:34,000 --> 00:33:35,600
We're not collecting modern art,
right? 

672
00:33:35,920 --> 00:33:38,120
Our diagrams serve a purpose, 
right? 

673
00:33:38,120 --> 00:33:41,080
They help us make better 
decisions, get a better 

674
00:33:41,080 --> 00:33:43,400
decision, Transparency and 
discipline. 

675
00:33:43,760 --> 00:33:49,080
So a the decision is, is clearer
and everybody's on the same page

676
00:33:49,080 --> 00:33:51,840
to what the decision was 
actually what drove the 

677
00:33:51,840 --> 00:33:53,840
decision. 
What trade-offs are we willing 

678
00:33:53,840 --> 00:33:56,720
to make? 
That's what adding the value and

679
00:33:56,720 --> 00:33:59,320
that the picture just has to be 
a means to that end. 

680
00:33:59,520 --> 00:34:02,480
If that end isn't there, the 
picture is also worthless. 

681
00:34:02,760 --> 00:34:06,720
Then it just becomes modern arts
and boxes and triangles and kind

682
00:34:06,720 --> 00:34:09,560
of kind of things one. 
Of the things you mentioned was 

683
00:34:09,560 --> 00:34:13,880
also how it can be incredibly 
dangerous to have had experience

684
00:34:13,880 --> 00:34:16,840
10 years ago but then be 10 
years kind of hands off and 

685
00:34:16,840 --> 00:34:19,159
still rely fully on that 
experience. 

686
00:34:19,400 --> 00:34:23,040
And then I look at modern day, 
current age where things are 

687
00:34:23,040 --> 00:34:26,040
evolving quite quickly. 
So then I don't know even from 

688
00:34:26,040 --> 00:34:28,560
your perspective, how do you 
keep up to date with things that

689
00:34:28,560 --> 00:34:30,639
are evolving? 
How do you keep on top of things

690
00:34:30,639 --> 00:34:33,600
with regards to your own skill 
set to then educate others as 

691
00:34:33,600 --> 00:34:35,840
well? 
And the danger is particularly 

692
00:34:35,840 --> 00:34:39,120
pronounced because you might be 
well meaning and you have a 

693
00:34:39,120 --> 00:34:41,360
technical foundation. 
You make rational and good 

694
00:34:41,360 --> 00:34:44,400
decisions, but based on outdated
constraints. 

695
00:34:44,719 --> 00:34:47,600
That is a very easy trap to to 
fall into. 

696
00:34:47,800 --> 00:34:49,679
I had a classic example I always
give us. 

697
00:34:49,960 --> 00:34:51,920
Like we were told, everything 
must scale out. 

698
00:34:52,639 --> 00:34:56,120
But the reality is Moore's Law 
grows faster than most 

699
00:34:56,120 --> 00:34:59,600
businesses do. 
So in reality, a lot of modern 

700
00:34:59,600 --> 00:35:03,560
business applications could run 
in a single server in memory. 

701
00:35:04,120 --> 00:35:07,240
You get a couple terabytes of 
RAM, no problem. 

702
00:35:07,560 --> 00:35:09,360
Yeah, your customer data, your 
things. 

703
00:35:09,360 --> 00:35:11,960
I mean, maybe not all your logs 
from all history, but all the 

704
00:35:11,960 --> 00:35:14,000
current data will fit in a 
couple terabytes of RAM. 

705
00:35:14,000 --> 00:35:16,920
Like what kind of business? 
I mean, if you're not Netflix or

706
00:35:16,920 --> 00:35:19,880
eBay, right? 
If it's like a normal kind of 

707
00:35:20,080 --> 00:35:23,880
transport insurance kind of 
thing, right, you'll probably be

708
00:35:23,880 --> 00:35:26,520
fine, right? 
So that's the what's that's the 

709
00:35:26,520 --> 00:35:29,240
danger there. 
So the main thing that 

710
00:35:29,240 --> 00:35:33,080
architects need to do is you 
need to revalidate your 

711
00:35:33,080 --> 00:35:36,920
heuristics because everybody is 
using your heuristics. 

712
00:35:36,960 --> 00:35:40,400
The example I always give, let's
say you a chief architect or 

713
00:35:40,400 --> 00:35:42,480
some important architect, 
somebody comes and they're like,

714
00:35:42,680 --> 00:35:44,880
they've all these little boxes 
in the middle is like a big 

715
00:35:44,880 --> 00:35:46,200
barrel. 
It's a database. 

716
00:35:46,240 --> 00:35:48,520
Everything talks to this 
database, right? 

717
00:35:48,520 --> 00:35:51,640
Your first heuristic is oh, 
performance bottleneck, right? 

718
00:35:51,640 --> 00:35:53,640
Everything talks to database. 
Bad news, right? 

719
00:35:53,960 --> 00:35:56,880
Then you find out, oh, or you 
like, or it could also be 

720
00:35:56,960 --> 00:35:59,320
brittle integration, right? 
Because the schema becomes 

721
00:35:59,320 --> 00:36:00,760
difficult to maintain, yadda, 
yadda. 

722
00:36:01,240 --> 00:36:03,960
And then you find out, oh, this 
is actually a no secret cloud 

723
00:36:03,960 --> 00:36:06,400
database. 
It scales more than you can 

724
00:36:06,400 --> 00:36:10,040
afford probably, right? 
It will scale and you have soft 

725
00:36:10,040 --> 00:36:11,600
schema or it's different 
documents. 

726
00:36:11,600 --> 00:36:14,560
So it turns out to be actually 
no problem, right? 

727
00:36:14,560 --> 00:36:17,560
This is where your heuristics 
are outdated. 

728
00:36:18,080 --> 00:36:21,920
So how do you compensate that? 
Well, the the easy answer, but 

729
00:36:21,920 --> 00:36:24,680
the easy, not easy answer is 
well, you should spend more 

730
00:36:24,680 --> 00:36:28,360
hands on time, but how are you 
going to spend hands on time 

731
00:36:28,520 --> 00:36:32,880
with all possible technologies. 
So what I do quite a bit is hang

732
00:36:32,880 --> 00:36:36,040
out with people who do this 
right, have a network. 

733
00:36:36,240 --> 00:36:39,160
I don't think it's something 
that you can expect to manage 

734
00:36:39,160 --> 00:36:43,360
yourself just too much going on 
like Jenny I is a great example.

735
00:36:43,360 --> 00:36:48,960
I decided to miss the first sort
of 12 to 18 months of the 

736
00:36:48,960 --> 00:36:51,840
madness because I was like, I'm 
busy and it was madness. 

737
00:36:51,840 --> 00:36:53,200
Yeah, Yeah. 
No time for this. 

738
00:36:53,360 --> 00:36:55,040
But then like, oh, I need to 
catch up quickly. 

739
00:36:55,280 --> 00:36:57,920
So what do I do? 
Well, I chat with my buddies, 

740
00:36:58,160 --> 00:36:59,040
Right. 
What do you guys build? 

741
00:36:59,040 --> 00:37:00,720
Like Rod Johnson, old friend of 
mine, he can buy. 

742
00:37:00,720 --> 00:37:03,000
Well, actually, it happened to 
live in the South of France. 

743
00:37:03,000 --> 00:37:05,560
So he stopped by. 
He stayed with me for two days 

744
00:37:05,920 --> 00:37:07,800
and in two days I got the full 
download. 

745
00:37:08,720 --> 00:37:10,080
Fantastic. 
Fantastic. 

746
00:37:10,320 --> 00:37:12,520
Great deal, right. 
It's like, Yep. 

747
00:37:12,520 --> 00:37:14,960
So he stayed with me and then, 
yeah, I got the whole download, 

748
00:37:14,960 --> 00:37:17,520
what he's working on. 
He showed me all the code he's 

749
00:37:17,520 --> 00:37:20,120
writing and then it becomes 
super high bandwidth. 

750
00:37:20,440 --> 00:37:25,600
And I highly recommend that you 
need some framework and you need

751
00:37:25,600 --> 00:37:27,080
to not waste people's time, 
right? 

752
00:37:27,080 --> 00:37:29,320
You need to be able to absorb at
the high rate. 

753
00:37:29,440 --> 00:37:32,560
So the reprocess your 
assumption, update your, 

754
00:37:32,560 --> 00:37:36,320
validate your heuristics. 
But don't think you can do this 

755
00:37:36,400 --> 00:37:39,880
all on your own. 
And unfortunately, I think 

756
00:37:39,880 --> 00:37:44,240
social media has become a not so
great source for catching up on 

757
00:37:44,240 --> 00:37:47,160
technologies and trade-offs 
because everybody's peddling 

758
00:37:47,400 --> 00:37:51,040
something on or, or other. 
So you got to have some trusted 

759
00:37:51,240 --> 00:37:55,560
sources that you can go buy them
a beer or coffee or something 

760
00:37:55,800 --> 00:37:57,680
and hang out with them. 
It's like, hey, that framework, 

761
00:37:57,680 --> 00:37:59,880
what do you actually think? 
Is that thing for real? 

762
00:37:59,880 --> 00:38:01,600
You know, what are you working 
on? 

763
00:38:01,880 --> 00:38:03,800
You know, basically be the 
social geek. 

764
00:38:04,000 --> 00:38:08,080
You know, talk to people. 
Talking to people can be a great

765
00:38:08,080 --> 00:38:10,520
strategy for being a good 
architect, right? 

766
00:38:10,520 --> 00:38:12,320
Don't lock yourself up in the 
chamber. 

767
00:38:12,560 --> 00:38:15,960
Go hang out with people, benefit
from the experience, learn from 

768
00:38:15,960 --> 00:38:18,040
what they're doing. 
I think that's the only way. 

769
00:38:18,240 --> 00:38:21,280
Yeah, I think it's also, even if
you don't have a professional 

770
00:38:21,280 --> 00:38:23,880
network yet, there are a lot of 
meet ups, There are a lot of 

771
00:38:23,880 --> 00:38:26,120
conferences, some are completely
free to attend. 

772
00:38:26,120 --> 00:38:29,600
You can be a social person to 
actually talk to people and see 

773
00:38:29,600 --> 00:38:32,520
what they're working on and gain
knowledge that way, even if you 

774
00:38:32,520 --> 00:38:35,160
haven't built up your, let's 
say, enterprise network yet. 

775
00:38:35,440 --> 00:38:38,040
And then once you actually have 
a role, find mentors, people 

776
00:38:38,040 --> 00:38:39,840
that will coach you on the job 
and off the job. 

777
00:38:40,320 --> 00:38:43,960
I think it's actually very 
achievable to grow quite quickly

778
00:38:43,960 --> 00:38:44,760
through that. 
As well. 

779
00:38:44,960 --> 00:38:46,240
Yeah. 
And I think for an architect 

780
00:38:46,240 --> 00:38:49,960
that's absolutely required. 
I mean, I even, I don't buy 

781
00:38:49,960 --> 00:38:53,480
quite into even like the senior 
engineer who sits in the quiet 

782
00:38:53,480 --> 00:38:56,320
chamber and never like talks to 
anybody and never looks anybody 

783
00:38:56,320 --> 00:38:59,040
in the eye and just kind of 
Mumbles to themselves even. 

784
00:38:59,040 --> 00:39:01,800
I doubt that. 
But as an architect I would go 

785
00:39:01,800 --> 00:39:04,160
as far as say like completely 
impossible. 

786
00:39:04,160 --> 00:39:06,160
Impossible. 
You can't be setting up in your 

787
00:39:06,160 --> 00:39:09,240
little chamber and people put 
requirements in and you spit 

788
00:39:09,240 --> 00:39:12,720
architectures out or something. 
It's like, I don't think, I 

789
00:39:12,720 --> 00:39:14,280
think that's how it works, 
right? 

790
00:39:14,280 --> 00:39:17,360
It's inherently, well, a from 
the impact that you're making, 

791
00:39:17,360 --> 00:39:18,600
right? 
If you want to make other people

792
00:39:18,600 --> 00:39:21,520
smarter, how are you going to do
that without talking to them? 

793
00:39:21,880 --> 00:39:24,600
And vice versa? 
For you to stay up to date, you 

794
00:39:24,600 --> 00:39:27,400
need to have a network. 
You need to be able to get 

795
00:39:27,400 --> 00:39:30,040
feedback. 
You need to learn maybe even 

796
00:39:30,040 --> 00:39:32,400
your own decisions, right? 
You need to learn what worked 

797
00:39:32,400 --> 00:39:34,200
and what maybe didn't work as 
well. 

798
00:39:34,240 --> 00:39:36,760
Yeah, maybe you thought it was a
great architecture and people 

799
00:39:36,760 --> 00:39:39,440
come back. 
It's like, you know, didn't 

800
00:39:39,440 --> 00:39:42,600
really do what we needed or too 
expensive or too cumbersome, 

801
00:39:42,680 --> 00:39:44,480
right? 
You want to be the person where 

802
00:39:44,480 --> 00:39:48,200
people come back and say, look, 
sounded like a good idea but 

803
00:39:48,480 --> 00:39:49,880
didn't work out that well, 
right? 

804
00:39:49,880 --> 00:39:52,840
You don't want to be the person 
where people are afraid, you 

805
00:39:52,840 --> 00:39:55,080
know, and they're like great 
advice, great advice. 

806
00:39:55,080 --> 00:39:56,640
And then they do the opposite 
thing, right? 

807
00:39:56,840 --> 00:39:59,680
Because you won't learn and you 
become out of date very quickly.

808
00:39:59,840 --> 00:40:03,040
So you have architect, you 
really have no choice, right? 

809
00:40:03,040 --> 00:40:06,440
You need to have a network. 
You can still be an individual 

810
00:40:06,920 --> 00:40:10,640
contributor, right? 
I was, you know, I was an ICI 

811
00:40:10,960 --> 00:40:14,280
love being an IC because I can 
focus on my stuff and I don't 

812
00:40:14,280 --> 00:40:17,560
need to deal with managing all 
these vacation plans and all 

813
00:40:17,640 --> 00:40:19,520
this other. 
So simple and just like all 

814
00:40:19,520 --> 00:40:22,560
these performance reviews, like 
right, It's just like I want 

815
00:40:22,560 --> 00:40:23,800
you. 
I want to be happy, but I don't 

816
00:40:23,800 --> 00:40:26,960
want to be the one who's like in
charge of all your happiness, 

817
00:40:27,200 --> 00:40:29,840
right? 
So I love being an IC, but 

818
00:40:29,840 --> 00:40:32,880
heavily networked. 
And I think that's the model 

819
00:40:32,880 --> 00:40:34,360
that you would be be looking 
for. 

820
00:40:34,560 --> 00:40:37,920
So there's a big difference 
between like a manager and where

821
00:40:37,920 --> 00:40:40,720
you have a large team and an 
architect, right? 

822
00:40:40,720 --> 00:40:42,640
Both. 
I think there's three viable 

823
00:40:42,640 --> 00:40:45,240
paths for people, right? 
You can stay an engineer and 

824
00:40:45,240 --> 00:40:48,720
just be a top notch engineer, no
problem whatsoever, right? 

825
00:40:48,960 --> 00:40:51,920
You can become a technical 
manager and that's great because

826
00:40:51,920 --> 00:40:55,280
you understand people skilled, 
sad, you understand motivations.

827
00:40:55,280 --> 00:40:58,360
You're not just a sort of box 
checker like you understand what

828
00:40:58,360 --> 00:41:00,160
people or you can be an 
architect. 

829
00:41:00,160 --> 00:41:04,000
And I would say all three are 
highly valuable and highly 

830
00:41:04,000 --> 00:41:07,200
satisfying career path, you 
know, just with a different 

831
00:41:07,200 --> 00:41:09,880
emphasis. 
From your role as architect, 

832
00:41:10,080 --> 00:41:13,000
you've mentioned a few times one
of your roles is to be a 

833
00:41:13,000 --> 00:41:15,440
multiplier, make other people 
smarter in the room. 

834
00:41:16,040 --> 00:41:19,200
But also you shared a story 
where typically you pick up a 

835
00:41:19,200 --> 00:41:21,440
pen when people have a question 
because that's the point when 

836
00:41:21,440 --> 00:41:24,400
they reach out to you, which 
means that's your value in a 

837
00:41:24,400 --> 00:41:26,560
vacuum. 
If you do that long enough, at 

838
00:41:26,560 --> 00:41:29,520
some point you're not needed 
anymore because you've enabled 

839
00:41:29,520 --> 00:41:32,320
enough people to do that. 
OPS is the same, right? 

840
00:41:32,520 --> 00:41:35,160
If you have great systems, then 
your operations team is going to

841
00:41:35,160 --> 00:41:36,880
do nothing and that is 
fantastic. 

842
00:41:36,880 --> 00:41:38,880
That's like the end goal, yet we
never get there. 

843
00:41:39,240 --> 00:41:41,480
Have you ever gotten yourself in
a position where you were like, 

844
00:41:41,920 --> 00:41:45,080
things are actually good and 
people are kind of practicing 

845
00:41:45,080 --> 00:41:47,000
what I've preached, then it's 
time to move on. 

846
00:41:47,760 --> 00:41:50,600
Well, I think I'm never afraid 
to run out of stuff. 

847
00:41:50,600 --> 00:41:53,600
So some people are hesitant. 
They want to monopolize 

848
00:41:53,960 --> 00:41:56,080
knowledge. 
All right, That's like this 

849
00:41:56,080 --> 00:41:59,200
Oracle kind of powerful 
architect kind of persona. 

850
00:41:59,360 --> 00:42:01,760
Unmissable. 
Yeah, which I totally don't 

851
00:42:01,760 --> 00:42:03,720
like. 
So I've never hesitated to 

852
00:42:03,720 --> 00:42:05,760
share. 
That's why I write books, right?

853
00:42:05,760 --> 00:42:08,360
That's why I read books, right? 
I never hesitated to share 

854
00:42:08,360 --> 00:42:11,680
everything I know. 
And I would say there's only two

855
00:42:11,680 --> 00:42:15,040
outcomes that can really happen.
The one thing is I really have 

856
00:42:15,040 --> 00:42:16,880
nothing to do. 
And then I just make sure nobody

857
00:42:16,880 --> 00:42:20,080
finds out, right? 
I just say, yeah, enjoy, great. 

858
00:42:20,480 --> 00:42:24,160
Everybody's doing that. 
Or more likely what happens is 

859
00:42:24,360 --> 00:42:29,400
more new stuff comes along and I
actually freed up some cycles to

860
00:42:29,400 --> 00:42:32,240
deal with new things. 
And quite honestly, I'm not that

861
00:42:32,240 --> 00:42:35,000
good at that. 
So I can definitely get better 

862
00:42:35,000 --> 00:42:38,240
at pushing more things into into
other folks. 

863
00:42:38,240 --> 00:42:39,600
For me, that would be the dream,
right? 

864
00:42:39,600 --> 00:42:43,640
If I uplift everybody to the 
point where they don't need me 

865
00:42:43,640 --> 00:42:47,120
for that anymore, there's a 
million other things that that I

866
00:42:47,120 --> 00:42:50,680
could be doing the the catches 
and large organizations. 

867
00:42:50,960 --> 00:42:53,960
I don't think it's a direction 
that you should be aiming to, 

868
00:42:54,240 --> 00:42:57,200
but I don't think you should 
measure yourself by achieving 

869
00:42:57,200 --> 00:42:59,440
that because that is rare, 
right? 

870
00:42:59,440 --> 00:43:01,920
You will have turnover. 
You have different people, you 

871
00:43:01,920 --> 00:43:06,080
have a large organization. 
So the odds that you can really 

872
00:43:06,080 --> 00:43:10,400
get so much into the ORC that 
everything runs on autopilot, 

873
00:43:11,440 --> 00:43:13,840
they're actually not that great 
either, right? 

874
00:43:13,840 --> 00:43:16,680
It's more a direction ideal to 
to aim for. 

875
00:43:16,920 --> 00:43:18,560
Yeah. 
But if you manage it, then yeah,

876
00:43:18,560 --> 00:43:21,840
just make sure nobody finds out 
and, you know, enjoy, enjoy the 

877
00:43:21,840 --> 00:43:23,040
time. 
You got a good gig then? 

878
00:43:23,720 --> 00:43:25,320
Exactly. 
Hasn't happened to me. 

879
00:43:25,720 --> 00:43:28,440
Gotcha. 
Yeah, we touched on hard skills.

880
00:43:28,480 --> 00:43:31,680
We also touched on soft skills. 
And one of my final thoughts was

881
00:43:32,200 --> 00:43:34,400
you're dealing with a lot of 
people as architect. 

882
00:43:34,400 --> 00:43:36,960
You already mentioned being an 
architect in a vacuum, not 

883
00:43:36,960 --> 00:43:38,920
talking to anyone. 
You're never going to be as 

884
00:43:38,920 --> 00:43:42,160
effective as you need to be, 
which also means that you need 

885
00:43:42,160 --> 00:43:45,840
to figure out kind of the 
internals of your organization, 

886
00:43:45,840 --> 00:43:49,760
the hierarchy, the politics, 
knowing when to fight and when 

887
00:43:49,760 --> 00:43:52,480
to actually be like, no, we'll, 
we'll let this one slip. 

888
00:43:53,000 --> 00:43:54,920
And that for me has always been 
an art. 

889
00:43:54,920 --> 00:43:58,360
Like I've only dealt with that 
from an IC role, which you don't

890
00:43:58,360 --> 00:44:01,200
get a lot of influence on or 
like your sphere of influence is

891
00:44:01,200 --> 00:44:03,320
quite small. 
I've dealt with that from a 

892
00:44:03,320 --> 00:44:05,880
product sense. 
You are more responsible, 

893
00:44:05,960 --> 00:44:08,440
dedicated for product. 
So you actually can move a 

894
00:44:08,440 --> 00:44:10,320
little bit more and choose when 
to play. 

895
00:44:11,560 --> 00:44:14,160
And there's also a concept that 
you teach with which is the 

896
00:44:14,160 --> 00:44:17,600
gesture, which might play a role
in actually building up that 

897
00:44:17,600 --> 00:44:20,000
political credit. 
But typically, how do you 

898
00:44:20,000 --> 00:44:23,360
develop those skills to move 
organizations and to get buy in 

899
00:44:23,360 --> 00:44:26,160
from people? 
It's incredibly difficult, 

900
00:44:26,160 --> 00:44:27,880
right? 
And the way, as you already 

901
00:44:27,880 --> 00:44:31,920
touched on, we have some, I have
some mental models and metaphors

902
00:44:32,200 --> 00:44:35,720
that help you be more conscious 
about it, but they're not 

903
00:44:35,720 --> 00:44:38,440
recipes I like. 
They mentioned the gesture, 

904
00:44:38,440 --> 00:44:40,160
right? 
The gesture is the person who is

905
00:44:40,160 --> 00:44:43,880
trusted and tells the truth 
because they have no other 

906
00:44:43,880 --> 00:44:46,200
agenda. 
And that's what a court jester 

907
00:44:46,200 --> 00:44:47,840
does. 
And I think that relates very 

908
00:44:47,840 --> 00:44:51,400
well to architects. 
We have little direct power. 

909
00:44:51,400 --> 00:44:53,760
We don't have the most headcount
in the budget, most budget, the 

910
00:44:53,760 --> 00:44:57,400
biggest teams, but we have high 
influential power and that's 

911
00:44:57,400 --> 00:45:00,520
just like a just and we can be 
trusted because we don't have a 

912
00:45:00,520 --> 00:45:02,080
hidden agenda. 
We're not trying to increase our

913
00:45:02,080 --> 00:45:05,000
budget or headcount to make our 
solutions overly complex. 

914
00:45:05,000 --> 00:45:07,880
So our resume looks looks better
kind of thing. 

915
00:45:07,880 --> 00:45:09,720
All right. 
So I think that's a metaphor I 

916
00:45:09,720 --> 00:45:11,680
use, but always careful with 
people. 

917
00:45:11,680 --> 00:45:15,320
The metaphor is just there to 
make you more conscious about 

918
00:45:15,320 --> 00:45:19,240
navigating this then never 
recipe for for success. 

919
00:45:19,560 --> 00:45:22,120
So I would say the the main 
thing, and you mention other 

920
00:45:22,120 --> 00:45:24,800
metaphor I pitch is the 
political capital that that you 

921
00:45:24,800 --> 00:45:27,040
have. 
So I would say 1 boundary 

922
00:45:27,040 --> 00:45:31,200
condition is like you got to be 
able to sleep at night knowing 

923
00:45:31,200 --> 00:45:34,080
that not everything is a perfect
state because otherwise you will

924
00:45:34,080 --> 00:45:37,120
get very little sleep, right? 
It's just like, it's like, oh, 

925
00:45:37,120 --> 00:45:39,040
the kids didn't clean up the 
room again. 

926
00:45:39,040 --> 00:45:42,120
It's like, whatever, right? 
It's like, it's like, yeah, 

927
00:45:42,160 --> 00:45:43,840
sometimes, yeah, yeah. 
People came to me. 

928
00:45:43,960 --> 00:45:46,520
Oh, did you know these people 
doing something that violates 

929
00:45:46,520 --> 00:45:48,440
the stand? 
And I'm like, what am I going to

930
00:45:48,440 --> 00:45:50,040
do, right? 
I have a cat, right? 

931
00:45:50,040 --> 00:45:52,720
It's like trying the cat, like 
telling the cat what to do. 

932
00:45:52,720 --> 00:45:55,600
It's like you also got to live 
with the fact that the cat will 

933
00:45:55,600 --> 00:45:57,680
not always do what you say. 
Especially. 

934
00:45:57,680 --> 00:46:00,440
Cats, especially the opposite. 
So you're going to be able to 

935
00:46:00,440 --> 00:46:02,040
live with that. 
If not, you're going to have 

936
00:46:02,040 --> 00:46:04,400
some ulcer or some other bad 
disease. 

937
00:46:04,480 --> 00:46:07,480
These or at least insomnia 
you're going to have, right? 

938
00:46:07,840 --> 00:46:11,120
But also another metaphor that 
helps me, we do this in the 

939
00:46:11,120 --> 00:46:15,000
workshop is to understand how 
much political capital you have 

940
00:46:15,280 --> 00:46:20,360
to bring change or, you know, 
challenge the status quo or rock

941
00:46:20,360 --> 00:46:24,040
the boat a little bit, right? 
And I think the the metaphor 

942
00:46:24,040 --> 00:46:27,560
helps you how much how can you 
earn goodwill and trust? 

943
00:46:27,880 --> 00:46:32,040
So by delivering things, keeping
your promises, being supportive,

944
00:46:32,320 --> 00:46:34,440
being transparent about what you
do, right? 

945
00:46:34,640 --> 00:46:39,160
These are all ways to build up 
trust, being fair, being open, 

946
00:46:39,400 --> 00:46:42,360
you know, sharing, bring what 
you do right, You build up 

947
00:46:42,360 --> 00:46:47,320
trust, but then the objective is
not to be the most liked and 

948
00:46:47,320 --> 00:46:49,840
most trusted person in the whole
organization. 

949
00:46:50,080 --> 00:46:53,280
You wanna go and spend some of 
the political capital, right? 

950
00:46:53,280 --> 00:46:56,920
You might be the gesture and 
say, hey, look, your baby, in 

951
00:46:56,920 --> 00:46:59,800
this case the project, he's 
actually a little bit ugly, 

952
00:47:00,000 --> 00:47:02,040
sorry to tell you, right? 
Somebody is going to tell you 

953
00:47:02,040 --> 00:47:04,800
that thing that is a train wreck
in the making. 

954
00:47:04,800 --> 00:47:06,800
And I'm convinced that I'm right
about it. 

955
00:47:06,800 --> 00:47:09,840
And here's here's why, right? 
And that will ruffle some 

956
00:47:09,840 --> 00:47:12,040
feathers, right? 
Some people are going to be, how

957
00:47:12,040 --> 00:47:13,560
can you say this? 
Everything is perfect. 

958
00:47:13,560 --> 00:47:17,480
Look at my project status report
work breakdown structure, right?

959
00:47:17,480 --> 00:47:20,480
It's like, you know, basically 
the implosion is scheduled for 

960
00:47:20,480 --> 00:47:23,400
Q3 and we're only in Q2. 
So everything is still perfect. 

961
00:47:23,440 --> 00:47:25,840
Yeah, it's still good. 
And you're just like, well, you 

962
00:47:25,840 --> 00:47:28,560
know, it's like, I call this the
Wiley Coyote, right, where the 

963
00:47:28,560 --> 00:47:31,280
guy is running off the Cliff and
he's still running until he 

964
00:47:31,280 --> 00:47:34,160
looks down. 
And then as soon as you look, so

965
00:47:34,160 --> 00:47:36,080
you got a crash. 
So there's a lot of projects 

966
00:47:36,080 --> 00:47:37,840
like this. 
If they don't look, they don't 

967
00:47:37,840 --> 00:47:40,680
drop kind of thing, right? 
So that will cost you some 

968
00:47:40,680 --> 00:47:43,680
political capital. 
And again, the model doesn't 

969
00:47:43,680 --> 00:47:46,240
give you the perfect answer. 
It doesn't say, oh, today is 

970
00:47:46,240 --> 00:47:48,480
Tuesday, today I should rock the
boat kind of thing. 

971
00:47:48,720 --> 00:47:52,080
That's not how it works. 
But it gives you a mental model 

972
00:47:52,080 --> 00:47:56,840
to not overspend political 
capital, especially early on, 

973
00:47:57,400 --> 00:47:59,960
right? 
You need to 1st earn, right. 

974
00:47:59,960 --> 00:48:01,320
You need to build up 
credibility. 

975
00:48:01,320 --> 00:48:03,520
People think like, oh, I'm the 
smartest guy here, I know 

976
00:48:03,520 --> 00:48:06,200
everything, so I can go and tell
everybody that is wrong, that is

977
00:48:06,200 --> 00:48:08,840
wrong, that is wrong and that 
project sucks, right, right. 

978
00:48:08,840 --> 00:48:11,560
Not going to not going to end 
well for you. 

979
00:48:11,560 --> 00:48:15,840
So the metaphor reminds you 
don't overspend, right? 

980
00:48:15,840 --> 00:48:18,880
I always tell people, you know, 
even your manager can give you a

981
00:48:18,880 --> 00:48:21,560
little line of credit. 
They can back you up a little 

982
00:48:21,560 --> 00:48:23,400
bit. 
So having friends in high levels

983
00:48:23,560 --> 00:48:25,400
helps What? 
Nobody has an infinite line of 

984
00:48:25,400 --> 00:48:27,560
credit. 
You can just run around, you 

985
00:48:27,560 --> 00:48:30,200
know, and call everybody wrong. 
And oh, this should have been 

986
00:48:30,200 --> 00:48:32,560
more loosely coupled and this 
should be cloud native and that 

987
00:48:32,560 --> 00:48:35,360
should be portable. 
It doesn't help anybody. 

988
00:48:35,360 --> 00:48:39,680
So the model of or the metaphor,
political capital reminds you, 

989
00:48:39,680 --> 00:48:44,160
you have to earn before you 
spend and you spend wisely. 

990
00:48:44,760 --> 00:48:48,680
And what I always say meant with
spending wisely means have one 

991
00:48:48,680 --> 00:48:52,320
thing where you really want to 
move the needle and you feel 

992
00:48:52,320 --> 00:48:54,760
it's worth spending some 
political capital on. 

993
00:48:55,040 --> 00:48:57,800
Like calling this big projects 
where they're pouring 10s of 

994
00:48:57,800 --> 00:49:01,360
millions in, calling that ugly 
or telling people that you think

995
00:49:01,360 --> 00:49:04,280
that's not going to deliver. 
That's spending a lot of 

996
00:49:04,280 --> 00:49:08,040
capital, but on one topic. 
And that is much better than 

997
00:49:08,040 --> 00:49:11,040
starting skirmishes everywhere 
and running around telling 

998
00:49:11,040 --> 00:49:13,680
people every project, oh, you 
could do this, you could do 

999
00:49:13,680 --> 00:49:16,240
this, you could do that. 
So that's where the metaphor 

1000
00:49:16,240 --> 00:49:19,280
helps you channel your your 
energy. 

1001
00:49:19,720 --> 00:49:21,720
And that's what I tried to do. 
And yes, sleep. 

1002
00:49:21,960 --> 00:49:25,080
Sleep at night knowing that not 
everything will be perfect. 

1003
00:49:25,240 --> 00:49:28,800
Yeah, I I really like this 
metaphor because even on an 

1004
00:49:28,800 --> 00:49:31,480
architect level, on a product 
level, on an IC software 

1005
00:49:31,480 --> 00:49:33,280
engineer level, everywhere it 
applies, right? 

1006
00:49:33,560 --> 00:49:37,160
I've talked to certain software 
engineers and we barely have a 

1007
00:49:37,160 --> 00:49:39,240
relationship. 
And then I feel like they only 

1008
00:49:39,240 --> 00:49:41,520
complain about everything. 
Everything is ugly or everything

1009
00:49:41,520 --> 00:49:43,200
sucks. 
And I'm like, yeah, we don't 

1010
00:49:43,200 --> 00:49:46,080
really have a like good enough 
relationship for me to be like, 

1011
00:49:46,080 --> 00:49:49,200
I can still trust you or I will 
immediately judge from that 

1012
00:49:49,200 --> 00:49:50,800
perspective. 
If it's some someone that I 

1013
00:49:50,800 --> 00:49:53,040
know, then I know exactly where 
they come from. 

1014
00:49:53,040 --> 00:49:55,400
We have a different perspective.
I have more history with this 

1015
00:49:55,400 --> 00:49:57,800
person to be like, they're not 
negative but they're actually 

1016
00:49:57,800 --> 00:49:59,440
critiquing and their critique is
valid. 

1017
00:49:59,880 --> 00:50:02,600
But if you just do that all the 
time, then people will judge and

1018
00:50:02,600 --> 00:50:04,040
people will label you 
automatically. 

1019
00:50:04,200 --> 00:50:05,840
You become sort of the grumpy 
old, yeah. 

1020
00:50:05,960 --> 00:50:07,160
Person, no one wants to talk to 
you. 

1021
00:50:07,240 --> 00:50:09,080
Yeah, I'm in that demographic, 
so I need to be. 

1022
00:50:09,560 --> 00:50:11,200
You're good. 
Double, double, double. 

1023
00:50:11,200 --> 00:50:12,160
Careful. 
Yeah. 

1024
00:50:12,360 --> 00:50:15,240
But basically everything sucks 
to some degree somehow. 

1025
00:50:15,240 --> 00:50:17,720
Are you right? 
There's never the perfect thing 

1026
00:50:17,960 --> 00:50:19,920
to me. 
There isn't even the notion of 

1027
00:50:19,920 --> 00:50:22,760
the perfect thing. 
I always say architecture isn't 

1028
00:50:22,760 --> 00:50:25,320
good or bad. 
It's not like a Hollywood movie,

1029
00:50:25,320 --> 00:50:28,240
movie or kind of thing. 
The battle between good and 

1030
00:50:28,240 --> 00:50:30,840
evil. 
It's suitable or not suitable, 

1031
00:50:31,080 --> 00:50:33,320
right? 
It does the job or it doesn't do

1032
00:50:33,320 --> 00:50:35,120
the job. 
That means you need to 

1033
00:50:35,120 --> 00:50:37,560
understand what job it needs to 
be doing, right. 

1034
00:50:37,560 --> 00:50:39,760
But then you say, oh, but it 
doesn't do that thing. 

1035
00:50:39,760 --> 00:50:43,080
It's like, well, perhaps we made
a conscious trade off. 

1036
00:50:43,080 --> 00:50:46,880
We valued this over that. 
So it does this and that's what 

1037
00:50:46,880 --> 00:50:48,920
it needs to do. 
And that thing, yeah, that 

1038
00:50:48,920 --> 00:50:51,960
sucks, but that's not what we 
need or that's a trade off that 

1039
00:50:51,960 --> 00:50:55,720
we consciously make. 
You can always throw a Pebble. 

1040
00:50:55,720 --> 00:50:57,680
You can always look at an 
architecture. 

1041
00:50:57,680 --> 00:51:00,320
That's why I'm cautious with 
cautious with architecture 

1042
00:51:00,320 --> 00:51:02,720
reviews. 
Like a third party comes in, 

1043
00:51:03,360 --> 00:51:06,600
generally they come with an 
agenda and they will always find

1044
00:51:06,600 --> 00:51:07,960
something. 
They will always follow the 

1045
00:51:07,960 --> 00:51:09,160
agenda. 
Yeah, yeah, exactly. 

1046
00:51:09,160 --> 00:51:11,360
They always find something 
matching the agenda, like the 

1047
00:51:11,360 --> 00:51:15,000
vendors always, oh, but you 
didn't use our latest product 

1048
00:51:15,200 --> 00:51:17,800
and they will find a very 
plausible case how not using the

1049
00:51:17,800 --> 00:51:20,320
latest product is a horrible 
architecture, right. 

1050
00:51:20,560 --> 00:51:24,040
So I'm always very cautious when
people ask me, it's like, oh, go

1051
00:51:24,040 --> 00:51:28,040
assess this architecture for me.
Like at least 2/3 of the 

1052
00:51:28,040 --> 00:51:31,880
exercise is do they really 
understand the needs and the 

1053
00:51:31,880 --> 00:51:35,600
decisions that they made? 
If the answer is no, well then 

1054
00:51:35,600 --> 00:51:37,920
we have a problem because we 
don't even know whether this 

1055
00:51:37,920 --> 00:51:40,480
architecture can be good or bad 
because we don't even know what 

1056
00:51:40,480 --> 00:51:44,040
we're trying to do, right? 
But if the answer is yes, if 

1057
00:51:44,040 --> 00:51:46,480
they understand the trade-offs, 
they can articulate the 

1058
00:51:46,480 --> 00:51:48,560
trade-offs. 
Oh, this is like this because of

1059
00:51:48,560 --> 00:51:50,960
that, right? 
And here's how we decided this, 

1060
00:51:51,240 --> 00:51:53,840
then I would be hard pressed to 
question that. 

1061
00:51:53,840 --> 00:51:57,280
It's like, well, you understand 
your priorities, you build this 

1062
00:51:57,280 --> 00:52:00,680
to your priorities. 
That means you made a certain 

1063
00:52:00,880 --> 00:52:03,400
trade off, You made a model or 
model list, right? 

1064
00:52:03,400 --> 00:52:05,800
Because the scaling thing wasn't
as important to you. 

1065
00:52:05,800 --> 00:52:09,240
I'm like, that is fine. 
I can't go in and say, oh, the 

1066
00:52:09,240 --> 00:52:11,920
monolith is bad, The micro 
service is good, right? 

1067
00:52:12,160 --> 00:52:13,960
It all depends what you were 
after. 

1068
00:52:13,960 --> 00:52:18,760
So I when I do architecture 
assessment, it's less looking at

1069
00:52:18,760 --> 00:52:22,440
the final product and saying 
whether this is good or bad, but

1070
00:52:22,440 --> 00:52:26,280
more like reviewing your thought
process and understanding 

1071
00:52:26,280 --> 00:52:30,000
whether you made conscious 
decisions and when the you 

1072
00:52:30,000 --> 00:52:33,120
understand the tradeoffs and 
whether those tradeoffs were 

1073
00:52:33,120 --> 00:52:35,080
aligned with what the business 
needed. 

1074
00:52:35,080 --> 00:52:37,880
And if the answer to that is 
yes, well, and the architecture 

1075
00:52:37,880 --> 00:52:42,800
is good because it does what it 
needed to do with sound exactly,

1076
00:52:42,800 --> 00:52:45,280
There's no, I say it's like 
suitable or appropriate. 

1077
00:52:45,280 --> 00:52:48,880
There's no global ranking of 
here's the best architecture in 

1078
00:52:48,880 --> 00:52:51,480
the world and then somehow it 
ranks down to the worst one. 

1079
00:52:51,800 --> 00:52:56,320
It's not a one-dimensional, it's
not a linear kind of space. 

1080
00:52:56,560 --> 00:52:59,880
So I think that is is very 
important to keep in mind when 

1081
00:52:59,880 --> 00:53:02,800
you reviewing someone else's 
architecture. 

1082
00:53:02,800 --> 00:53:06,000
You can always find something 
that sucks in one way or 

1083
00:53:06,000 --> 00:53:08,760
another. 
That's not the exercise, it's 

1084
00:53:08,760 --> 00:53:12,360
understanding the trade-offs. 
Maybe it was quick and easy and 

1085
00:53:12,360 --> 00:53:15,520
cheap to build, right? 
I often remind people, we 

1086
00:53:15,520 --> 00:53:18,200
generally don't like the big 
bottle of mud so much, right? 

1087
00:53:18,200 --> 00:53:20,000
We say, oh, the big bottle of 
mud is bad. 

1088
00:53:20,280 --> 00:53:23,840
Now The funny thing is, I know 
Brian Food and Joyota quite well

1089
00:53:24,240 --> 00:53:28,200
and they're actually unhappy 
that their pattern got labeled 

1090
00:53:28,200 --> 00:53:30,320
into this. 
Yes, they're like, oh, this is 

1091
00:53:30,320 --> 00:53:33,360
bad, don't do it. 
So they actually rewrote the 

1092
00:53:33,360 --> 00:53:38,440
pattern to drive the balance 
more in people because the big 

1093
00:53:38,440 --> 00:53:42,560
ball of mud is quick, cheap and 
requires limited skill set. 

1094
00:53:43,920 --> 00:53:46,720
Are those bad qualities? 
Not necessarily. 

1095
00:53:47,000 --> 00:53:48,880
They're actually very good 
qualities, right? 

1096
00:53:49,040 --> 00:53:52,280
Who doesn't want something that 
cheap and quick and made with 

1097
00:53:52,280 --> 00:53:55,720
limited skill set? 
Now it has downsides in terms of

1098
00:53:55,920 --> 00:53:58,480
shared infrastructure 
maintainability, right? 

1099
00:53:58,480 --> 00:54:03,280
There's a trade off, but if if 
it was completely bad in all 

1100
00:54:03,280 --> 00:54:06,480
regards, we wouldn't have any 
big ball of muds because 

1101
00:54:06,480 --> 00:54:09,120
everybody would say, oh, that is
stupid, That's a bad idea. 

1102
00:54:09,360 --> 00:54:11,760
No, it has some desirable 
qualities. 

1103
00:54:12,000 --> 00:54:15,360
Being quick and cheap is a very 
desirable quality, right? 

1104
00:54:15,600 --> 00:54:19,040
Deliver fast, deliver cheap and 
with a given skill set without 

1105
00:54:19,040 --> 00:54:21,520
sending people to like a nine 
month training kind of thing. 

1106
00:54:21,760 --> 00:54:24,960
That is a great quality, right. 
So it's interesting. 

1107
00:54:24,960 --> 00:54:28,360
It's all about understanding a 
nuances trade off. 

1108
00:54:28,360 --> 00:54:32,760
So even the big ball of mud, you
could easily say all that sucks.

1109
00:54:33,160 --> 00:54:36,160
It might have been exactly what 
they needed at that time. 

1110
00:54:36,160 --> 00:54:39,480
They had a launch deadline of 
like one month that they can't 

1111
00:54:39,480 --> 00:54:42,000
miss. 
They can't hire more people and 

1112
00:54:42,000 --> 00:54:43,880
they gotta get something out of 
the door. 

1113
00:54:43,880 --> 00:54:46,680
What you gonna say? 
Oh, build a giant platform and 

1114
00:54:46,680 --> 00:54:48,480
send everybody to cloud native 
training. 

1115
00:54:48,480 --> 00:54:49,360
Go for it. 
Exactly. 

1116
00:54:49,760 --> 00:54:52,640
The month is gonna be up. 
Be like, well, you hammer that 

1117
00:54:52,640 --> 00:54:55,640
thing together quick, easy, 
cheap, simple, right? 

1118
00:54:55,640 --> 00:54:59,680
Make it work somehow and then 
you sort out might have been the

1119
00:54:59,680 --> 00:55:03,680
best decision. 
So it's so easy to judge and be 

1120
00:55:03,680 --> 00:55:07,280
the smarty pants, but I don't 
think that's going to make you 

1121
00:55:07,560 --> 00:55:09,600
the architect that people want 
to come to. 

1122
00:55:09,840 --> 00:55:12,280
Yeah, right. 
The architect that people want 

1123
00:55:12,280 --> 00:55:15,680
to come to is a person who 
understands the trade-offs, 

1124
00:55:16,080 --> 00:55:18,280
maybe highlights the trade-offs 
in your mind. 

1125
00:55:18,280 --> 00:55:21,120
So the people who might have 
made the big ball of mud might 

1126
00:55:21,120 --> 00:55:23,560
not have thought ahead for 
maintainability. 

1127
00:55:23,560 --> 00:55:26,800
So I think it's totally fair to 
say like, oh, that was great for

1128
00:55:26,800 --> 00:55:29,920
how far you gotten, but what 
does the future look like? 

1129
00:55:30,280 --> 00:55:31,800
Do you understand the next 
requirements? 

1130
00:55:31,800 --> 00:55:33,560
How big does the system have to 
get? 

1131
00:55:34,040 --> 00:55:36,480
What's the expected lifespan of 
the system right? 

1132
00:55:36,760 --> 00:55:40,240
And then you can guide them a 
little bit to what's like maybe 

1133
00:55:40,240 --> 00:55:43,000
a little bit of modularity, 
maybe a little bit of structure 

1134
00:55:43,000 --> 00:55:45,560
isn't all bad, right? 
But this comes out of the 

1135
00:55:45,560 --> 00:55:49,160
dialogue, not out of some 
judgements or recipe. 

1136
00:55:49,160 --> 00:55:52,760
Oh, that is a bad pattern. 
And that is a is a good pattern.

1137
00:55:53,200 --> 00:55:56,440
Doesn't help anyone. 
Now I must say, I take this 

1138
00:55:56,440 --> 00:55:59,760
stance a lot to 1st understand 
and then to actually reason. 

1139
00:55:59,760 --> 00:56:03,200
And for me it's also fun to just
probe for information if there's

1140
00:56:03,200 --> 00:56:06,200
a decision, how well do people 
understand the business context?

1141
00:56:06,200 --> 00:56:09,520
What's the reasoning behind it? 
Like is it actually sound or can

1142
00:56:09,520 --> 00:56:12,160
I poke holes through it? 
And it's not bad, I think to 

1143
00:56:12,160 --> 00:56:14,680
poke holes. 
It's to give perspective and to 

1144
00:56:14,680 --> 00:56:17,960
also do it in a way and to take 
someone with them or with you 

1145
00:56:17,960 --> 00:56:19,920
instead of putting someone down.
Yeah. 

1146
00:56:19,920 --> 00:56:22,440
And coming back full circle sort
of on the architect, the 

1147
00:56:22,440 --> 00:56:25,160
elevator executives do this 
quite a lot. 

1148
00:56:25,520 --> 00:56:29,640
So I always tell people the 
chances that top decision makers

1149
00:56:29,640 --> 00:56:35,240
like board members, CIOSCOOS, 
right, CF OS, the chances that 

1150
00:56:35,240 --> 00:56:39,240
they come and say, oh, there's a
syntax error in your home chart 

1151
00:56:39,240 --> 00:56:42,440
kind of thing or I think your 
Kubernetes node allocation 

1152
00:56:42,440 --> 00:56:46,520
policy should have been X, but 
it was why I'm very low, right? 

1153
00:56:46,960 --> 00:56:51,240
Is this is your domain, right? 
And then their domain, but they 

1154
00:56:51,240 --> 00:56:55,200
have a really good smell for if 
you don't have your story 

1155
00:56:55,200 --> 00:56:57,040
together or your logic together,
right? 

1156
00:56:57,040 --> 00:57:00,800
If you say, oh, Kubernetes, and 
if they say why, it's like, 

1157
00:57:00,800 --> 00:57:02,720
well, what dumb question that 
is. 

1158
00:57:02,720 --> 00:57:04,880
Of course, everything must be 
always in Kubernetes because 

1159
00:57:05,000 --> 00:57:08,280
it's the future and Google used 
it and it's best practice, 

1160
00:57:08,280 --> 00:57:09,760
right? 
And you're like hand waving, 

1161
00:57:09,760 --> 00:57:12,920
hand waving, hand waving. 
Then they will quickly sense 

1162
00:57:12,920 --> 00:57:17,240
that your thought process 
probably has a few little gaps 

1163
00:57:17,440 --> 00:57:20,280
and they're much more likely to 
zoom on this, right? 

1164
00:57:20,280 --> 00:57:23,240
They're less. 
They're very unlikely to like 

1165
00:57:23,240 --> 00:57:27,800
question your actual technical 
decision and acumen, but they're

1166
00:57:27,800 --> 00:57:30,520
very likely to find holes in 
your thought process. 

1167
00:57:30,520 --> 00:57:32,440
Oh, what alternatives have to 
consider? 

1168
00:57:32,440 --> 00:57:35,680
What metrics have to use? 
How do we know what the success 

1169
00:57:35,680 --> 00:57:38,640
looks like, right? 
How much upfront investment does

1170
00:57:38,640 --> 00:57:41,240
this take? 
Could we defer this decision 

1171
00:57:41,240 --> 00:57:42,840
until later? 
Can we start with something 

1172
00:57:42,840 --> 00:57:45,800
simple and then we put it on 
Kubernetes later, right? 

1173
00:57:46,040 --> 00:57:49,720
That's what they're after. 
So having the skill of being 

1174
00:57:49,720 --> 00:57:53,400
sound and being able to 
elaborate your thought process, 

1175
00:57:53,640 --> 00:57:56,920
that actually helps you with 
executives as well. 

1176
00:57:57,080 --> 00:58:00,680
And once they feel like you 
thought this through, you're an 

1177
00:58:00,680 --> 00:58:02,920
expert. 
You, you work through this by 

1178
00:58:02,920 --> 00:58:06,360
reasoning. 
They have very little reason to 

1179
00:58:06,360 --> 00:58:08,800
doubt you because it is your 
domain. 

1180
00:58:09,000 --> 00:58:13,720
But if you cheat, generally they
have, they like dogs who can 

1181
00:58:13,720 --> 00:58:15,280
smell fear kind of things, 
right? 

1182
00:58:15,560 --> 00:58:19,680
They can smell jumps in logic, 
like if you go like, oh, from 

1183
00:58:19,680 --> 00:58:21,880
here to here and they like come 
again. 

1184
00:58:21,880 --> 00:58:27,200
So exactly like exactly. 
They're very good at that 

1185
00:58:27,200 --> 00:58:30,240
because that's where hidden 
assumptions are buried. 

1186
00:58:30,600 --> 00:58:33,200
This is where risks are being 
buried. 

1187
00:58:33,480 --> 00:58:36,840
This is where people reverse 
engineering from the preferred 

1188
00:58:36,840 --> 00:58:39,080
answer, right? 
They know what the answer was, 

1189
00:58:39,080 --> 00:58:41,080
and now they reverse engineer 
guitar process. 

1190
00:58:41,360 --> 00:58:44,120
Exactly. 
They are very trained at 

1191
00:58:44,120 --> 00:58:47,760
figuring these things out. 
And ironically, that's where a 

1192
00:58:47,760 --> 00:58:52,160
lot of engineers stumble when 
they talk to executives because 

1193
00:58:52,160 --> 00:58:54,960
they don't have their reasoning 
straight. 

1194
00:58:55,200 --> 00:58:58,000
And again, that's why I think 
we're having a sounding board, 

1195
00:58:58,120 --> 00:59:01,160
having an architecture sort of 
probes a little bit, right? 

1196
00:59:01,160 --> 00:59:03,040
Does this all make sense? 
Is this sound? 

1197
00:59:03,040 --> 00:59:05,720
Do we understand the traders and
the decisions we make? 

1198
00:59:05,960 --> 00:59:10,560
That is also perfect preparation
to take something to the to the 

1199
00:59:10,560 --> 00:59:13,920
upper floors of the elevator. 
Yeah, I think especially now 

1200
00:59:13,920 --> 00:59:17,280
when it's easier to cheat, have 
gaps in your logic just through 

1201
00:59:17,480 --> 00:59:19,440
content. 
As a generator, you can ask a 

1202
00:59:19,440 --> 00:59:22,320
magical boxer question and the 
answer pops out. 

1203
00:59:22,920 --> 00:59:27,040
You will distinguish yourself if
you do that, sure, but also if 

1204
00:59:27,040 --> 00:59:29,040
you have solid reasoning you can
use whatever tool. 

1205
00:59:29,040 --> 00:59:31,120
In the end, your path to 
whatever answer, it doesn't 

1206
00:59:31,120 --> 00:59:33,480
really matter as long as your 
reasoning is solid. 

1207
00:59:33,760 --> 00:59:35,960
You will distinguish yourself in
those conversations 

1208
00:59:35,960 --> 00:59:38,920
specifically. 
And it's a a huge way, I feel 

1209
00:59:38,920 --> 00:59:41,320
like, to build trust with people
that have limited time. 

1210
00:59:41,640 --> 00:59:42,320
Yeah. 
And I have. 

1211
00:59:42,480 --> 00:59:44,280
I have a very good news. 
I mean, if we're talking about 

1212
00:59:44,280 --> 00:59:48,200
tools like LLMS and stuff, I 
have a very good news about it. 

1213
00:59:48,200 --> 00:59:49,920
Yeah, I looked at some 
architecture documents, 

1214
00:59:49,920 --> 00:59:53,200
marketing growth. 
I asked like 2 questions and I'm

1215
00:59:53,200 --> 00:59:58,160
like, who wrote this? 
And he was like, well, yeah, I 

1216
00:59:58,160 --> 01:00:00,560
use LLM. 
I was like, I'm like, well 

1217
01:00:00,560 --> 01:00:02,640
that's great to use it as a 
tool. 

1218
01:00:02,840 --> 01:00:06,280
But if you paste the output of 
the LLM into your architecture 

1219
01:00:06,280 --> 01:00:09,360
document, you can only lose 
because either it's really good,

1220
01:00:09,520 --> 01:00:12,520
then we don't need you, that's 
not a good outcome for you, or 

1221
01:00:12,520 --> 01:00:14,920
it's not good and it's also not 
good for you, right? 

1222
01:00:15,080 --> 01:00:18,120
You can only lose. 
I always say your starting 

1223
01:00:18,120 --> 01:00:22,240
point, the output of the tool. 
That's your starting point, and 

1224
01:00:22,240 --> 01:00:24,520
you put the. 
Value on top and if you don't do

1225
01:00:24,520 --> 01:00:28,600
that, that's a slippery slope 
that you're not gonna win. 

1226
01:00:28,600 --> 01:00:32,560
And yeah, my nose, my nose is 
extremely good and a lot of 

1227
01:00:32,560 --> 01:00:36,960
other senior IT people, the 
decision makers also these long 

1228
01:00:36,960 --> 01:00:40,960
lists and the fluffy wording on 
this like overconfidence and 

1229
01:00:40,960 --> 01:00:42,960
things like it always shines 
through. 

1230
01:00:42,960 --> 01:00:46,040
I'm like, and then if you're not
sure, you ask people one or two 

1231
01:00:46,040 --> 01:00:49,640
questions, right and then. 
The House of Cards. 

1232
01:00:49,640 --> 01:00:51,640
Exactly. 
The House of Cards falls down. 

1233
01:00:51,640 --> 01:00:55,200
So use the tool. 
But always says make sure the 

1234
01:00:55,200 --> 01:00:58,840
tool works for you, not you work
for the tool, right? 

1235
01:00:58,840 --> 01:01:01,960
Use it as an amplifier of your 
own abilities, not as a 

1236
01:01:01,960 --> 01:01:04,360
substitute. 
If it becomes a substitute, as I

1237
01:01:04,360 --> 01:01:07,600
said, you cannot win, right? 
If it's a great substitute, 

1238
01:01:07,760 --> 01:01:09,520
you're still lost. 
Because we're like, well, why 

1239
01:01:09,520 --> 01:01:10,880
are we paying the architect for 
this, right? 

1240
01:01:11,040 --> 01:01:13,160
We just use the same tool that 
person is using. 

1241
01:01:13,160 --> 01:01:15,280
Right, I love that. 
I'm going to quote that many 

1242
01:01:15,280 --> 01:01:16,360
times. 
So be. 

1243
01:01:16,360 --> 01:01:19,400
Careful, be careful. 
Don't be the tool. 

1244
01:01:19,760 --> 01:01:22,160
I'd be on top of the tool. 
Yeah, absolutely. 

1245
01:01:22,600 --> 01:01:24,920
There's a lot of people 
listening and we've covered the 

1246
01:01:24,920 --> 01:01:28,040
grounds for what architects do, 
but also how to get there, some 

1247
01:01:28,040 --> 01:01:30,920
of the traps and also some of 
the highlight points that makes 

1248
01:01:30,920 --> 01:01:33,080
great architects. 
Before you round off, is there 

1249
01:01:33,080 --> 01:01:34,200
still something you want to 
share? 

1250
01:01:35,000 --> 01:01:36,920
Well, there's probably a lot 
more things we could share, but 

1251
01:01:36,920 --> 01:01:40,080
we are also going to going to 
test people's people's patience,

1252
01:01:40,080 --> 01:01:43,240
right? 
Couple things that I share with 

1253
01:01:43,240 --> 01:01:45,680
people. 
There's a few traps set out for 

1254
01:01:45,680 --> 01:01:48,400
architects. 
So let's say you're you're good 

1255
01:01:48,400 --> 01:01:52,200
at bringing things in the model 
with the sketching, visualizing,

1256
01:01:52,200 --> 01:01:56,280
sharpening people's thinking. 
Sometimes that ends up being or 

1257
01:01:56,280 --> 01:01:59,360
feeling anti climatic. 
So people come to you and it's 

1258
01:01:59,360 --> 01:02:02,680
all unclear in the head and you 
draw a picture and suddenly it 

1259
01:02:02,680 --> 01:02:06,960
all seems easy. 
It sometimes feels like you 

1260
01:02:06,960 --> 01:02:09,080
haven't done a great job because
it's like too easy. 

1261
01:02:09,080 --> 01:02:11,160
Like could it be true? 
Could it be the simple? 

1262
01:02:11,400 --> 01:02:14,560
So that advice I always gave is 
give is don't stumble on the 

1263
01:02:14,560 --> 01:02:17,680
finish line, right? 
If you made sense out of this 

1264
01:02:17,680 --> 01:02:20,960
and suddenly it seems obvious, 
you've done a fantastic job, 

1265
01:02:21,480 --> 01:02:23,920
right? 
We've gotten so in love with 

1266
01:02:23,920 --> 01:02:28,600
complexity that if we actually 
cut through the complexity, we 

1267
01:02:28,600 --> 01:02:31,520
sometimes doubt ourselves. 
We're like, oh, could it be so 

1268
01:02:31,520 --> 01:02:32,560
easy? 
Right. 

1269
01:02:32,720 --> 01:02:34,800
So don't do that. 
That's probably the biggest 

1270
01:02:34,800 --> 01:02:38,440
advice I give is like, don't 
doubt yourself if it all makes 

1271
01:02:38,440 --> 01:02:43,480
sense, it's probably you found 
the right model, the right to 

1272
01:02:43,480 --> 01:02:46,280
picture, and you abstracted away
the right things. 

1273
01:02:46,480 --> 01:02:49,160
You did your sense making and 
now it all makes sense. 

1274
01:02:49,160 --> 01:02:53,200
That's what success looks like. 
So don't doubt that. 

1275
01:02:53,280 --> 01:02:57,800
So that would be stumbling point
#1 and stumbling point #2 is, 

1276
01:02:58,160 --> 01:03:00,040
and we do this in the workshop 
quite a bit. 

1277
01:03:00,760 --> 01:03:04,080
A big part of being an architect
is to to unearth hidden 

1278
01:03:04,080 --> 01:03:06,320
assumptions. 
I like assumptions that are 

1279
01:03:06,320 --> 01:03:09,880
baked into a design that might 
come back later to haunt you. 

1280
01:03:10,120 --> 01:03:11,480
Right. 
Oh, you're assuming this is 

1281
01:03:11,480 --> 01:03:13,760
always there. 
Well, is that necessarily true 

1282
01:03:13,760 --> 01:03:14,440
or not? 
Right. 

1283
01:03:14,840 --> 01:03:18,680
The catch with assumptions is 
once you state them, they're 

1284
01:03:18,680 --> 01:03:21,200
obvious. 
And this is another, these 

1285
01:03:21,200 --> 01:03:24,840
points where you can stumble, 
where you unearth an assumption 

1286
01:03:24,840 --> 01:03:27,360
and you clarify an assumption 
that people were like, oh, that 

1287
01:03:27,360 --> 01:03:29,480
was obvious. 
It's like, well, if it was that 

1288
01:03:29,480 --> 01:03:32,560
obvious, why didn't you? 
Why didn't you state it? 

1289
01:03:32,560 --> 01:03:34,640
So I have some confidence, 
right? 

1290
01:03:34,640 --> 01:03:37,960
There's some skills that you 
have as an architect to tease 

1291
01:03:37,960 --> 01:03:40,640
things out or make things 
obvious. 

1292
01:03:41,080 --> 01:03:45,440
Don't ever let yourself be sold 
undervalue by people saying, oh,

1293
01:03:45,440 --> 01:03:48,880
that's too simple or it's too 
simplistic or that was obvious, 

1294
01:03:48,880 --> 01:03:51,240
right? 
If that was the case, they would

1295
01:03:51,240 --> 01:03:54,320
have come up with it with it 
themselves and you wouldn't have

1296
01:03:54,320 --> 01:03:58,080
to do anything. 
So really understand that this 

1297
01:03:58,080 --> 01:04:03,960
is a highly valuable skill and 
being a catalyst, like making it

1298
01:04:04,480 --> 01:04:07,880
easy for other people to 
articulate things, sketching 

1299
01:04:07,880 --> 01:04:11,080
something out that is actually 
highly valuable. 

1300
01:04:11,080 --> 01:04:14,400
And if the end result is simple,
well, that is even better. 

1301
01:04:14,400 --> 01:04:17,520
That's like the absolute peak 
performance. 

1302
01:04:17,720 --> 01:04:20,960
So don't let people ever tell 
you that, oh, that was easy or, 

1303
01:04:21,120 --> 01:04:23,480
you know, that was obvious. 
No, it wasn't. 

1304
01:04:24,320 --> 01:04:26,440
Gregor, thank you so, so much 
for coming on. 

1305
01:04:26,720 --> 01:04:29,840
I honestly learned a lot and I'm
going to have a joy listening 

1306
01:04:29,840 --> 01:04:31,800
back to this episode. 
Yeah, me too. 

1307
01:04:31,840 --> 01:04:33,320
Actually, I do listen to my 
audience. 

1308
01:04:34,080 --> 01:04:36,520
Yeah, that's a good test, 
because if I don't want to 

1309
01:04:36,520 --> 01:04:38,760
listen to this for like 30 
minutes, you know, I would 

1310
01:04:38,760 --> 01:04:41,320
expect the audience to. 
So I always. 

1311
01:04:41,440 --> 01:04:43,240
But let me know what your 
feedback is going to be. 

1312
01:04:43,320 --> 01:04:45,520
Will do definitely. 
If you're still here with us, 

1313
01:04:45,520 --> 01:04:47,800
let me know in the comments 
section what you think of this 

1314
01:04:47,800 --> 01:04:49,280
episode and we'll see you in the
next one.