1
00:00:00,024 --> 00:00:02,968
What you are getting paid and 
what 
 you pay for is always a 

2
00:00:02,968 --> 00:00:04,589
skill. 
You'll never pay for typing 

3
00:00:04,589 --> 00:00:07,041
speed, do you? 
Today's guest is Ajey Gore, 

4
00:00:07,041 --> 00:00:10,597
former Group 
 CTO of Gojek and 
Operating Partner at Peak XV 

5
00:00:10,597 --> 00:00:14,457
Partners, who scaled Gojek from 
 300,000 to 120 million monthly 

6
00:00:14,457 --> 00:00:16,254
orders. 
Today, he builds and advises 
 

7
00:00:16,262 --> 00:00:18,763
at the intersection of AI, 
infrastructure, and product. 

8
00:00:19,034 --> 00:00:20,783
AI brings a lot of 
 anxiety to 
people today. 

9
00:00:21,053 --> 00:00:23,354
We think we are gonna 
 have no 
job in the future. 

10
00:00:23,413 --> 00:00:25,764
So what does this AI era 
 give 
you perspective? 

11
00:00:25,764 --> 00:00:28,221
So first perspective is 
engineering 
 practices are 

12
00:00:28,221 --> 00:00:30,503
getting reinforced. 
Second thing, age of agents 
 

13
00:00:30,511 --> 00:00:33,458
will reward us as builders. 
We can build things at different

14
00:00:33,458 --> 00:00:35,611
speed. 
The problem is people are trying

15
00:00:35,611 --> 00:00:37,483

 to use AI at a very minimal 
level. 

16
00:00:37,573 --> 00:00:40,354
They're not trying to offload 
 
their job and think more. 

17
00:00:40,613 --> 00:00:43,854
When expertise is becoming 
 
almost free, whether design, 

18
00:00:43,874 --> 00:00:46,983
security, audit, everywhere. 
Then what becomes expensive? 

19
00:00:47,034 --> 00:00:49,164
And my answer is judgment 
 
becomes expensive. 

20
00:00:49,683 --> 00:00:52,243
And judgment layer is gonna be 

very important for everybody. 

21
00:00:52,284 --> 00:00:54,895
Because if agents can do good 
things 
 fast, they'll do bad 

22
00:00:54,895 --> 00:00:57,253
things fast as well. 
For leaders out there, many 
 

23
00:00:57,261 --> 00:01:00,123
are thinking AI means I can 
reduce a lot of people, layoff. 

24
00:01:00,163 --> 00:01:02,463
What is the risk? 
It's not about people reduction.

25
00:01:02,523 --> 00:01:06,005
It's about basic job going away.
Structured work is getting 

26
00:01:06,005 --> 00:01:07,984
replaced. 
Cognitive work is not. 

27
00:01:08,214 --> 00:01:09,404
If they're reducing it, it's 
fine. 

28
00:01:09,514 --> 00:01:11,124
But they're creating cognitive 
gap. 

29
00:01:11,203 --> 00:01:13,633
They're not gonna go faster. 
They're gonna go same speed. 

30
00:01:13,703 --> 00:01:16,153
If they're okay with that 
 
speed, let people go, it's fine.

31
00:01:16,253 --> 00:01:18,441
But if you want to accelerate 
something, 
 then instead of 

32
00:01:18,441 --> 00:01:21,582
letting people go, think through
and enable them with more 

33
00:01:21,582 --> 00:01:23,469
skills. 
Look, 88% of AI companies are 

34
00:01:23,469 --> 00:01:25,023
adopting AI. 
Adoption has increased. 

35
00:01:25,194 --> 00:01:26,933
But only 6% have successful 
projects. 

36
00:01:26,984 --> 00:01:45,233
So why? 
Hey, quick pause. 

37
00:01:45,383 --> 00:01:47,243
My goal with Tech Lead Journal 
is simple. 

38
00:01:47,753 --> 00:01:50,513
Learn from the best in tech, 
 
so we can all grow together. 

39
00:01:50,753 --> 00:01:54,053
If this resonates with you, hit 
 subscribe to follow the 

40
00:01:54,053 --> 00:01:56,027
channel. 
It's the biggest way for you to 

41
00:01:56,027 --> 00:01:58,598
support 
 the show and help us 
keep bringing great guests and 

42
00:01:58,598 --> 00:02:00,863
insights to you. 
Thanks for being here, 
 and 

43
00:02:00,863 --> 00:02:03,563
let's get back to it. 
Hello. 

44
00:02:03,683 --> 00:02:06,043
Welcome to another new episode 

of the in-person podcast 

45
00:02:06,103 --> 00:02:09,299
recording of Tech Lead Journal. 
Today, I'm very excited to meet 

46
00:02:09,299 --> 00:02:12,473
you, Ajey. 
So if you remember, five years 


47
00:02:12,481 --> 00:02:16,883
plus ago, we did an online 
recording, I think episode 23. 

48
00:02:18,244 --> 00:02:20,680
Very excited to have this chance

 to actually talk to you in 

49
00:02:20,680 --> 00:02:21,323
person. 
Welcome. 

50
00:02:21,735 --> 00:02:22,675
Yeah, thank you. 
Thank you so much. 

51
00:02:22,735 --> 00:02:24,616
That was like COVID, I 
 think 
so, if I remember. 

52
00:02:25,275 --> 00:02:27,455
And we were locked in our 
 
rooms and we were like kind of 

53
00:02:27,556 --> 00:02:28,935
recording and trying to do 
things. 

54
00:02:29,075 --> 00:02:32,145
This is much, much better. 
I was like-- And it has been 

55
00:02:32,145 --> 00:02:35,227
always 
 excited, exciting to 
actually revisit what we have 

56
00:02:35,227 --> 00:02:39,195
done in last five years 
 and 
see the perspective change. 

57
00:02:39,235 --> 00:02:40,975
And I think COVID changed many 
things. 

58
00:02:41,416 --> 00:02:46,595
One of the things, if AI has 
 
accelerated the coding, COVID 

59
00:02:46,656 --> 00:02:48,675
has accelerated the 
digitization. 

60
00:02:50,416 --> 00:02:51,839
Yeah. 
And also perspectives of life, I

61
00:02:51,839 --> 00:02:53,097
guess. 
Yes, yes, yes. 

62
00:02:53,967 --> 00:02:57,611
And COVID also created some 
false 
 hope, saying, people who

63
00:02:57,611 --> 00:03:00,035
work remote, can work remote all
the time. 

64
00:03:00,635 --> 00:03:03,615
And then after some time people 
 realize, no, that's not true. 

65
00:03:03,875 --> 00:03:06,315
We need people in person. 
So we are social animals. 

66
00:03:06,336 --> 00:03:09,535
We need interactions, we need 
ideas 
 to go through and all 

67
00:03:09,535 --> 00:03:11,932
this stuff. 
So I think, there are also 

68
00:03:11,932 --> 00:03:15,344
lessons 
 in terms of what you 
can digitize and what you cannot

69
00:03:15,344 --> 00:03:18,470
digitize. 
But that evolution would have 

70
00:03:18,470 --> 00:03:22,411
happened 
 over many, many 
years, but COVID just forced 

71
00:03:22,411 --> 00:03:24,283
that acceleration so fast. 
Yeah. 

72
00:03:25,083 --> 00:03:27,463
And yeah, hence we have this 
 
in-person podcast recording. 

73
00:03:27,583 --> 00:03:31,291
So I think that's one part that 
I feel 
 digitization does not 

74
00:03:31,291 --> 00:03:34,923
give you the same kind of like 
interactions, the richness. 

75
00:03:34,939 --> 00:03:38,241
Yep. 
Yeah, it always is like 
 very 

76
00:03:38,241 --> 00:03:41,620
different in person. 
Also what happens is, I'll tell 

77
00:03:41,620 --> 00:03:43,452
you this. 
When you were-- when I was 
 

78
00:03:43,460 --> 00:03:47,119
meeting so many people on Zoom, 
so you only see still this. 

79
00:03:48,059 --> 00:03:51,119
So you don't know what the 
 
person really looks like. 

80
00:03:51,839 --> 00:03:54,446
And then sometimes what happens,
you 
 have like, you assume the 

81
00:03:54,446 --> 00:03:56,539
person must be very tall, and 
they're not that tall. 

82
00:03:56,579 --> 00:03:58,979
And then sometimes this, and 
 
this, sometimes you assume that 

83
00:03:59,019 --> 00:04:01,824
person won't be that tall, but 

they're very tall or whatever 

84
00:04:01,824 --> 00:04:03,163
way. 
So you get always surprised, 
 

85
00:04:03,171 --> 00:04:06,299
"Oh, I did not think like that."
Has that happened to you at all?

86
00:04:06,380 --> 00:04:07,359
Yeah, for sure. 
For sure. 

87
00:04:07,560 --> 00:04:09,713
Okay, cool. 
So it's n- I'm not the 
 only 

88
00:04:09,713 --> 00:04:11,130
one who does that. 
Yeah. 

89
00:04:11,780 --> 00:04:16,541
So I think I've seen, you know, 
your posts, your blogs, and all 

90
00:04:16,541 --> 00:04:18,728
that. 
There's a little bit of change 


91
00:04:18,737 --> 00:04:21,861
in terms of life, right? 
I think maybe like three months 

92
00:04:21,861 --> 00:04:25,879
 plus ago you stopped your last 
role in Peak XV, you know? 

93
00:04:26,579 --> 00:04:30,579
So tell us maybe what have you 

been up to, what's being… Like, 

94
00:04:30,739 --> 00:04:36,562
what's keep you busy these day? 
So, like I wrote in that blog 
 

95
00:04:36,570 --> 00:04:40,764
post, like sometimes you run so 
much or you walk so much, you 
 

96
00:04:40,772 --> 00:04:42,383
talk so much or you work so 
much. 

97
00:04:44,264 --> 00:04:47,903
You assume, you accept that 
 is
a new normal, correct? 

98
00:04:48,323 --> 00:04:50,843
And then you don't know 
 when 
to stop, what to do. 

99
00:04:51,123 --> 00:04:54,035
You feel you don't understand, 
oh, 
 I have been running all 

100
00:04:54,035 --> 00:04:56,243
the time. 
What it feels like to walk? 

101
00:04:57,203 --> 00:04:59,733
What that slow pace means? 
What does slow thinking mean? 

102
00:04:59,943 --> 00:05:03,543
What does being slow and 
 
having time mean, right? 

103
00:05:04,023 --> 00:05:08,123
So- and there are multiple, 
there 
 are multiple reasons I 

104
00:05:08,123 --> 00:05:10,449
did that. 
For example, there were some 

105
00:05:10,449 --> 00:05:13,083
things 
 around family which I 
want to focus on. 

106
00:05:13,203 --> 00:05:15,443
There are some things around 
 
myself which I want to focus on.

107
00:05:15,923 --> 00:05:18,837
And one of the thoughts which 
also 
 came across is like we 

108
00:05:18,837 --> 00:05:22,523
work a lot. 
All of us earn money we work for

109
00:05:22,523 --> 00:05:25,043

 many, many years, and we spend
money. 

110
00:05:25,173 --> 00:05:29,164
But we forget that we earn 
 
time as well with that money. 

111
00:05:29,644 --> 00:05:33,983
And we end up spending time 
 
most probably at the end of when

112
00:05:33,983 --> 00:05:36,083
somebody calls retirement. 
I don't believe in retirement. 

113
00:05:36,304 --> 00:05:38,524
But yeah, I'll do that. 
I'll chill at sixty-five. 

114
00:05:39,403 --> 00:05:42,133
Yeah, but that is something like

 we are doing it when you have 

115
00:05:42,133 --> 00:05:45,444
money and time, but you don't 
have health. 

116
00:05:45,634 --> 00:05:48,909
So I wanted to bring that 
perspective, 
 and I'm fortunate

117
00:05:48,909 --> 00:05:52,196
that I can do that. 
It's very difficult to 
 do, I 

118
00:05:52,196 --> 00:05:54,104
can tell you that. 
It's very difficult to do. 

119
00:05:54,204 --> 00:05:57,604
There are some external forces, 
 there are some internal forces.

120
00:05:57,644 --> 00:06:00,924
But sometime it just feels 
 
like, let's do, let's pause. 

121
00:06:01,184 --> 00:06:02,884
Let's do some tinkering. 
Let's do something else. 

122
00:06:02,984 --> 00:06:06,215
Let's play those things which 
you 
 have been willing and 

123
00:06:06,215 --> 00:06:08,964
wishing to play for many, many, 
many days, right? 

124
00:06:09,544 --> 00:06:14,364
So that's what I'm doing now. 
I am going after my passion. 

125
00:06:15,064 --> 00:06:20,683
I have-- I am building things. 
And then I had, I feel very 
 

126
00:06:20,691 --> 00:06:25,774
much associated with the youth 
mental health and other issues. 

127
00:06:26,124 --> 00:06:28,924
So I'm volunteering with 
 the 
helloally or Ally. 

128
00:06:29,424 --> 00:06:33,595
Ally is a Indian not-for-profit 
which 
 is building mental 

129
00:06:33,595 --> 00:06:37,204
health solutions to train 
therapists, teachers, and 

130
00:06:37,204 --> 00:06:41,524
related 
 people to identify and
help youth, right? 

131
00:06:41,944 --> 00:06:44,456
And that is very good for us. 
And everybody should actually 
 

132
00:06:44,464 --> 00:06:46,424
do something like that. 
So I'm doing that. 

133
00:06:46,484 --> 00:06:48,264
I'm doing that volun- 
 that is 
my volunteer work. 

134
00:06:48,714 --> 00:06:51,024
And then I'm helping few 
organizations. 

135
00:06:51,084 --> 00:06:55,984
Like for-- I just accepted being

 advisor to NUS Enterprise, 

136
00:06:55,984 --> 00:07:00,226
because I was excited by their 
mission of 
 doing things around

137
00:07:00,226 --> 00:07:02,734
data and AI. 
So I'm doing that. 

138
00:07:03,184 --> 00:07:07,444
And then on the other hand, 
 we
are just going through this 

139
00:07:07,484 --> 00:07:12,329
craze of bots and this and that.
And I have been tinkering around

140
00:07:12,329 --> 00:07:15,559
that. 
We built ClawStation just for 

141
00:07:15,559 --> 00:07:18,104
sake of it. 
It is okay. 

142
00:07:18,124 --> 00:07:20,014
It's side hustle. 
If it takes off, it takes off. 

143
00:07:20,014 --> 00:07:22,044
If it doesn't take off, I'll go 
 build, tinker something else. 

144
00:07:22,424 --> 00:07:24,764
So yes. 
And reading book. 

145
00:07:24,824 --> 00:07:27,450
And as you know that you-- when 
you 
 called me saying, "Ajey, 

146
00:07:27,450 --> 00:07:29,134
can we do this?" 
And I go, "Yeah, I can do this. 

147
00:07:29,194 --> 00:07:30,415
I have time. 
I'm spending time." 

148
00:07:30,415 --> 00:07:31,984
So I'm spending 
 time with 
friends like you. 

149
00:07:32,284 --> 00:07:33,564
So yeah, that's what is going 
on. 

150
00:07:33,584 --> 00:07:36,524
It's like pretty exciting 
 
journey in that sense. 

151
00:07:37,064 --> 00:07:42,004
Because what is happening is 
that you 
 have-- you create 

152
00:07:42,004 --> 00:07:47,007
slack in your life. 
And then slack allows 
 you to 

153
00:07:47,007 --> 00:07:50,537
think through. 
When you meet people without 
 

154
00:07:50,545 --> 00:07:54,027
any agenda, then you have 
different flow of conversations.

155
00:07:54,847 --> 00:07:58,487
They go in many directions and 

then you come out mostly happy. 

156
00:07:58,587 --> 00:08:01,043
"You know, I felt good meeting 
this 
 person." 

157
00:08:01,043 --> 00:08:03,967
And then we talk about -- 
because we don't have any 

158
00:08:03,967 --> 00:08:05,815
agenda, we don't 
 have any 
transaction scheduled, we don't 

159
00:08:05,815 --> 00:08:10,227
have anything to go after or 
 
before that -- then the fluidity

160
00:08:10,307 --> 00:08:12,687
of conversation is very 
positive. 

161
00:08:13,607 --> 00:08:16,723
So that is what is happening. 
And then a lot of people are 

162
00:08:16,723 --> 00:08:18,747
reaching 
 out to me, asking me 
for advice. 

163
00:08:20,087 --> 00:08:24,542
They are in mid-career stuff. 
They are going through this AI 


164
00:08:24,550 --> 00:08:28,187
stuff, and they're like scared. 
Some of them are excited. 

165
00:08:28,968 --> 00:08:33,432
So I'm telling them what we 
should not 
 do, because I can't

166
00:08:33,432 --> 00:08:37,167
tell you what to do. 
But I can tell you what not to 

167
00:08:37,167 --> 00:08:40,811
do 
 from my experience, which 
actually kind of reduces the 

168
00:08:40,811 --> 00:08:43,107
possibility of 
 going and doing
something stupid. 

169
00:08:44,563 --> 00:08:48,335
Yeah. 
So I think you brought a very 
 

170
00:08:48,343 --> 00:08:50,972
interesting perspective, right? 
Because if I look at your 

171
00:08:50,972 --> 00:08:53,544
career, you've 
 been kind of 
like hustling or let's call it 

172
00:08:53,544 --> 00:08:56,244
hustling, like working hard, you
know, 
 from ThoughtWorks, you 

173
00:08:56,244 --> 00:08:59,360
know, spent 10 years, GO-JEK 
maybe five years, Sequoia, 
 

174
00:08:59,368 --> 00:09:01,824
Peak XV is about another five 
years. 

175
00:09:02,123 --> 00:09:03,519
Yeah. 
So you've been like, kind of 

176
00:09:03,519 --> 00:09:05,323
like 
 running in full mode all 
these years. 

177
00:09:05,933 --> 00:09:08,827
And I think when you mention 
about 
 earning time, spending 

178
00:09:08,827 --> 00:09:12,413
time, I think it's in everyone's
interest now to actually 
 

179
00:09:12,421 --> 00:09:14,783
understand what does the concept
mean. 

180
00:09:15,203 --> 00:09:17,934
So maybe if you can elaborate 
for 
 people, especially 

181
00:09:17,934 --> 00:09:20,903
mid-career and, you know, later 
stage of career people. 

182
00:09:21,143 --> 00:09:24,023
Actually sometimes we don't 
realize 
 that actually we miss 

183
00:09:24,023 --> 00:09:26,599
this opportunity to create slack
time, spend time 
 on our 

184
00:09:26,599 --> 00:09:29,797
passion and all that. 
Yeah, so I wrote this 
 on my 

185
00:09:29,797 --> 00:09:33,963
blog, the Space Between 
Chapters. 

186
00:09:33,973 --> 00:09:36,416
So I'm living in the space 
 
between the chapters. 

187
00:09:37,276 --> 00:09:39,703
One chapter is over and next 
chapter is 
 supposed to start 

188
00:09:39,703 --> 00:09:44,348
at some point of time. 
And I found this space, or I've 

189
00:09:44,348 --> 00:09:48,293
 kind of created this space and 
I'm going to live through that 

190
00:09:48,293 --> 00:09:51,516
space. 
Lot of time, what happens is 

191
00:09:51,516 --> 00:09:55,475
people don't 
 believe that they
can create that space. 

192
00:09:56,666 --> 00:10:00,254
That is first thing. 
If you don't believe that I 
 

193
00:10:00,262 --> 00:10:03,016
can create that space, then you 
will not be able to do it. 

194
00:10:04,336 --> 00:10:08,136
And I started that blog saying 

someti-- the most courageous 

195
00:10:08,296 --> 00:10:13,696
thing is to do stop running. 
It's like just stop running. 

196
00:10:13,976 --> 00:10:16,548
That's the most courageous thing
to 
 do because there are a lot 

197
00:10:16,548 --> 00:10:18,996
of things which we have 
preconceived notion. 

198
00:10:19,036 --> 00:10:23,180
We worry about a lot of things 
which 
 have-- which are not yet

199
00:10:23,180 --> 00:10:25,756
come, and we ponder and regret 
about a lot of 
 things which 

200
00:10:25,756 --> 00:10:30,221
have gone in the past. 
It's so difficult to implement 


201
00:10:30,229 --> 00:10:34,721
that whatever happening in 
present, can you enjoy that, 

202
00:10:34,721 --> 00:10:37,712
right? 
And I'll tell you, Henry, 
 last

203
00:10:37,712 --> 00:10:40,196
three months, it's like 
February, March, April is gone. 

204
00:10:40,356 --> 00:10:42,136
We are sitting on the 
 
thirtieth April over here. 

205
00:10:43,856 --> 00:10:46,826
And it's so funny that three 
months 
 or ninety days are 

206
00:10:46,826 --> 00:10:50,349
gone, and I'm like, I am still 
like busy and 
 meeting people, 

207
00:10:50,349 --> 00:10:52,897
and I'm enjoying it. 
I don't know where those 
 three

208
00:10:52,897 --> 00:10:56,180
months are gone. 
So first thing is that I spent a

209
00:10:56,180 --> 00:10:58,796
lot 
 of time with my friends 
and family. 

210
00:11:00,876 --> 00:11:04,236
I took my kids to bouldering, 
which 
 I never thought I would 

211
00:11:04,236 --> 00:11:06,296
do, right? 
They climbed the thing and all 


212
00:11:06,304 --> 00:11:10,156
the stuff which is amazing. 
And then I'm again going to 

213
00:11:10,156 --> 00:11:13,036
spend 
 a week with them just 
like that. 

214
00:11:13,716 --> 00:11:19,284
And then what happens is the 
slack allows 
 me to think, and 

215
00:11:19,284 --> 00:11:25,000
that is very important. 
So I will try to give you the 

216
00:11:25,000 --> 00:11:27,416
analogy 
 of getting up late and
getting early. 

217
00:11:28,496 --> 00:11:30,656
So if you get up late in the 
morning, 
 what happens is, 

218
00:11:30,656 --> 00:11:33,056
suppose you have to go to office
at, like, eight thirty, nine. 

219
00:11:33,716 --> 00:11:35,656
And you get up at eight fifteen,

 then what are you doing? 

220
00:11:35,676 --> 00:11:37,756
The first thing is you 
 are 
reacting to the time. 

221
00:11:38,486 --> 00:11:41,186
So you're hurry, hurry, hurry, 

and then you've-- somehow you 

222
00:11:41,186 --> 00:11:44,392
get into MRT or car or whatever 
and 
 get to your office by, 

223
00:11:44,392 --> 00:11:48,116
like nine. 
And then whole day is going on 

224
00:11:48,116 --> 00:11:50,696
reactions 
 to reactions to 
reactions to reactions, and end 

225
00:11:50,696 --> 00:11:54,407
of day you come home, and then 

you actually just do default 

226
00:11:54,407 --> 00:11:57,694
things. 
But now just flip this 
 and 

227
00:11:57,694 --> 00:12:02,244
say, I get up at five. 
Now I can think and deliberate 

228
00:12:02,244 --> 00:12:04,652
on 
 things and I can create a 
slack. 

229
00:12:05,272 --> 00:12:10,656
Yes, I had to sleep early at 
ten, but that 
 slack is way 

230
00:12:10,656 --> 00:12:13,512
more important than actually 
working like a night owl till 

231
00:12:13,512 --> 00:12:15,677
two o'clock 
 and then getting 
out at eight fifteen and do… 

232
00:12:15,677 --> 00:12:19,817
actually, and we don't work. 
We actually spend time during 

233
00:12:19,817 --> 00:12:22,592
work 
 hours, but not useful 
spend time. 

234
00:12:23,032 --> 00:12:24,852
I'm not saying that 
 don't go 
on social media. 

235
00:12:24,872 --> 00:12:26,312
I'm not saying that don't watch 
movies. 

236
00:12:26,452 --> 00:12:29,489
I want to watch movies. 
But what I'm trying to say is 
 

237
00:12:29,497 --> 00:12:33,142
create a slack in the system. 
And it is very true for creating

238
00:12:33,142 --> 00:12:35,612

 the slack, not in your 
personal life, but within your 

239
00:12:35,612 --> 00:12:37,652
teams as well, 
 else innovation
will not happen. 

240
00:12:38,192 --> 00:12:41,512
Lot of engineering teams I 
 
see, they always react, right? 

241
00:12:42,712 --> 00:12:45,764
Now, second thing. 
Many people will say, "Ajey, you

242
00:12:45,764 --> 00:12:49,210

 are saying it because you can 
create slack, and you have 

243
00:12:49,210 --> 00:12:51,652
money, 
 and you are financially
secure." 

244
00:12:52,212 --> 00:12:55,522
I don't think so it's like that.
There are many, many examples, 

245
00:12:55,522 --> 00:12:59,610
and I have 
 seen many people 
who don't have a lot of money, 

246
00:12:59,610 --> 00:13:04,712
but they have done this as well.
I think I watched a TEDx talk. 

247
00:13:04,792 --> 00:13:06,412
I don't remember the talk, 
 but
I watched a TEDx talk. 

248
00:13:06,412 --> 00:13:13,406
Somebody was in Thailand, in 
Bangkok, and he 
 actually, I 

249
00:13:13,406 --> 00:13:17,477
think the name was Danai, and he
actually went back to his 

250
00:13:17,477 --> 00:13:20,393
village. 
And now they-- he has, runs a 

251
00:13:20,393 --> 00:13:23,266
very 
 big, large village where 
he like… creating a house is 

252
00:13:23,266 --> 00:13:26,211
very simple. 
You just dig up the soil and 

253
00:13:26,211 --> 00:13:29,172
build a 
 mud wall, and you have
a house, right? 

254
00:13:29,412 --> 00:13:31,511
You don't need to do that. 
But if you have to live in 

255
00:13:31,511 --> 00:13:33,392
Bangkok, he 
 has to spend so 
much money or Hong Kong. 

256
00:13:33,472 --> 00:13:35,402
I don't remember. 
The point is he-- I watched that

257
00:13:35,402 --> 00:13:38,627
talk 
 around ten years back and
I was like, "This is amazing." 

258
00:13:38,627 --> 00:13:43,492
So a lot of time we 
 have this 
preconceived notion of about our

259
00:13:43,532 --> 00:13:46,848
work and job security and all 
the stuff, 
 and that's where I 

260
00:13:46,848 --> 00:13:49,004
go to my principle. 
If-- When I talked to you last 


261
00:13:49,012 --> 00:13:51,904
time, I think I talked about it,
being uncomfortable is one of 

262
00:13:51,904 --> 00:13:53,572
the 
 most important thing for 
anybody. 

263
00:13:54,752 --> 00:13:58,279
If you be uncomfortable in your 
job then 
 you will be able to 

264
00:13:58,279 --> 00:14:00,796
take new challenges. 
Second, be replace-- being 
 

265
00:14:00,804 --> 00:14:03,632
replaceable is also one of the 
most important thing in our job.

266
00:14:03,692 --> 00:14:06,160
Because if you can be 
replaceable, 
 then you can do 

267
00:14:06,160 --> 00:14:07,792
something new and you can be 
uncomfortable. 

268
00:14:07,832 --> 00:14:10,332
And if you be uncomfortable, 
 
you will acquire more skills. 

269
00:14:11,072 --> 00:14:15,263
That will eventually lead you to

 take you out of one important 

270
00:14:15,263 --> 00:14:17,356
fear. 
Will I get next job or will I 
 

271
00:14:17,364 --> 00:14:20,602
have something to do or not? 
If you have done enough many 

272
00:14:20,602 --> 00:14:22,712
things, then 
 you will be able 
to do something, right? 

273
00:14:23,972 --> 00:14:29,047
And that is very important. 
The way in continuous delivery, 

274
00:14:29,047 --> 00:14:34,536
or 
 continuous integration, the
only way we can do that, like 

275
00:14:34,536 --> 00:14:37,662
one-click deployment, 
 is 
because we have test harness 

276
00:14:37,662 --> 00:14:40,972
that we know our code will not 
fail, and 
 it fails, we'll know

277
00:14:40,972 --> 00:14:42,872
about it before it actually we 
can click the button. 

278
00:14:42,912 --> 00:14:44,292
The button will not get enabled,
right? 

279
00:14:46,162 --> 00:14:48,855
That is a harness which allows, 
gives 
 us confidence that our 

280
00:14:48,855 --> 00:14:51,448
code will work. 
This is also the harness which 

281
00:14:51,448 --> 00:14:53,648
will give 
 us the confidence 
that we'll get a job. 

282
00:14:55,207 --> 00:14:58,628
I can do this today because 
 
somewhere I have a little bit of

283
00:14:58,688 --> 00:15:03,077
confidence that I can get a job.
Somewhere I have confidence that

284
00:15:03,077 --> 00:15:04,427

 I'm useful for somebody, 
right? 

285
00:15:04,907 --> 00:15:07,487
For various reasons, various 
 
experiences and all stuff. 

286
00:15:07,967 --> 00:15:13,062
People do value experience. 
You must have heard this joke 

287
00:15:13,062 --> 00:15:17,415
where 
 there's somebody was 
called for some electric or some

288
00:15:17,415 --> 00:15:21,241
computer error. 
And he came and, like I think 

289
00:15:21,241 --> 00:15:25,766
took a 
 hammer and str- s- like
put it on some right place and 

290
00:15:25,766 --> 00:15:28,405
the thing start working. 
And somebody asked like, "So how

291
00:15:28,405 --> 00:15:31,171
much 
 should we pay you for?" 
He's like, "A thousand dollars."

292
00:15:31,171 --> 00:15:32,815
And like, "Why 
 a thousand 
dollars?" 

293
00:15:32,815 --> 00:15:35,877
Like, "No, put- striking with 
hammer is like $1. 

294
00:15:36,127 --> 00:15:39,577
The rest is to know 
 where to 
strike." 

295
00:15:39,577 --> 00:15:41,951
Right? 
My point is people should 

296
00:15:41,951 --> 00:15:44,622
experience, 
 people should 
gather those experience, and if 

297
00:15:44,622 --> 00:15:46,351
they do, they gather that 
 
experience, they'll become 

298
00:15:46,351 --> 00:15:47,676
valuable. 
If they become valuable, 
 then 

299
00:15:47,676 --> 00:15:49,733
they'll be useful. 
If they are useful, then they 

300
00:15:49,733 --> 00:15:52,251
can 
 create slack, and they can
create a space between the 

301
00:15:52,251 --> 00:15:53,015
chapters. 
Yeah. 

302
00:15:53,675 --> 00:15:56,192
So I think the importance of 
creating 
 slack, you know, 

303
00:15:56,192 --> 00:15:58,685
sometimes we see it in the 
literature, right, productivity 

304
00:15:58,685 --> 00:16:01,165
 and creativity and all that. 
Definitely creating the slack 
 

305
00:16:01,173 --> 00:16:03,795
is very important I feel. 
And sometimes it doesn't have 
 

306
00:16:03,803 --> 00:16:07,314
to be not having a job, right? 
You can create slack without -- 

307
00:16:07,314 --> 00:16:11,015
for 
 example, like maybe go for
running for a long time or, you 

308
00:16:11,015 --> 00:16:13,806
know, you meditate or, 
 you 
know, just do something that 

309
00:16:13,806 --> 00:16:17,963
doesn't really fully engage you 
in terms of, 
 you know, work or

310
00:16:17,963 --> 00:16:21,064
career and all that. 
Sometimes that could let 
 your 

311
00:16:21,064 --> 00:16:23,393
mind wander and… Oh, I'll tell 
you this. 

312
00:16:23,723 --> 00:16:28,699
One other hack I have is go for 
walks, 
 runs, but without any 

313
00:16:28,699 --> 00:16:31,003
digital stuff with you, not even
with your phone. 

314
00:16:31,943 --> 00:16:35,769
Now you are forced… But the, 
 
another problem is like a lot of

315
00:16:35,769 --> 00:16:37,844
time people want to record this 
on Strava. 

316
00:16:37,903 --> 00:16:40,107
I don't know why. 
I do. 

317
00:16:40,167 --> 00:16:41,303
I do that. 
I know. 

318
00:16:41,663 --> 00:16:43,184
So my point is like, yeah, it's 
okay. 

319
00:16:43,983 --> 00:16:46,184
Let it go for one day 
 and just
go for a walk. 

320
00:16:46,223 --> 00:16:47,764
Then you'll suddenly realize 
 
there's so much of world. 

321
00:16:48,196 --> 00:16:53,291
Yeah I go out for a walk without
my phone, 
 and I come back in 

322
00:16:53,291 --> 00:16:56,284
like one hour or two hour, and 
it's like amazing. 

323
00:16:56,703 --> 00:16:59,424
Like I can stop at some hawker 

center on the way and eat some 

324
00:16:59,483 --> 00:17:02,383
food, have some coffee, or I can

 do something else and come 

325
00:17:02,383 --> 00:17:04,723
back. 
And you're absolutely right. 

326
00:17:05,083 --> 00:17:07,624
People can create slack in 
 
their life just like that. 

327
00:17:07,782 --> 00:17:11,396
Very simple. 
See, I-- when I was in Gojek, 

328
00:17:11,396 --> 00:17:13,784
somebody 
 asked me, "How did 
you scale Gojek?" 

329
00:17:13,843 --> 00:17:17,685
I go, "Okay, the only way we 
could 
 scale Gojek was 

330
00:17:17,685 --> 00:17:19,887
discipline." 
Discipline brings the scale, and

331
00:17:19,887 --> 00:17:21,963
actually 
 scale also brings the
discipline. 

332
00:17:22,303 --> 00:17:24,663
If you want to scale, you 
 
can't be indisciplined. 

333
00:17:25,163 --> 00:17:27,373
And if you're at scale, then 
 
you have to be disciplined. 

334
00:17:27,613 --> 00:17:29,063
So my point is just be 
disciplined. 

335
00:17:29,063 --> 00:17:34,148
Say I will do this once a week 
and that's 
 okay to be 

336
00:17:34,148 --> 00:17:37,392
disconnected for once a week. 
If you can't disconnect for once

337
00:17:37,392 --> 00:17:39,464
a week 
 for two hours, then 
there is a problem. 

338
00:17:39,504 --> 00:17:40,639
But I agree with you. 
Yeah. 

339
00:17:41,159 --> 00:17:44,112
Like for me, I've been doing a 
lot 
 of running as well lately,

340
00:17:44,112 --> 00:17:47,524
like maybe three times a week, 
like an 
 hour, more than that 

341
00:17:47,524 --> 00:17:50,880
sometimes. 
I think it gives a lot of 
 

342
00:17:50,888 --> 00:17:53,459
benefit, especially sometimes 
it's just creativity flowing. 

343
00:17:53,660 --> 00:17:53,949
Yeah. 
Yeah. 

344
00:17:54,099 --> 00:17:57,079
Because the interconnectedness 

with things that you read, maybe

345
00:17:57,120 --> 00:18:00,795
things that you didn't think 
about, 
 somehow it blends 

346
00:18:00,795 --> 00:18:02,183
during the session. 
Yeah. 

347
00:18:02,943 --> 00:18:08,187
And you think about many things,

 which you wouldn't think about

348
00:18:08,187 --> 00:18:11,523
if you are engaged with your 
devices. 

349
00:18:12,183 --> 00:18:14,660
Yeah. 
So one thing that you mentioned 

350
00:18:14,660 --> 00:18:16,915
about 
 earning time, spending 
time, I think that's also very 

351
00:18:16,915 --> 00:18:18,826
important, right? 
Because we all feel that we 
 

352
00:18:18,834 --> 00:18:22,284
have to work hard during young. 
This is probably also 
 true for

353
00:18:22,284 --> 00:18:24,985
Asian, right? 
Our teaching is that, yeah, when

354
00:18:24,985 --> 00:18:27,847

 you're young, work hard, earn 
a lot, save it, and then when 

355
00:18:27,847 --> 00:18:29,623
you 
 retire, you kinda like 
spend that. 

356
00:18:29,964 --> 00:18:33,224
I think this is like a mis- 
 
big misconception in life. 

357
00:18:33,959 --> 00:18:36,899
True for Asians, Indians, 
 and 
everybody else as well. 

358
00:18:37,459 --> 00:18:39,854
Yeah. 
I have a saying, saying 
 we 

359
00:18:39,854 --> 00:18:43,731
live poor, die rich. 
Yeah. 

360
00:18:45,669 --> 00:18:50,113
So you are right. 
So there is a… I got inspired 
 

361
00:18:50,121 --> 00:18:53,919
by this guy called, again, I 
forget his name all the time. 

362
00:18:54,499 --> 00:18:59,979
He was Sembcorp in Brazil. 
He wrote the book Maverick. 

363
00:19:00,559 --> 00:19:01,799
Why I forget his name every 
time? 

364
00:19:01,859 --> 00:19:03,659
I, and so he has a very amazing 
TikTok. 

365
00:19:04,239 --> 00:19:08,454
But anyway. 
So he says something like this, 

366
00:19:08,454 --> 00:19:12,288
"Think 
 about…" and I think I 
asked you this question, and I 

367
00:19:12,288 --> 00:19:16,701
actually wrote about it. 
I wrote a small utility on 
 my 

368
00:19:16,701 --> 00:19:20,388
blog called life.html. 
If you go to my blog at 
 

369
00:19:20,396 --> 00:19:22,911
ajeygore.in/life.html, it'll ask
you a few questions, and it'll 


370
00:19:22,919 --> 00:19:26,051
show you how much life is left. 
Wow, okay. 

371
00:19:28,379 --> 00:19:30,969
I really wanted to create a 
gadget 
 around it, like a 

372
00:19:30,969 --> 00:19:33,631
digital panel which like beeps 
today's year and 
 tells you how

373
00:19:33,631 --> 00:19:34,839
much is left, right? 
Yeah. 

374
00:19:35,779 --> 00:19:43,594
That, we, I talk about lot of 
 
people that they live as if they

375
00:19:43,594 --> 00:19:47,058
never, they will never die. 
But then they die as if 
 they 

376
00:19:47,058 --> 00:19:49,554
never lived, right? 
And this philosophical, but it 


377
00:19:49,562 --> 00:19:53,259
has been there with me for like 
almost like 10, 10-12 years. 

378
00:19:55,680 --> 00:19:57,620
So there is guy called Ricardo 
Semler. 

379
00:19:59,260 --> 00:20:01,020
Ricardo Semler has a very good 
TED Talk. 

380
00:20:02,330 --> 00:20:05,406
He talks about many, many, many,
things, 
 but one of the things 

381
00:20:05,406 --> 00:20:09,535
he talks about is think about 
this, Ricardo, one day your 
 

382
00:20:09,543 --> 00:20:11,879
doctor tells you that here is 
your scan. 

383
00:20:12,440 --> 00:20:14,680
You have one year to live 
 or 
six months to live. 

384
00:20:15,040 --> 00:20:16,740
What will you do in that 
 six 
months or one year? 

385
00:20:17,480 --> 00:20:18,800
I think your perspective 
changes. 

386
00:20:19,380 --> 00:20:23,780
And it changes very 
 
dramatically or drastically. 

387
00:20:25,120 --> 00:20:27,805
Up to you how you want to think 
 through it, but dramatically or

388
00:20:27,805 --> 00:20:31,420
drastically changes, right? 
And then you will do many, many,

389
00:20:31,420 --> 00:20:34,390

 many things which you had in 
your bucket list or non-bucket 


390
00:20:34,398 --> 00:20:36,760
list or unbucket list, whatever.
You unbucket many things. 

391
00:20:36,860 --> 00:20:38,580
You had the bucket list 
 and 
you unbucket that. 

392
00:20:39,240 --> 00:20:43,668
So what he said, he's like, "Oh,
instead 
 of waiting for that 

393
00:20:43,668 --> 00:20:46,782
moment, I brought my terminal 
days forward. 

394
00:20:46,782 --> 00:20:51,408
So every Wednesday 
 I go out 
with my wife. 

395
00:20:51,408 --> 00:20:55,374
Every Sunday night or every 
Friday I go out an afternoon for

396
00:20:55,374 --> 00:20:57,230

 movie." 
I have actually done that once. 

397
00:20:57,680 --> 00:21:00,120
We went out for a movie in 
 
afternoon, Monday afternoon. 

398
00:21:00,280 --> 00:21:01,135
It was amazing. 
Yeah. 

399
00:21:03,020 --> 00:21:04,500
There was no rush. 
Everybody's chill. 

400
00:21:04,660 --> 00:21:06,340
The hall for ourself, we 
enjoyed. 

401
00:21:06,380 --> 00:21:09,200
We had a food. 
No, there's no rush over there. 

402
00:21:09,220 --> 00:21:12,227
And good restaurants like try to
do 
 that on Friday, Saturday, 

403
00:21:12,227 --> 00:21:13,100
Sunday. 
Why? 

404
00:21:14,360 --> 00:21:17,027
Correct? 
So that is where the 
 slack 

405
00:21:17,027 --> 00:21:19,870
comes in, right? 
And it's very fulfilling. 

406
00:21:20,440 --> 00:21:24,775
So lot of us also, and I tell, I
usually 
 tell this to my Gojek 

407
00:21:24,775 --> 00:21:28,499
colleagues, I say, "Take leave, 
take off, go somewhere, 
 do 

408
00:21:28,499 --> 00:21:30,530
something." 
But spend your leaves. 

409
00:21:31,040 --> 00:21:34,984
Like, don't let it… Like there's
no 
 pride in carrying forward 

410
00:21:34,984 --> 00:21:36,967
your leave. 
"Oh, I have like 45 leaves." 
 

411
00:21:36,975 --> 00:21:38,220
Like what will you do with 
those? 

412
00:21:39,280 --> 00:21:41,060
Correct? 
People don't take leave. 

413
00:21:41,760 --> 00:21:44,818
They want to work. 
And I'm saying I don't see 
 

414
00:21:44,826 --> 00:21:49,739
things as a work-life balance. 
I don't see thing as oh, 0% , 0%

415
00:21:49,739 --> 00:21:51,120
or 100%. 
No, it's not like that. 

416
00:21:53,080 --> 00:21:55,868
It's a blend. 
It is somewhere in the 
 shade 

417
00:21:55,868 --> 00:21:57,990
of gray, right? 
It's not like I work eight 
 

418
00:21:57,998 --> 00:21:59,920
hours and I switch off. 
No, you can't do that. 

419
00:22:00,400 --> 00:22:02,290
If you have to work 16 
 hours, 
you work 16 hours. 

420
00:22:02,360 --> 00:22:03,819
We worked like three days night 
outs. 

421
00:22:03,980 --> 00:22:04,920
I have done that. 
Do it. 

422
00:22:05,530 --> 00:22:08,930
But then slack hard as well, 
yeah? 

423
00:22:09,150 --> 00:22:12,428
And if you do that, if you 
practice 
 that, then nobody's 

424
00:22:12,428 --> 00:22:14,900
gonna question you. 
Then, "Hey, why you are not 

425
00:22:14,900 --> 00:22:16,530
coming on 
 Friday?" 
Or, "Why you are not coming on 

426
00:22:16,530 --> 00:22:19,066
Monday?" 
Because they know that if there 

427
00:22:19,066 --> 00:22:23,033
 is something on Friday night or
Saturday night and with that 

428
00:22:23,033 --> 00:22:26,120
require Henry's, or 
 Ajey's 
attention, they will be there. 

429
00:22:27,020 --> 00:22:30,280
So we have to build that trust, 
and then 
 we can slack during 

430
00:22:30,280 --> 00:22:33,758
the work as well. 
But having this kind of slack, 

431
00:22:33,758 --> 00:22:37,360
what I'm 
 doing right now with 
advising companies, doing small 

432
00:22:37,360 --> 00:22:41,751
things, I'm… Just say, I'm 
 
having small meals, let's put it

433
00:22:41,751 --> 00:22:44,235
that. 
I'm having small bites the, 
 

434
00:22:44,243 --> 00:22:46,940
during the whole day, right? 
But I'm still doing this. 

435
00:22:47,260 --> 00:22:49,540
I'm still working. 
It's not like I'm not working. 

436
00:22:49,600 --> 00:22:52,618
I'm still working with NUS, I'm 
working 
 with a few companies, 

437
00:22:52,618 --> 00:22:56,391
I'm doing this. 
But not in a traditional way. 

438
00:22:56,981 --> 00:23:00,525
I have sprints, so I run 100 
meters, 
 then I take a rest of 

439
00:23:00,525 --> 00:23:01,399
one hour. 
Yeah. 

440
00:23:01,579 --> 00:23:04,375
What I learned as well from some
people, 
 yeah, the concept of 

441
00:23:04,375 --> 00:23:07,299
work-life balance sometimes is 
not appropriate, right? 

442
00:23:07,379 --> 00:23:10,630
It's more like seasonal thing, 
right, 
 where sometimes you can

443
00:23:10,630 --> 00:23:12,946
work hard. 
But just don't forget that you 

444
00:23:12,946 --> 00:23:14,839
also have 
 to go through, like,
the winter season. 

445
00:23:15,359 --> 00:23:18,685
Maybe take a break, you know, 
chill 
 a little bit, not like 

446
00:23:18,685 --> 00:23:21,863
running 100% all the time, which
probably is… I have done that 

447
00:23:21,863 --> 00:23:24,795
during my Gojek 
 days, and I 
was very proud of that. 

448
00:23:24,955 --> 00:23:27,453
And then I realized I was like, 
 "Shit." 

449
00:23:27,453 --> 00:23:31,960
I wrote about that as well. 
I talked of blockers of pit 

450
00:23:31,960 --> 00:23:33,387
stop. 
Like if you don't take a pit 
 

451
00:23:33,395 --> 00:23:36,715
stop in your car and you just go
around, the tires will bust. 

452
00:23:37,565 --> 00:23:39,884
So yeah. 
They go fast, but they 
 take a 

453
00:23:39,884 --> 00:23:41,015
pit stop as well. 
Yeah. 

454
00:23:41,135 --> 00:23:44,582
And I like the, like you brought

 the importance about life, 

455
00:23:44,582 --> 00:23:46,531
right? 
So I also follow stoicism and 

456
00:23:46,531 --> 00:23:48,575
all that. 
They also have the same 
 

457
00:23:48,583 --> 00:23:51,725
concept, like memento mori, you 
love, for the love of death. 

458
00:23:52,195 --> 00:23:54,915
So, you know, the illustration 

of number of days left and you 

459
00:23:54,915 --> 00:23:57,645
just cross it every day. 
I think it gives you perspective

460
00:23:57,645 --> 00:23:59,815

 that actually, yeah, life is 
kind of like limited, right? 

461
00:24:00,035 --> 00:24:03,220
So if you don't realize how much
left, 
 you wouldn't be able to 

462
00:24:03,220 --> 00:24:05,675
spend the time. 
And also it happens like 
 you 

463
00:24:05,675 --> 00:24:11,454
ask this, what for? 
And most probably so, so people 

464
00:24:11,454 --> 00:24:15,497
 who have extremely successful 
or extremely not successful, 

465
00:24:15,497 --> 00:24:19,873
they reach 
 a peak and either 
they go down in the valley or 

466
00:24:19,873 --> 00:24:22,901
they keep climbing, right? 
But at one point of time, their 

467
00:24:22,901 --> 00:24:25,450
 time, if you ask them this 
question, and we don't know 

468
00:24:25,450 --> 00:24:27,903
when. 
So a lot of people who are 

469
00:24:27,903 --> 00:24:30,976
hearing this 
 right now may say
no to answer, because that 

470
00:24:30,976 --> 00:24:34,463
question, the answer to that 
question 
 comes when there is a

471
00:24:34,463 --> 00:24:35,928
realization. 
And that realization can happen 

472
00:24:35,928 --> 00:24:37,680
at 
 eighty years old or 
eighteen years old. 

473
00:24:37,710 --> 00:24:41,939
So it may not come to many, many

 people and said, "No, it was, 

474
00:24:42,139 --> 00:24:45,006
it is all worth it." 
But at one 
 point of time, when

475
00:24:45,006 --> 00:24:47,146
you ask the question, "Is it 
worth it?" 

476
00:24:47,146 --> 00:24:52,279
Like, no. 
What worth is, what worth it 
 

477
00:24:52,287 --> 00:24:55,639
is, is friends, family, kids. 
Yeah, relationship 

478
00:24:55,739 --> 00:24:57,299
Relationships. 
That's what is worth it. 

479
00:24:57,339 --> 00:24:57,639
Yeah. 
Yeah. 

480
00:24:58,560 --> 00:25:01,694
So speaking about this, right? 
So you mentioned about, you 

481
00:25:01,694 --> 00:25:04,403
know, 
 you have to challenge 
yourself, be uncomfortable about

482
00:25:04,403 --> 00:25:07,579
stuff, 
 and these days it's all
about AI. 

483
00:25:07,579 --> 00:25:11,409
I mean, if can we switch 
 topic
from life perspective. 

484
00:25:11,639 --> 00:25:14,939
So I think AI brings a lot of 
 
anxiety to people today, right? 

485
00:25:15,219 --> 00:25:19,160
And also brings a lot of 
 maybe
existential crisis. 

486
00:25:19,299 --> 00:25:21,839
You know, some, for example, 
developers, 
 we think we are 

487
00:25:21,839 --> 00:25:26,499
gonna have no job in the future,
in the short, you know, time. 

488
00:25:26,779 --> 00:25:27,979
So what's your perspective with 
all this? 

489
00:25:28,079 --> 00:25:30,698
Especially you have dealt with, 
you 
 know, like software 

490
00:25:30,698 --> 00:25:32,583
engineering, building 
engineering teams for quite a 

491
00:25:32,583 --> 00:25:36,607
long time. 
So what does this AI era, you 
 

492
00:25:36,615 --> 00:25:39,805
know, give you perspective? 
Oh, there are three perspectives

493
00:25:39,805 --> 00:25:44,001
I have. 
I was talking about, I was 
 

494
00:25:44,009 --> 00:25:47,701
reading and talking at people. 
So first perspective is 

495
00:25:47,701 --> 00:25:50,056
engineering 
 practices are 
getting reinforced. 

496
00:25:50,215 --> 00:25:52,890
That is first perspective. 
People are talking about the 
 

497
00:25:52,898 --> 00:25:57,043
spec-driven development and TDD 
and like-- I also wrote a post 

498
00:25:57,043 --> 00:25:58,733
about, 
 it's called "expensive 
thing". 

499
00:25:59,573 --> 00:26:03,913
Though my point is that people 
are 
 trying to make sure that 

500
00:26:03,913 --> 00:26:06,852
whatever is being developed is 
developed correct, 
 right, and 

501
00:26:06,852 --> 00:26:09,455
all the stuff, right? 
So that is first perspective. 

502
00:26:09,495 --> 00:26:13,779
So I'm happy that people are 
 
reinforcing those software 

503
00:26:13,779 --> 00:26:16,355
engineering principles which we 
lived and breathed. 

504
00:26:16,415 --> 00:26:19,851
We both are from ThoughtWorks, 
and 
 we have lived and breathed

505
00:26:19,851 --> 00:26:23,415
the whole Agile, TDD stuff, and 
we know the value. 

506
00:26:23,925 --> 00:26:27,931
And we know that many, many 
people, when 
 we go, we would 

507
00:26:27,931 --> 00:26:30,201
go and tell them, "Please write 
tests." 

508
00:26:30,201 --> 00:26:32,465
They say, "No, tests will 
 slow
us down." 

509
00:26:32,465 --> 00:26:34,235
But that is coming now. 
Yeah. 

510
00:26:35,399 --> 00:26:38,272
Earlier, I think we did trust 
 
developers because they were 

511
00:26:38,272 --> 00:26:40,170
humans. 
Now we don't trust agents 
 

512
00:26:40,178 --> 00:26:42,479
because they are not humans. 
But both developers and 
 agents

513
00:26:42,479 --> 00:26:44,770
did the same thing. 
They both do the same mistakes. 

514
00:26:45,500 --> 00:26:48,269
But earlier we did not verify, 
and 
 we verified in very manual

515
00:26:48,269 --> 00:26:51,038
way. 
But now the amount of code AI 
 

516
00:26:51,046 --> 00:26:54,399
is producing, you can't do that.
So that is first thing. 

517
00:26:54,439 --> 00:26:56,999
The engineering principles 
 are
getting in place. 

518
00:27:00,632 --> 00:27:06,817
Second thing. 
This is exactly like a cognitive

519
00:27:06,817 --> 00:27:10,203

 industrial revolution at par 
with the industrial revolution 

520
00:27:10,203 --> 00:27:12,235
which happened 
 hundred fifty, 
two hundred years ago. 

521
00:27:13,565 --> 00:27:17,816
Like think about it, when, steam

 engines replaced horse carts. 

522
00:27:19,035 --> 00:27:21,943
Think about it when handlooms 
got replaced 
 with the 

523
00:27:21,943 --> 00:27:24,475
mechanical or industrial looms. 
What happened then? 

524
00:27:25,715 --> 00:27:30,799
What did those people do? 
One of the interesting thing 

525
00:27:30,799 --> 00:27:34,784
those 
 people did, industrial 
revolution did not bring 

526
00:27:34,784 --> 00:27:37,371
unemployment. 
It actually bring more 

527
00:27:37,371 --> 00:27:40,029
employment. 
That means on the sidelines, 
 

528
00:27:40,037 --> 00:27:43,694
there are different jobs now. 
Instead of putting the horse 

529
00:27:43,694 --> 00:27:46,596
cart, 
 you are actually-- you 
have to learn to become engine 

530
00:27:46,596 --> 00:27:48,983
driver. 
And then you need somebody at 

531
00:27:48,983 --> 00:27:50,879
the 
 end of the ca-- train as 
well. 

532
00:27:51,380 --> 00:27:53,514
And then you need a conductor, 
then you 
 need many more 

533
00:27:53,514 --> 00:27:55,979
things, and then you get more 
efficiency and all the stuff, 

534
00:27:55,979 --> 00:27:59,719
right? 
Like think about then when 

535
00:27:59,719 --> 00:28:02,979
people got 
 airplanes, did 
ships and cruises stop? 

536
00:28:03,139 --> 00:28:05,400
No, they did not. 
But they reduced. 

537
00:28:05,939 --> 00:28:08,922
But then there's a new industry 
 got created, and people 

538
00:28:08,922 --> 00:28:11,003
transformed and moved, and 
operators became 
 better 

539
00:28:11,003 --> 00:28:13,403
operator, all the stuff. 
So this is industrial 

540
00:28:13,403 --> 00:28:16,011
revolution. 
What I'm trying to say is that 

541
00:28:16,011 --> 00:28:18,699
age 
 of agents will reward us 
as builders. 

542
00:28:20,439 --> 00:28:23,712
The problem is lot of developers

 think they are just 

543
00:28:23,712 --> 00:28:28,399
developers. 
I have always debated that 
 

544
00:28:28,407 --> 00:28:32,099
software engineering is a 
balance of art and science. 

545
00:28:33,859 --> 00:28:36,599
That means software engineering 
 is not industrial process. 

546
00:28:36,759 --> 00:28:39,491
That means we all are creative. 
If programmers are not 
 

547
00:28:39,499 --> 00:28:43,131
creative, they can't program. 
Unless they're very dumb 

548
00:28:43,131 --> 00:28:45,197
programmers. 
But if you're good-- and, those 

549
00:28:45,197 --> 00:28:46,919
 guys will get out of job 
anyway. 

550
00:28:47,869 --> 00:28:50,359
Like they will get let go 
anyway. 

551
00:28:50,679 --> 00:28:53,745
So-- And they have to do 
 
something about it, and that is 

552
00:28:53,745 --> 00:28:56,809
completely different stuff. 
And this-- what I'm saying is 
 

553
00:28:56,817 --> 00:28:59,939
may backfire, but it's okay. 
But we are creative people. 

554
00:29:00,199 --> 00:29:03,767
Now if software engineers are 
 
creative people, that means this

555
00:29:03,767 --> 00:29:07,887
is age of builders. 
If it is age of builders, 
 it's

556
00:29:07,887 --> 00:29:11,813
a boon for us. 
We can build things at different

557
00:29:11,813 --> 00:29:15,379

 speed, different thing. 
So what it means, right? 

558
00:29:16,920 --> 00:29:18,659
Now what it means is very 
simple. 

559
00:29:21,120 --> 00:29:24,540
Think about-- I'm trying to 
 
figure out a very bad analogy. 

560
00:29:24,620 --> 00:29:26,588
I have a very bad analogy, but 
anyway, 
 I'm gonna give you 

561
00:29:26,588 --> 00:29:29,398
that bad analogy. 
Think about we have a dog, and 

562
00:29:29,398 --> 00:29:32,889
we 
 have a ball, and we throw 
the ball, and the dog goes and 

563
00:29:32,889 --> 00:29:34,240
fetches, and 
 we're trying to 
eat something. 

564
00:29:34,620 --> 00:29:37,023
But dog is so fast that it comes
back 
 immediately, so we are 

565
00:29:37,023 --> 00:29:39,430
not able to try. 
And then we take it and throw it

566
00:29:39,430 --> 00:29:42,210
very far. 
And we try to eat one bite, 
 

567
00:29:42,218 --> 00:29:44,886
and instantly it comes back. 
And then this time we're gonna 


568
00:29:44,894 --> 00:29:48,890
throw it really very far, right?
Now the dog and the ball fetch, 

569
00:29:48,890 --> 00:29:52,140
 this whole thing is agent. 
Coding agent. 

570
00:29:53,280 --> 00:29:57,662
So what happens now, we type 
 
something, do this, it does 

571
00:29:57,662 --> 00:29:59,818
that. 
But we have not thought about 

572
00:29:59,818 --> 00:30:01,270
it. 
Now we are reacting, okay, it's 

573
00:30:01,270 --> 00:30:02,383
free. 
Now I have to type something 

574
00:30:02,383 --> 00:30:03,860
else. 
Now I'm typing something else. 

575
00:30:04,380 --> 00:30:07,420
And that fetch and pull 
 is 
happening, right? 

576
00:30:08,360 --> 00:30:11,730
Now what has to happen, there 
 
has to be a paradigm shift. 

577
00:30:13,390 --> 00:30:18,280
See, in software industry, 
 it 
is not a people problem. 

578
00:30:19,080 --> 00:30:22,200
It is a management problem. 
Think about it. 

579
00:30:24,040 --> 00:30:29,340
Earlier, our bottleneck 
 was 
the people who execute. 

580
00:30:30,670 --> 00:30:34,846
Software developers. 
And then the bottleneck was not 

581
00:30:34,846 --> 00:30:38,650
the people 
 who ideate or give 
the problem, like product 

582
00:30:38,650 --> 00:30:42,350
managers, BAs, and all stuff. 
Now the thing have flipped. 

583
00:30:43,140 --> 00:30:46,580
Now we have to create 
 enough 
thought process. 

584
00:30:46,680 --> 00:30:50,370
We have to flip it on the 
 
things and start thinking like, 

585
00:30:51,440 --> 00:30:56,723
how can we create more specs? 
How can we create more, better 

586
00:30:56,723 --> 00:30:58,580
PRDs? 
How can we define correct? 

587
00:30:59,160 --> 00:31:02,167
Like one of the things which TDD
works 
 and the reason we had 

588
00:31:02,167 --> 00:31:04,535
Gojek with like nine hundred 
plus microservices, thousands 
 

589
00:31:04,543 --> 00:31:07,784
of deploys, and we had 
downtimes, but we did not have 

590
00:31:07,784 --> 00:31:10,643
that much crazy downtimes. 
I mean, we could do root cause 

591
00:31:10,643 --> 00:31:13,240
analysis 
 and people could 
deploy ten times a day. 

592
00:31:13,280 --> 00:31:15,522
There were like easily two 
thousand 
 deploys or three 

593
00:31:15,522 --> 00:31:17,680
thousand deploys a day, and we 
still could do that. 

594
00:31:17,740 --> 00:31:21,745
We did not wait for weekends. 
Because we were defining, all 

595
00:31:21,745 --> 00:31:24,700
the 
 time all of us define what
is correct. 

596
00:31:24,800 --> 00:31:27,350
See, tests define what is 
correct. 

597
00:31:27,940 --> 00:31:30,720
What is that judgment which says

 what is acceptance criteria? 

598
00:31:31,900 --> 00:31:33,788
When we define acceptance 
criteria, 
 we define both 

599
00:31:33,788 --> 00:31:36,078
things, right? 
What it means when it works 

600
00:31:36,078 --> 00:31:38,876
perfectly 
 fine, and what it 
means when it does not work 

601
00:31:38,876 --> 00:31:41,120
perfectly fine. 
We write both, right? 

602
00:31:41,920 --> 00:31:45,760
If I give a wrong username and 

wrong password, it should fail. 

603
00:31:45,840 --> 00:31:50,039
We have to write that. 
While we write also that given a

604
00:31:50,039 --> 00:31:53,389

 right username and right 
password, it should log me in, 

605
00:31:53,389 --> 00:31:56,830
right? 
So that defining that correct 

606
00:31:56,830 --> 00:31:59,520
means 
 it also encompasses 
incorrect. 

607
00:32:01,440 --> 00:32:05,322
Now given that, if you now think

 through my view, now we have 

608
00:32:05,322 --> 00:32:10,760
to spend more time, more chunky 
job, 
 so that agent go on for a

609
00:32:10,760 --> 00:32:13,010
long time. 
Correct? 

610
00:32:13,560 --> 00:32:18,324
So instead of saying, "What did 
you do 
 yesterday?" 

611
00:32:18,324 --> 00:32:22,890
The standup questions change. 
"Henry, what did we deliver 

612
00:32:22,890 --> 00:32:25,134
yesterday?" 
 
And what did you deliver 

613
00:32:25,134 --> 00:32:28,760
yesterday means, what did you 
validate yesterday, right? 

614
00:32:29,260 --> 00:32:32,560
So now the whole-- there 
 is a 
whole paradigm shift. 

615
00:32:32,640 --> 00:32:35,500
Now you're a product manager, 
but instead 
 of having like one

616
00:32:35,500 --> 00:32:39,564
BA and four developers, you have
four BAs and one developer, 
 

617
00:32:39,572 --> 00:32:43,020
and maybe, 0.5 QA or maybe no 
QA. 

618
00:32:44,500 --> 00:32:46,809
Because verification has 
 to go
up to the level. 

619
00:32:47,740 --> 00:32:51,582
Then what would QA do, right? 
They also have their work, but 


620
00:32:51,590 --> 00:32:54,716
their work is now changing to 
becoming more QA automation 

621
00:32:54,716 --> 00:33:00,295
people. 
So the whole paradigm shift is 

622
00:33:00,295 --> 00:33:05,055
flipping. 
The builders getting more hand, 

623
00:33:05,055 --> 00:33:10,212
more 
 control, and 
executioners, which are like 

624
00:33:10,212 --> 00:33:14,350
developers are getting less. 
So if you are gonna remain 

625
00:33:14,350 --> 00:33:19,124
developer, 
 eventually you will
have to go up and get your 

626
00:33:19,124 --> 00:33:22,991
better skill at creating a 
better 
 judgment, experiences, 

627
00:33:22,991 --> 00:33:25,970
and then you will be able to get
more useful skills. 

628
00:33:26,600 --> 00:33:30,237
What you pay for, what you 
 are
getting paid and what you pay 

629
00:33:30,237 --> 00:33:33,660
for is always a skill. 
You'll never pay for typing 

630
00:33:33,660 --> 00:33:38,366
speed, do you? 
So now what has happened is 

631
00:33:38,366 --> 00:33:42,580
typing speed 
 and some of the 
structured skill is gone. 

632
00:33:43,940 --> 00:33:45,500
But this is not only happening 
here. 

633
00:33:45,660 --> 00:33:46,860
You don't define correct over 
here. 

634
00:33:46,900 --> 00:33:48,880
Now you have to define 
 correct
in domains as well. 

635
00:33:48,960 --> 00:33:50,900
Financial domain, you 
 have to 
define correct. 

636
00:33:51,520 --> 00:33:54,445
If you don't define correct, 
then you'll 
 have wrong 

637
00:33:54,445 --> 00:33:56,920
financial audits, right? 
Lawyers, you have to define 

638
00:33:56,920 --> 00:33:59,515
correct. 
We know that whole hallucination

639
00:33:59,515 --> 00:34:03,910
and whole 
 legal issues and 
LLMs generated have fake cases 

640
00:34:03,910 --> 00:34:06,210
and all stuff because they're 
just 
 token generators at the 

641
00:34:06,210 --> 00:34:10,960
end of day, right? 
So the more structured jobs 
 

642
00:34:10,967 --> 00:34:17,676
will increase the circle of 
skills, and you have to 
 learn 

643
00:34:17,676 --> 00:34:20,280
peripheral skills. 
And then you can-- and 
 you 

644
00:34:20,280 --> 00:34:21,969
should be a builder. 
All of us are a builder. 

645
00:34:22,280 --> 00:34:25,420
So it gives a fantastic chance 

for us to actually look forward 

646
00:34:25,500 --> 00:34:30,306
for doing that- instead of 
trying 
 to resist AI and trying

647
00:34:30,306 --> 00:34:34,715
to resist agents and like, oh, 
we'll still-- 
 Yeah, you can 

648
00:34:34,715 --> 00:34:36,342
create artisanal code. 
Yeah. 

649
00:34:36,830 --> 00:34:38,577
Yeah. 
Like I was telling somebody, I 

650
00:34:38,577 --> 00:34:41,228
was 
 telling my, one of my 
ex-colleagues saying, "Remember 

651
00:34:41,228 --> 00:34:44,612
ten years back, 
 we actually 
wrote every bit of that line?" 

652
00:34:44,612 --> 00:34:45,889
Actually, I'm telling you. 
Yes. 

653
00:34:45,971 --> 00:34:48,068
Same. 
Yeah, we wrote every bit of that

654
00:34:48,068 --> 00:34:49,780
line. 
Can't believe it. 

655
00:34:50,780 --> 00:34:53,347
And now, I'm actually happy, and

 people should be happy about 

656
00:34:53,347 --> 00:34:54,520
it. 
They should embrace this. 

657
00:34:54,699 --> 00:34:55,110
Yeah. 
Yeah. 

658
00:34:55,583 --> 00:34:57,154
Yeah, I still remember 
 writing
getter, setter. 

659
00:34:57,184 --> 00:34:59,954
That's like the boring- Every 
time, yeah. 

660
00:34:59,954 --> 00:35:02,784
Control+Space, Control- 
 
everytime, it's so boring. 

661
00:35:02,844 --> 00:35:05,303
It's so, so, so boring. 
So yeah. 

662
00:35:05,443 --> 00:35:11,023
So I was saying, so when 
expertise 
 is becoming almost 

663
00:35:11,023 --> 00:35:15,480
free, whether it is design, and 
it's happening every 
 day, 

664
00:35:15,480 --> 00:35:19,463
whether design, security, audit,
everywhere, then what becomes 

665
00:35:19,463 --> 00:35:21,925
expensive? 
And my answer is judgment 
 

666
00:35:21,933 --> 00:35:25,225
becomes expensive. 
So we need to fundamentally 
 

667
00:35:25,233 --> 00:35:27,075
start focusing on judgment 
layer. 

668
00:35:28,436 --> 00:35:31,776
And judgment layer is gonna be 

very important for everybody, 

669
00:35:32,235 --> 00:35:35,454
like how agents are doing. 
Because if agents can do good 

670
00:35:35,454 --> 00:35:38,235
things 
 fast, they'll do bad 
things fast as well. 

671
00:35:38,355 --> 00:35:40,716
Yeah. 
Yes, exactly. 

672
00:35:40,995 --> 00:35:43,666
We all see it in the 
 
cybersecurity space, right? 

673
00:35:43,666 --> 00:35:45,355
You have hackers and you have 
defenders. 

674
00:35:45,710 --> 00:35:49,006
Yeah. 
And, like we had like a few days

675
00:35:49,006 --> 00:35:53,397
back, 
 I read this document 
where a coding agent went and 

676
00:35:53,397 --> 00:35:58,780
destroyed the entire 
 AWS cloud
environment, and said, "Oh, 

677
00:35:58,820 --> 00:36:01,704
sorry, I don't have backups. 
I should 
 have checked before 

678
00:36:01,704 --> 00:36:03,804
doing this." 
Like, yeah, but now actually you

679
00:36:03,804 --> 00:36:05,780
don't 
 have anybody to scream 
at as well. 

680
00:36:08,731 --> 00:36:11,575
But one thing that doesn't help 
is 
 actually all these big AI 

681
00:36:11,575 --> 00:36:14,866
companies are saying, you know, 
it's like 
 fearmongering in a 

682
00:36:14,866 --> 00:36:16,975
sense, right? 
Maybe in the interest of 

683
00:36:16,975 --> 00:36:18,511
increasing 
 their valuation, 
whatever that is. 

684
00:36:18,572 --> 00:36:21,148
They're saying like, "Okay, 
software 
 developers, we don't 

685
00:36:21,148 --> 00:36:24,343
need anymore. 
Now going to legal, accountant, 

686
00:36:24,343 --> 00:36:27,367
and all that." 
 
So there's this perspective 

687
00:36:27,367 --> 00:36:30,071
that, yeah, maybe I need to 
really switch my skill. 

688
00:36:30,410 --> 00:36:34,598
No, no, no, no. 
Look, 88% of AI companies are 

689
00:36:34,598 --> 00:36:37,520
actually 
 adopt-- or companies 
are adopting AI. 

690
00:36:38,760 --> 00:36:42,130
Adoption has increased. 
But only 6% have successful 

691
00:36:42,130 --> 00:36:44,410
projects. 
So why? 

692
00:36:45,310 --> 00:36:49,380
Because look, it is-- 
 Okay, 
I'll tell you this. 

693
00:36:49,560 --> 00:36:51,200
I'll let me put-- give you this 
example. 

694
00:36:51,260 --> 00:36:53,720
I gave this example about Linux 
 many, many, many years ago. 

695
00:36:54,030 --> 00:36:56,020
I'll give the same example about
AI now. 

696
00:36:58,000 --> 00:36:59,460
So when you get into Unix shell,
right? 

697
00:37:00,100 --> 00:37:03,709
And it is powerful, and what you
get 
 is a blank screen and one 

698
00:37:03,709 --> 00:37:06,176
prompt. 
And then you are trying to 

699
00:37:06,176 --> 00:37:08,000
install one 
 package, and you 
don't know how to do it. 

700
00:37:08,930 --> 00:37:13,240
And then you struggle. 
Like, AI is like that. 

701
00:37:14,360 --> 00:37:16,360
You are given a very 
 
beautiful, amazing car. 

702
00:37:17,760 --> 00:37:20,780
Take any ex-- any luxury brand, 
 and you're given the key. 

703
00:37:20,840 --> 00:37:23,820
You started the car, and it 
 
started because you touched it. 

704
00:37:25,140 --> 00:37:28,940
Starting a car and running it 
 
on a big field is very different

705
00:37:29,260 --> 00:37:32,040
than starting a car and running 
 in a controlled chaos road. 

706
00:37:34,280 --> 00:37:37,920
So a lot of people, what they're

 experiencing right now is they

707
00:37:37,960 --> 00:37:40,180
got this car, and they got this 
 big field, and they're running 

708
00:37:40,220 --> 00:37:41,500
like anything, and it's all 
good. 

709
00:37:42,180 --> 00:37:44,437
But when they get out of that 
 
field, and they exit the field 

710
00:37:44,437 --> 00:37:48,152
and main-- come to main road, 
now you 
 have to put a lot of 

711
00:37:48,152 --> 00:37:49,942
guardrails. 
Who is gonna put those 

712
00:37:49,942 --> 00:37:52,485
guardrails? 
Only people who know how to put 

713
00:37:52,485 --> 00:37:53,860
 guardrails will put the 
guardrails. 

714
00:37:54,220 --> 00:37:58,680
So our job is changing from… 
 
So earlier what used to happen, 

715
00:37:58,980 --> 00:38:01,985
we do cognitive job, right? 
All of us, like lawyers and 

716
00:38:01,985 --> 00:38:03,420
financial, 
 everybody does 
cognitive job. 

717
00:38:03,480 --> 00:38:07,090
We sit and do. 
We type, and the machine 
 does 

718
00:38:07,090 --> 00:38:09,920
something, right? 
We don't do heavy lifting job. 

719
00:38:10,040 --> 00:38:15,260
So all that cognitive job was 
 
about to build guardrails, 

720
00:38:15,440 --> 00:38:17,000
always correcting, always 
fixing. 

721
00:38:17,620 --> 00:38:20,640
And that was, as a software 
 
developer, that part in our life

722
00:38:20,640 --> 00:38:24,110
was very, very small part. 
In the beginning, like morning, 

723
00:38:24,110 --> 00:38:26,980
 you read your story, you read 
your card, et cetera, you start 

724
00:38:26,980 --> 00:38:28,648
typing. 
Now the whole day you're just 
 

725
00:38:28,656 --> 00:38:30,792
typing, typing, typing, typing, 
typing, and following that 

726
00:38:30,792 --> 00:38:33,952
thing. 
But now our job is the typing, 

727
00:38:33,952 --> 00:38:36,319
if 
 typing is gonna, if-- Think
about this. 

728
00:38:37,900 --> 00:38:40,090
If typing is done by somebody 
 
else, then what will you do? 

729
00:38:42,507 --> 00:38:46,383
Then you'll think more. 
My point is, it is 
 telling us 

730
00:38:46,383 --> 00:38:50,315
to think more. 
It is allowing us to getting 
 

731
00:38:50,323 --> 00:38:55,127
free from that typist job to 
more creative tinkerer job. 

732
00:38:56,268 --> 00:38:58,267
So I don't think so that is 
gonna go away. 

733
00:38:58,768 --> 00:39:03,028
Yes, job will reduce because 
 
typing capacity is going away. 

734
00:39:04,284 --> 00:39:06,551
Yeah. 
So if you're doing that, 
 then 

735
00:39:06,551 --> 00:39:09,602
you are going away. 
Like for example, if you are 
 

736
00:39:09,610 --> 00:39:13,617
just a front-end developer, 
yeah, like two years old, three 

737
00:39:13,617 --> 00:39:15,567
years 
 experience front-end 
developer, yeah. 

738
00:39:16,687 --> 00:39:18,768
Because that basic stuff finesse
has come. 

739
00:39:19,428 --> 00:39:23,900
But if you want to create a UX, 
yeah, 
 Claude Design can do 

740
00:39:23,900 --> 00:39:27,832
many things. 
But if you really want to create

741
00:39:27,832 --> 00:39:31,063

 products and do something, 
experiences, we are the one who 

742
00:39:31,063 --> 00:39:34,032
can create. 
Machines will imitate, but 

743
00:39:34,032 --> 00:39:37,997
they'll imitate 
 so nicely, it 
almost feels like creation. 

744
00:39:39,927 --> 00:39:44,020
Correct? 
But if you go little bit further

745
00:39:44,020 --> 00:39:48,953
on that 
 edge and treat 
yourself as a builder, then you 

746
00:39:48,953 --> 00:39:51,735
will suddenly figure out, oh 
 
my God, I have so many 

747
00:39:51,735 --> 00:39:55,158
opportunities. 
Yeah And people are not thinking

748
00:39:55,158 --> 00:39:58,908
like that. 
People are trying to use 
 AI at

749
00:39:58,908 --> 00:40:02,617
a very minimal level. 
They're not trying to offload 
 

750
00:40:02,625 --> 00:40:06,469
their job and think more. 
I think I want to encourage 

751
00:40:06,469 --> 00:40:09,621
people 
 to offload their more 
job to AI so that they can do 

752
00:40:09,621 --> 00:40:14,958
something better. 
But if you always have been that

753
00:40:14,958 --> 00:40:21,308

 guy who goes and reboots the 
server, yes, your job is gone. 

754
00:40:21,527 --> 00:40:22,663
I'm sorry to say. 
Yeah. 

755
00:40:23,224 --> 00:40:25,403
So I think what you brought 
 up
a very good point, right? 

756
00:40:25,483 --> 00:40:27,503
So obviously we need to 
 level 
up ourselves, right? 

757
00:40:27,604 --> 00:40:31,375
In the value creation, right? 
No more typing as like a skill 

758
00:40:31,375 --> 00:40:34,661
or 
 pressing that button, but 
more like the builder, the 

759
00:40:34,661 --> 00:40:37,614
creative thinker, you know? 
Putting the judgment, I think 

760
00:40:37,614 --> 00:40:39,823
this is also 
 something that is
very important, right? 

761
00:40:39,883 --> 00:40:42,052
Putting the guardrails. 
Define correctness, 

762
00:40:42,052 --> 00:40:46,383
incorrectness, and 
 you know, 
the judgment on what is good. 

763
00:40:46,563 --> 00:40:48,223
Maybe some people also 
 call it
taste, right? 

764
00:40:48,563 --> 00:40:52,413
Because only humans, I hope, 
only humans 
 that can give the 

765
00:40:52,413 --> 00:40:55,423
taste, you know, opinion, right,
rather than the machine. 

766
00:40:55,943 --> 00:40:59,885
So earlier you mentioned it, 
it's 
 very hard for anyone to 

767
00:40:59,885 --> 00:41:02,690
predict now what's gonna happen 
with this 
 AI, but one thing 

768
00:41:02,690 --> 00:41:05,734
for sure, you know what people 
should not do. 

769
00:41:06,223 --> 00:41:10,808
So maybe tell us what we should 
not do 
 in order to navigate 

770
00:41:10,808 --> 00:41:14,980
this time better. 
We should not resist. 

771
00:41:15,923 --> 00:41:17,284
Okay. 
First thing is that, accept, 

772
00:41:17,284 --> 00:41:22,160
yeah. 
I, somebody asked me and we were

773
00:41:22,160 --> 00:41:25,000
debating 
 on the framework and 
all the stuff. 

774
00:41:25,810 --> 00:41:28,284
And I tweeted about it, like 
two, three 
 days back as 

775
00:41:28,284 --> 00:41:31,372
saying: "Hey, are you here to 
ship the framework or are you 
 

776
00:41:31,380 --> 00:41:32,980
here to ship the product?" 
Correct? 

777
00:41:33,860 --> 00:41:37,031
That's very common, right? 
So we should-- And when I ask 

778
00:41:37,031 --> 00:41:40,167
people, 
 a lot of them ask the 
people, like, are you-- and my, 

779
00:41:40,167 --> 00:41:42,246
one of my friends also asks 
 
these people, "What is your 

780
00:41:42,246 --> 00:41:43,948
job?" 
If I ask you, you are in 

781
00:41:43,948 --> 00:41:46,819
company, what is your job? 
They'll say, "My job is this 
 

782
00:41:46,827 --> 00:41:48,272
and that." 
And I know that, and I ask 

783
00:41:48,272 --> 00:41:50,026
again, "What is your job? 
Think more." 

784
00:41:50,026 --> 00:41:56,040
And few people reach a 
 place 
where say, my job is to have 

785
00:41:56,080 --> 00:42:00,320
value creation for the company. 
Whatever role you play, the 

786
00:42:00,320 --> 00:42:03,380
problem is 
 people mistake 
their role for a job. 

787
00:42:04,980 --> 00:42:07,676
If our job is to create value 
 
for a company, because our-- we 

788
00:42:07,676 --> 00:42:10,260
are giving our time to them, 
 
and that's why they will pay. 

789
00:42:10,340 --> 00:42:13,040
If we stop value creating, 
 
then they'll not pay us. 

790
00:42:13,240 --> 00:42:15,866
Very simple. 
So there are multiple 
 ways to 

791
00:42:15,866 --> 00:42:21,396
create value. 
So that's where I think, first, 

792
00:42:21,396 --> 00:42:25,156
accept it. 
Second, figure out multiple 
 

793
00:42:25,164 --> 00:42:29,403
ways of value creation. 
We all are very good cognitive 

794
00:42:29,403 --> 00:42:31,751
thinkers. 
We are very few-- We are one 
 

795
00:42:31,759 --> 00:42:35,470
of very few people in the world 
who are getting exposed to this 

796
00:42:35,470 --> 00:42:41,315
 and getting disrupted by this. 
People who are in, like, selling

797
00:42:41,315 --> 00:42:43,640

 hawker center things, they're 
not. 

798
00:42:43,880 --> 00:42:46,020
They're happy. 
People who are delivering 
 

799
00:42:46,028 --> 00:42:49,096
food, they're happy. 
They'll get, at some point of 

800
00:42:49,096 --> 00:42:52,270
time, 
 and bots and drones will
replace the food delivery or 

801
00:42:52,270 --> 00:42:54,600
some other delivery. 
But still there will be more 

802
00:42:54,600 --> 00:42:57,286
things. 
So my point is, this is a 

803
00:42:57,286 --> 00:42:59,380
cognitive 
 industrial 
revolution where this happens. 

804
00:42:59,410 --> 00:43:05,467
So first, accept change. 
Second, think about your value 

805
00:43:05,467 --> 00:43:09,212
creation. 
And third, what not to do is 

806
00:43:09,212 --> 00:43:11,280
very 
 important, is get worried
about it. 

807
00:43:12,400 --> 00:43:14,000
There's no reason to get worried
about it. 

808
00:43:14,760 --> 00:43:22,079
Because unless you stop worrying
about it, you 
 can't think 

809
00:43:22,079 --> 00:43:27,148
about something more, right? 
Maybe I'll not get a job, and 

810
00:43:27,148 --> 00:43:29,740
it's okay. 
Can I bake a bread? 

811
00:43:31,020 --> 00:43:33,815
Maybe yes. 
But I have this responsibility, 

812
00:43:33,815 --> 00:43:36,240
 I have that responsibility, I 
have that responsibility. 

813
00:43:36,280 --> 00:43:39,953
Fair. 
If you don't do anything, those 

814
00:43:39,953 --> 00:43:41,836
things 
 are still constant in 
your life. 

815
00:43:42,166 --> 00:43:46,255
As you said, if you are think 
about a 
 stressor, those things

816
00:43:46,255 --> 00:43:49,590
don't change. 
If worrying about things can 

817
00:43:49,590 --> 00:43:52,236
change 
 things, then I would 
worry whole day. 

818
00:43:53,275 --> 00:43:56,258
And I'm culprit of that. 
I worry about my son, my 

819
00:43:56,258 --> 00:43:59,940
daughter, 
 not daughter, my 
son, my wife, my relatives, my 

820
00:43:59,940 --> 00:44:01,115
friends. 
I worry about them, right? 

821
00:44:02,236 --> 00:44:06,689
But that worry gets away. 
But worried about them or 

822
00:44:06,689 --> 00:44:09,232
thinking 
 about them, like 
somebody falls sick, we worry 

823
00:44:09,232 --> 00:44:13,411
about them, but we pray, right? 
Not that religious pray, 
 but a

824
00:44:13,411 --> 00:44:15,481
spiritual pray. 
So my point is there is no 
 

825
00:44:15,489 --> 00:44:19,698
reason to worry about it. 
There is-- If you stop worrying,

826
00:44:19,698 --> 00:44:23,298
then 
 you can think about path 
forward, right? 

827
00:44:23,887 --> 00:44:28,225
A fourth thing which, which I 
think is 
 very important is 

828
00:44:28,225 --> 00:44:31,667
that stop running. 
Close your eyes. 

829
00:44:33,567 --> 00:44:38,111
Stop looking at your device. 
Stop getting consumed by stupid 

830
00:44:38,111 --> 00:44:41,287
internet. 
Give yourself time. 

831
00:44:42,588 --> 00:44:46,747
And then think how world, how 
 
beautiful world is around you. 

832
00:44:47,588 --> 00:44:49,807
And then you will think 
 about 
what can you do. 

833
00:44:54,051 --> 00:44:57,351
That concept many of people 
 
will not agree with me. 

834
00:44:57,811 --> 00:45:01,662
They'll give me excuses, they'll
give 
 me reasons, real reasons 

835
00:45:01,662 --> 00:45:04,781
saying, "Hey, you are talking 
about that, but I have 
 to pay 

836
00:45:04,781 --> 00:45:07,320
bills, I have to do this, I have
to do that, and I just got laid 

837
00:45:07,320 --> 00:45:12,311
off." 
There is no way you can get 
 

838
00:45:12,319 --> 00:45:16,416
back to the same company. 
And it's unfortunate. 

839
00:45:17,636 --> 00:45:19,596
But now you got a time, 
 you 
got a forced time. 

840
00:45:21,036 --> 00:45:24,492
And if you stop worrying and 
stop immed-- 
 A lot of people 

841
00:45:24,492 --> 00:45:26,586
just start immediately applying 
for a job somewhere else. 

842
00:45:27,636 --> 00:45:30,618
That used to happen earlier. 
But the problem is you got 
 

843
00:45:30,626 --> 00:45:32,716
laid off because there is some 
skill which got automated. 

844
00:45:33,496 --> 00:45:35,996
If it's got automated at company

 A, it will get automated at 

845
00:45:35,996 --> 00:45:38,429
company B as well, and company C

 as well, and company D as 

846
00:45:38,429 --> 00:45:40,271
well. 
So you're not gonna-- You can 

847
00:45:40,271 --> 00:45:43,854
apply 
 for five places, and if 
you are not a value creator for 

848
00:45:43,854 --> 00:45:46,166
them, they're 
 not gonna hire a
typist again. 

849
00:45:46,756 --> 00:45:49,196
This is a very crude way of 
saying it. 

850
00:45:49,276 --> 00:45:54,442
I'm so sorry to say this. 
So my point is, people 
 should 

851
00:45:54,442 --> 00:45:56,951
actually learn. 
Actually, you know what, 
 

852
00:45:56,959 --> 00:46:00,609
Henry, I'll tell you this. 
Don't you think our job has been

853
00:46:00,609 --> 00:46:03,477

 always be-- Like, we, our 
industry also has been always 

854
00:46:03,477 --> 00:46:06,046
been about learning. 
Like when you started, maybe 
 

855
00:46:06,054 --> 00:46:11,129
you are working in Java or C++. 
Now you work at Python, Rust, 

856
00:46:11,129 --> 00:46:12,656
Go. 
You learn on the way. 

857
00:46:13,356 --> 00:46:19,696
I think people forgot that we, 

our learning, learning exposure 

858
00:46:19,736 --> 00:46:23,768
or learning attack area or 
exposed 
 surface area has to be

859
00:46:23,768 --> 00:46:28,691
much larger now. 
So, like for example, I'll 
 

860
00:46:28,699 --> 00:46:33,986
give you one more example. 
Fifteen, twenty years back, you 

861
00:46:33,986 --> 00:46:38,196
could 
 go to a GP and not go to
specialist. 

862
00:46:38,486 --> 00:46:42,156
But today, if you got a tooth 
problem, 
 you'll go to dentist,

863
00:46:42,156 --> 00:46:45,965
but you'll go to the best 
dentist who has done master's 
 

864
00:46:45,973 --> 00:46:47,796
and post graduate, everything, 
right? 

865
00:46:49,156 --> 00:46:53,263
If you had to go to, like kidney

 or whatever, you'll go to MD 

866
00:46:53,263 --> 00:46:55,076
and whatnot and bunch of things,
right? 

867
00:46:55,836 --> 00:46:58,214
We seek. 
I think that also, I was 
 

868
00:46:58,222 --> 00:47:00,836
telling somebody, that also has 
come to software engineering. 

869
00:47:01,256 --> 00:47:04,076
Now you could just can't 
 have 
comp science degree. 

870
00:47:05,206 --> 00:47:07,388
You had a comp science degree, 
then 
 you had to do MS, then 

871
00:47:07,388 --> 00:47:10,138
you had to do something more, 
and because this value 
 

872
00:47:10,146 --> 00:47:12,136
creation has just gone really 
high. 

873
00:47:15,588 --> 00:47:18,021
Yes, it has gone very high in 
 
like very short amount of time 

874
00:47:18,021 --> 00:47:21,275
and people are reacting to it, 
but I 
 think eventually it will

875
00:47:21,275 --> 00:47:23,666
be fine. 
There are different kind 
 of 

876
00:47:23,666 --> 00:47:25,573
jobs will get created. 
For example, people actually are

877
00:47:25,573 --> 00:47:27,088
not 
 good at prompt engineering
at all. 

878
00:47:27,928 --> 00:47:30,983
That is a new job got created. 
And people are not good at that 

879
00:47:30,983 --> 00:47:32,427
at all. 
People should be good at that. 

880
00:47:32,468 --> 00:47:34,172
There is-- They should-- 
Actually, there 
 are papers 

881
00:47:34,172 --> 00:47:36,434
about prompt engineering. 
I was reading a paper from 

882
00:47:36,434 --> 00:47:38,890
Google. 
They are saying, if you repeat 


883
00:47:38,898 --> 00:47:41,719
the prompt, LLM does it better. 
Oh, really? 

884
00:47:41,859 --> 00:47:45,742
Yeah. 
Now, that is about learning, 

885
00:47:45,742 --> 00:47:47,551
right? 
So people are not doing 
 prompt

886
00:47:47,551 --> 00:47:50,443
engineering. 
A lot of people just use 
 

887
00:47:50,451 --> 00:47:55,228
Claude Code and just use Claude 
Cowork and Codex and all stuff. 

888
00:47:55,808 --> 00:47:59,034
But they're not thinking 
through, 
 how can I automate 

889
00:47:59,034 --> 00:48:03,348
the whole stuff? 
I don't know how many 

890
00:48:03,348 --> 00:48:06,998
organizations 
 actually have 
built and put agents on their 

891
00:48:06,998 --> 00:48:09,368
GitHub repositories 
 to 
auto-merge the PRs. 

892
00:48:10,328 --> 00:48:11,448
First of all, PR shouldn't be 
there. 

893
00:48:12,308 --> 00:48:15,277
But that is different software 

engineering, talking about 

894
00:48:15,277 --> 00:48:17,288
trunk-based development and all 
that stuff. 

895
00:48:17,768 --> 00:48:21,888
But have they put in PRs? 
Like, how many of them? 

896
00:48:22,808 --> 00:48:24,920
Whatever people talk about 
automation, 
 you know, they're 

897
00:48:24,920 --> 00:48:26,608
talking about automation on 
their laptops. 

898
00:48:27,419 --> 00:48:29,898
Yeah. 
They're not talking about 
 

899
00:48:29,906 --> 00:48:34,084
automation at the central level.
They're not talking about 

900
00:48:34,084 --> 00:48:36,424
saying, 
 "Hey, we will not 
develop on our laptops anymore. 

901
00:48:36,424 --> 00:48:39,787
We'll just do a 
 cloud-based 
workspaces and work over there."

902
00:48:39,787 --> 00:48:42,378
How many people are doing? 
Because once everybody works on 

903
00:48:42,378 --> 00:48:45,745
the 
 same box, same machine, 
shared terminal, then you can 

904
00:48:45,745 --> 00:48:47,528
have agent which just 
 can look
at everybody's source code. 

905
00:48:47,548 --> 00:48:48,868
And how many people have done 
that? 

906
00:48:49,588 --> 00:48:51,358
That is a skill they should be 
doing. 

907
00:48:52,208 --> 00:48:56,598
Like, remember, we went through 
 this DevOps thing, right? 

908
00:48:57,448 --> 00:49:00,888
The thought process was that 
 
we don't need anybody else, and 

909
00:49:00,928 --> 00:49:02,108
developers should be doing 
operator. 

910
00:49:02,608 --> 00:49:05,628
What happened? 
Other companies which could not 

911
00:49:05,628 --> 00:49:07,868
upskill 
 their job, they 
created DevOps role, man. 

912
00:49:08,505 --> 00:49:09,815
Yes. 
They're, are still there. 

913
00:49:11,398 --> 00:49:14,206
I don't like it. don't like it. 
Because DevOps is a set of 
 

914
00:49:14,214 --> 00:49:17,858
practices, it's not a role. 
But now big companies are 

915
00:49:17,858 --> 00:49:21,508
advertising 
 it as a role, and 
there's a DevOps role. 

916
00:49:21,628 --> 00:49:24,498
I'm like, no, that just sounds 
so wrong. 

917
00:49:25,788 --> 00:49:32,050
So that's where it is, right? 
So my point is, you have to 

918
00:49:32,050 --> 00:49:34,948
now-- 
 Your full stack 
engineer, developer engineer 

919
00:49:34,948 --> 00:49:38,172
definition has changed. 
Earlier, full stack was 

920
00:49:38,172 --> 00:49:41,248
front-end, 
 back-end, and 
whatever, and now it is like 

921
00:49:41,248 --> 00:49:43,248
front-end, back-end, 
 how to 
use tools, and how to write 

922
00:49:43,318 --> 00:49:45,368
agents, and you have to do that.
That's it. 

923
00:49:45,900 --> 00:49:48,179
Yeah, so I think really 
 good 
point, don't resist. 

924
00:49:48,739 --> 00:49:52,475
You know, do not get worried. 
Think about value creation 

925
00:49:52,475 --> 00:49:55,049
rather 
 than, you know, your 
tasks, your day-to-day job, 

926
00:49:55,049 --> 00:49:56,759
right? 
And I think participate, right? 

927
00:49:56,859 --> 00:50:00,579
You need to learn about all 
 
this AI stuff, you know. 

928
00:50:01,119 --> 00:50:03,927
Otherwise, I think you'll get 
 
replaced or you'll get 

929
00:50:03,927 --> 00:50:06,340
irrelevant in the world, in the 
industry, right? 

930
00:50:07,000 --> 00:50:11,203
So speaking about engineering 
 
practices, you mentioned it is 

931
00:50:11,203 --> 00:50:15,125
getting reinforced, and I think 
for people 
 who wants to 

932
00:50:15,125 --> 00:50:19,325
leverage AI in the right way to 
always produce correct things, 


933
00:50:19,333 --> 00:50:23,836
you really need these practices.
Maybe if you can give your, I 

934
00:50:23,836 --> 00:50:26,929
don't 
 know, like few top 
engineering practices that needs

935
00:50:26,929 --> 00:50:31,618
to be practiced 
 really so that
we can leverage and benefit a 

936
00:50:31,618 --> 00:50:34,016
lot from this AI revolution. 
Oh. 

937
00:50:36,096 --> 00:50:40,416
So there is a book called 
 
Philosophy of Software Design. 

938
00:50:40,911 --> 00:50:43,436
Yeah, John Ousterhout, yeah. 
I love that book. 

939
00:50:44,876 --> 00:50:48,736
And there's a paper, 1975, 
 
paper called No Silver Bullet. 

940
00:50:48,832 --> 00:50:50,151
Yeah. 
Fred Brooks, yeah. 

941
00:50:50,256 --> 00:50:51,596
Yes, love these two things, 
right? 

942
00:50:52,516 --> 00:50:55,416
And now then there is a-- there 
 are books around patterns and 

943
00:50:55,476 --> 00:50:57,696
software engineering and 
enterprise 
 architecture, all 

944
00:50:57,696 --> 00:51:01,190
that stuff. 
But it kind of comes down to 

945
00:51:01,190 --> 00:51:02,996
like 
 three or four basic 
principles. 

946
00:51:03,956 --> 00:51:07,236
For example, software 
 is about
composition. 

947
00:51:07,796 --> 00:51:11,804
That means you have to be able 
to build 
 software layer by 

948
00:51:11,804 --> 00:51:14,436
layer, brick by brick. 
That means abstraction. 

949
00:51:14,696 --> 00:51:18,276
So first one is abstraction. 
I always say, abstract 

950
00:51:18,276 --> 00:51:21,148
everything out. 
Like when I-- when people ask 

951
00:51:21,148 --> 00:51:23,716
me, 
 when people-- when 
sometimes I ask people, what are

952
00:51:23,716 --> 00:51:25,162
you building? 
They'll give me very big, and 
 

953
00:51:25,170 --> 00:51:27,886
you know, very big, fancy thing.
We are building this. 

954
00:51:28,096 --> 00:51:30,436
And I say, no, no, what 
 are 
you really building? 

955
00:51:31,656 --> 00:51:35,778
And if you go down like two or 
three or 
 five whys, basically 

956
00:51:35,778 --> 00:51:38,474
five whys, if you go through 
that, finally they'll figure 
 

957
00:51:38,482 --> 00:51:42,016
out you are-- they are actually 
building some CRM or some CMS or

958
00:51:42,016 --> 00:51:44,716
some SKU listing 
 thing or some
payment checkout stuff. 

959
00:51:45,416 --> 00:51:47,560
You put them all together and 
this becomes 
 fancy product, 

960
00:51:47,560 --> 00:51:50,856
which is great, right? 
Software engineers should think 

961
00:51:50,856 --> 00:51:52,642
like that. 
And we should now who are 
 

962
00:51:52,650 --> 00:51:54,996
builders, they should think like
that in terms of abstractions. 

963
00:51:55,176 --> 00:51:57,816
Build thing-- like build one 
thing which 
 does only one 

964
00:51:57,816 --> 00:52:00,381
thing, and then build one thing 
on top of it, like Linux 

965
00:52:00,381 --> 00:52:02,774
commands. 
If you want to count the files 


966
00:52:02,782 --> 00:52:07,896
on a directory, you say ls -l 
pipe to wc -l and you're done. 

967
00:52:07,896 --> 00:52:13,386
So one is that. 
Second is about, define correct 

968
00:52:13,386 --> 00:52:16,796
or like 
 domain-driven 
development, behavior-driven 

969
00:52:16,856 --> 00:52:18,766
development, test-driven 
development, 
 whatever you want

970
00:52:18,766 --> 00:52:24,726
to call it, do that. 
That will give you confidence 
 

971
00:52:24,734 --> 00:52:32,546
in the software produced, right?
Third, now I never liked PRs, 

972
00:52:32,546 --> 00:52:35,356
and I 
 don't like PRs still 
today. 

973
00:52:35,396 --> 00:52:38,696
There's no reason to do PR 
 
because now, now more so, 

974
00:52:38,696 --> 00:52:43,596
because my agent will generate 
like 
 three thousand lines in a

975
00:52:43,596 --> 00:52:44,901
day. 
Your agent will generate like 
 

976
00:52:44,909 --> 00:52:47,882
four thousand lines in a day. 
And we both do a PR, and then 

977
00:52:47,882 --> 00:52:50,491
they, 
 we'll pull our head and 
their agents will pull our head,

978
00:52:50,491 --> 00:52:52,936
and they'll do absolutely 
 
wrong things to merge those PRs.

979
00:52:53,636 --> 00:52:56,366
It just becomes a race 
 who 
checks in first. 

980
00:52:56,616 --> 00:52:58,573
Henry checked in five thirty, I 
 checked in five thir-- Oh my 

981
00:52:58,573 --> 00:53:00,247
God, I should have know-- I 
should have 
 known when he's 

982
00:53:00,247 --> 00:53:02,651
checking in, right? 
And this PR merge hell, 

983
00:53:02,651 --> 00:53:05,048
companies 
 are still going 
through that, but they don't 

984
00:53:05,048 --> 00:53:08,890
realize it. 
So now trunk-based development, 

985
00:53:08,890 --> 00:53:13,316
do a 
 branch, agreed, put 
autonomous code, merge the 

986
00:53:13,316 --> 00:53:14,976
branch, but don't do PRs. 
Doesn't matter. 

987
00:53:15,036 --> 00:53:17,528
Nobody's there to review. 
You are the one who is there to 

988
00:53:17,528 --> 00:53:19,403
review. 
So trunk-based development, 
 go

989
00:53:19,403 --> 00:53:24,679
move towards that. 
You do get commit a hash when 

990
00:53:24,679 --> 00:53:28,228
you do 
 a check-in, so you can 
actually pick your hash out and 

991
00:53:28,228 --> 00:53:31,378
do something, right? 
Fourth thing, feature flags 
 

992
00:53:31,386 --> 00:53:34,940
become very important. 
Super important now compared to 

993
00:53:34,940 --> 00:53:36,754
any time. 
Like check in your feature 
 

994
00:53:36,762 --> 00:53:39,119
with a feature flag. 
Check and disable or enable 
 

995
00:53:39,127 --> 00:53:42,319
based on your environment. 
So feature flags. 

996
00:53:45,159 --> 00:53:48,507
So we talked about abstraction, 
 we talked about spec, 

997
00:53:48,507 --> 00:53:51,103
trunk-based, spec and also then 
trunk-based 
 development, and 

998
00:53:51,103 --> 00:53:54,329
then feature flags. 
Fifth thing, which is very 

999
00:53:54,329 --> 00:53:58,479
important, is 
 that try to run 
the whole damn test suite. 

1000
00:53:59,779 --> 00:54:04,169
Like bring CI. 
Lot of people use CI and CD as a

1001
00:54:04,169 --> 00:54:07,221
buzzword. 
I want to again reiterate, CI 
 

1002
00:54:07,229 --> 00:54:10,994
and CD are two different things.
But because of industry and 
 

1003
00:54:11,002 --> 00:54:13,981
lot of things, we say CI/CD. 
Because continuous integration 

1004
00:54:13,981 --> 00:54:15,739
is very 
 different than 
continuous delivery. 

1005
00:54:16,519 --> 00:54:20,319
One validates your software and 
 gives you a producible build. 

1006
00:54:21,260 --> 00:54:23,199
Many people have forgotten 
 the
concept of golden build. 

1007
00:54:23,580 --> 00:54:27,274
Bring that concept back. 
And then CD, continuous 

1008
00:54:27,274 --> 00:54:30,734
delivery, 
 is about taking that
golden build and deploying it on

1009
00:54:30,734 --> 00:54:33,671
your server. 
CD does a last mile delivery, 

1010
00:54:33,671 --> 00:54:35,919
and 
 CI does a first mile 
verification. 

1011
00:54:36,799 --> 00:54:38,479
So these are the four or five 
things. 

1012
00:54:38,660 --> 00:54:41,620
If people do that, I 
 think 
that is good enough. 

1013
00:54:42,440 --> 00:54:44,260
And then there are many more 
things. 

1014
00:54:44,739 --> 00:54:48,343
But I live by these, and these 
are 
 the easier one to actually

1015
00:54:48,343 --> 00:54:49,532
crack. 
Yeah. 

1016
00:54:49,991 --> 00:54:53,194
So I think thanks for sharing 
all this 
 for listeners who are

1017
00:54:53,194 --> 00:54:57,072
curious how to actually excel in
this, you know, AI era. 

1018
00:54:57,411 --> 00:55:00,363
Some of these practices actually
has 
 been around for years and 

1019
00:55:00,363 --> 00:55:01,882
decades. 
Yes. 

1020
00:55:02,151 --> 00:55:06,800
But somehow we always tend 
 to 
either forget, mispractice or 

1021
00:55:06,800 --> 00:55:08,974
whatever that is, right? 
So I think thanks for sharing 

1022
00:55:08,974 --> 00:55:11,825
that. 
I know that you're also at the 

1023
00:55:11,825 --> 00:55:14,903
side, 
 right, building this, 
you know, the claw thing by 

1024
00:55:14,903 --> 00:55:18,222
yourself, you know? 
Some people call it vibe 
 

1025
00:55:18,230 --> 00:55:20,691
coding, some people call it solo
developer, whatever that is, 

1026
00:55:20,691 --> 00:55:23,866
right? 
How can you succeed 
 actually 

1027
00:55:23,866 --> 00:55:26,663
building something? 
I assume it's also not trivial, 

1028
00:55:26,663 --> 00:55:28,449
right? 
It's not like to-do lists and 
 

1029
00:55:28,457 --> 00:55:30,351
something like that, right? 
It's actually not trivial. 

1030
00:55:30,501 --> 00:55:33,451
It involves infrastructure and 

some kind of reliability thing 

1031
00:55:33,511 --> 00:55:37,532
if it doesn't work, right? 
So how would you advise 
 vibe 

1032
00:55:37,532 --> 00:55:40,981
coders out there? 
Do they actually need to 

1033
00:55:40,981 --> 00:55:42,451
practice 
 all this as well by 
themselves? 

1034
00:55:43,535 --> 00:55:47,815
At least first three. 
First three are the very 

1035
00:55:47,815 --> 00:55:53,119
important, 
 which is I think 
abstraction, specs- and the 

1036
00:55:53,119 --> 00:55:56,191
trunk-based thing. 
I think that is very important, 

1037
00:55:56,191 --> 00:55:57,995
 and I think that's how I built 
it. 

1038
00:55:58,235 --> 00:56:00,207
And you have used it, and using 
it. 

1039
00:56:00,547 --> 00:56:05,009
Yeah. 
And that is little bit of 

1040
00:56:05,009 --> 00:56:08,581
complex 
 piece of software 
because of, we dynamically 

1041
00:56:08,581 --> 00:56:11,875
prod-- like procure 
 
infrastructure, then we set up a

1042
00:56:11,935 --> 00:56:14,835
machine, then we put your 
OpenClaw there. 

1043
00:56:14,875 --> 00:56:17,533
We have baked image, then we 
take 
 your parameters, we set 

1044
00:56:17,533 --> 00:56:20,095
with OpenClaw. 
There is a proxy, there is a 

1045
00:56:20,095 --> 00:56:21,763
dynamic DNS. 
So there are a lot of 
 pieces, 

1046
00:56:21,763 --> 00:56:25,150
moving pieces. 
But the way I built it was very 

1047
00:56:25,150 --> 00:56:27,979
simple. 
I wrote about it, it's called 

1048
00:56:27,979 --> 00:56:31,235
the Small 
 Island of Delight, 
what OpenClaw is. 

1049
00:56:33,215 --> 00:56:35,899
And then there are-- I wrote 
 
next thing, What Ten Walls All 

1050
00:56:35,899 --> 00:56:39,299
Developers Feel and Face. 
So basically the way I built 
 

1051
00:56:39,307 --> 00:56:42,855
it, it started very small. 
First one was to just automate 


1052
00:56:42,863 --> 00:56:45,960
OpenClaw installation on my 
machine- so I can give that a 

1053
00:56:45,960 --> 00:56:47,665
script to my 
 friend, and he 
can, they can do it. 

1054
00:56:48,545 --> 00:56:52,967
But then once you have done 
that, then you have to make 
 

1055
00:56:52,975 --> 00:56:57,689
that machine, make the script 
execute using a web trigger. 

1056
00:56:57,749 --> 00:56:59,279
So you write a small web 
application. 

1057
00:56:59,289 --> 00:57:01,109
Then behind the scene, 
 you 
write a small worker. 

1058
00:57:01,656 --> 00:57:04,296
But now that worker needs to 
know 
 which machine, so you put

1059
00:57:04,296 --> 00:57:07,890
a job in the queue, and it goes 
to make API 
 call to cloud and 

1060
00:57:07,890 --> 00:57:10,175
get a VM for you. 
And that's how I built it. 

1061
00:57:10,836 --> 00:57:16,535
But what I did, I spent a lot 
 
of time, like these story cards.

1062
00:57:16,535 --> 00:57:18,015
I'll always carry the 
 bunch of
story cards. 

1063
00:57:18,095 --> 00:57:20,955
And I wrote, literally I wrote, 
 sat down and wrote the stories.

1064
00:57:22,682 --> 00:57:25,928
And then I-- And we know that 
 
we do this stacking of the story

1065
00:57:25,968 --> 00:57:29,195
and fan them out, like what are 
 dependents, what can are 

1066
00:57:29,195 --> 00:57:30,328
independent. 
I did all that stuff. 

1067
00:57:31,548 --> 00:57:35,050
To start with, I had like first,
first 
 day I had around like 

1068
00:57:35,050 --> 00:57:38,161
five cards. 
But next day they became 
 like 

1069
00:57:38,161 --> 00:57:42,413
12 or 15 cards. 
And then now we have like 300 

1070
00:57:42,413 --> 00:57:44,949
cards. 
We went through that. 

1071
00:57:45,349 --> 00:57:48,315
But then what I did, and I'll 
tell 
 you about that as well, I

1072
00:57:48,315 --> 00:57:53,632
did a role play with the agents,
but I used 
 issues, GitHub 

1073
00:57:53,632 --> 00:57:59,276
issues as the anchor. 
So I'll get a PM agent to read 


1074
00:57:59,284 --> 00:58:02,588
my description of card and then 
create a detailed card. 

1075
00:58:02,729 --> 00:58:05,676
Then I'll get a tech lead agent 
 to verify that card, and I 

1076
00:58:05,676 --> 00:58:08,078
will, I would use different 
Claude provider. 

1077
00:58:08,829 --> 00:58:11,753
And then I come back saying, 
"You got 
 as a…" And then that 

1078
00:58:11,753 --> 00:58:15,005
I have shown you that the PM 
agent actually, the tech 
 lead 

1079
00:58:15,005 --> 00:58:17,048
agent actually puts a comment on
it. 

1080
00:58:17,528 --> 00:58:20,986
Then I get one more agent 
saying, "Hey, 
 you are the 

1081
00:58:20,986 --> 00:58:22,644
implement developer. 
How would you implement?" 

1082
00:58:22,644 --> 00:58:25,371
Put the how as well. 
And then we implement the whole 

1083
00:58:25,371 --> 00:58:27,628
thing. 
And I have defined my agents.md,

1084
00:58:27,628 --> 00:58:30,809
which 
 talks about SDLC cycle, 
what they should do, what they 

1085
00:58:30,809 --> 00:58:33,208
should not do. 
Always, always deploy in the 

1086
00:58:33,208 --> 00:58:34,748
staging. 
Always. 

1087
00:58:35,549 --> 00:58:40,286
Never deploy in production. 
If you're… I deploy to 

1088
00:58:40,286 --> 00:58:42,161
production. 
I never ask Claude to deploy 

1089
00:58:42,161 --> 00:58:43,852
production. 
I always click that button 

1090
00:58:43,852 --> 00:58:45,524
myself. 
So I had to click that 
 button 

1091
00:58:45,524 --> 00:58:47,768
15 times a day. 
So I set up a CI over there. 

1092
00:58:48,108 --> 00:58:49,469
So that's how. 
Very small. 

1093
00:58:49,649 --> 00:58:53,108
But I did not write code. 
I did not write code. 

1094
00:58:54,448 --> 00:58:56,528
That is amazing actually. 
I'm very impressed with that. 

1095
00:58:57,532 --> 00:59:03,404
So what makes it possible for 
 
someone who can let go writing 

1096
00:59:03,404 --> 00:59:05,988
code? 
You know, like, because I 
 see 

1097
00:59:05,988 --> 00:59:08,717
these extremes, right? 
Some people even like in big 
 

1098
00:59:08,725 --> 00:59:11,228
companies like Anthropic, 
Spotify, they say no developers 

1099
00:59:11,228 --> 00:59:13,352
writing 
 code anymore by hand, 
right? 

1100
00:59:13,652 --> 00:59:15,633
But some people still 
 think 
it's not possible. 

1101
00:59:16,152 --> 00:59:18,732
AI hallucinate. 
There will be incorrect code. 

1102
00:59:19,332 --> 00:59:21,270
What to do about it? 
Like maybe you can share your 

1103
00:59:21,270 --> 00:59:23,500
experience. 
How do you actually kind of like

1104
00:59:23,500 --> 00:59:26,332
let go 
 this, you know, having 
to deal with code? 

1105
00:59:26,344 --> 00:59:34,325
First, I actually, that's when 
I-- honestly, that's 
 when I 

1106
00:59:34,325 --> 00:59:36,585
realized that judgment layer is 
very important. 

1107
00:59:39,101 --> 00:59:43,101
First 10 days in February were 
nuts. 

1108
00:59:45,481 --> 00:59:49,465
I was trying to make it work 
 
and because, like, the way… And 

1109
00:59:49,465 --> 00:59:52,131
of course, I improved. 
As I told you, first day five 
 

1110
00:59:52,139 --> 00:59:55,179
cards, 10 cards were there. 
But those 10 cards didn't have 


1111
00:59:55,187 --> 00:59:56,901
anything to say how will I 
verify. 

1112
00:59:57,221 --> 00:59:59,100
What process I'm telling you 
right 
 now after like, it's 

1113
00:59:59,100 --> 01:00:03,536
like after, like, so many days 
and so many iterations 
 going 

1114
01:00:03,536 --> 01:00:06,932
through like 300 cards, right? 
And then one day it struck to 

1115
01:00:06,932 --> 01:00:09,261
me, no, 
 no, no, I'm not gonna 
do this way. 

1116
01:00:09,841 --> 01:00:12,861
I'm gonna do my way. 
I'm outsourcing typing. 

1117
01:00:13,851 --> 01:00:18,184
That's how I did it. 
And I tested the system and all 

1118
01:00:18,184 --> 01:00:22,206
the stuff. 
So that check and balance, 
 so 

1119
01:00:22,206 --> 01:00:25,701
people who are not ready to let 
it go, think about it. 

1120
01:00:26,021 --> 01:00:27,781
If you're not typing, then 
 
what would you be doing? 

1121
01:00:29,011 --> 01:00:31,862
And that's what I thought. 
And if I'm not typing, I'm… 

1122
01:00:31,862 --> 01:00:36,187
Well, that 
 guy or that it 
thing or it or guy or girl, I 

1123
01:00:36,187 --> 01:00:38,666
don't know what it should be 
called, 
 but it sometimes is 

1124
01:00:38,666 --> 01:00:40,341
very amazing, huh? 
Like it talks to you. 

1125
01:00:40,421 --> 01:00:41,921
Oh, I'm sorry, I did not do 
this. 

1126
01:00:41,971 --> 01:00:44,453
It's like so amazing sometimes. 
Like, oh, wait a minute, 
 we 

1127
01:00:44,453 --> 01:00:46,461
should have done that. 
Like, okay, fine, whatever. 

1128
01:00:48,821 --> 01:00:53,329
So put it in sandbox. 
First thing is that, put it in 

1129
01:00:53,329 --> 01:00:57,537
sandbox. 
Have your exact staging, 
 

1130
01:00:57,545 --> 01:00:59,541
production-like repli- 
environment. 

1131
01:00:59,921 --> 01:01:02,571
So I have two staging, 
 two 
staging environments. 

1132
01:01:02,921 --> 01:01:04,221
I tear them down and bring them 
up. 

1133
01:01:05,441 --> 01:01:07,513
Do automation. 
Actually, one more point is very

1134
01:01:07,513 --> 01:01:10,021

 important, is do automation. 
Automation is very important. 

1135
01:01:10,281 --> 01:01:12,401
Do automation on all small 
 
things, so you don't have to do.

1136
01:01:12,801 --> 01:01:16,617
So in my software development 
 
lifecycle, only automation I 

1137
01:01:16,617 --> 01:01:18,921
don't have is whichever I go and
deploy. 

1138
01:01:19,761 --> 01:01:21,251
Apart from that, everything 
 is
automated, right? 

1139
01:01:21,681 --> 01:01:26,207
So given that, if you do a 
sandbox, 
 if you abstract your 

1140
01:01:26,207 --> 01:01:29,298
small things out, if you define 
correctness, if 
 you spend a 

1141
01:01:29,298 --> 01:01:33,221
lot of time on judgment layer, I
think it does not hallucinate. 

1142
01:01:34,401 --> 01:01:37,788
Now, at least. 
Yeah, I think the model has 

1143
01:01:37,788 --> 01:01:40,361
become 
 so advanced now, and 
especially if you have a good 

1144
01:01:40,361 --> 01:01:42,359
harness tools, 
 right, like 
Claude Code and all that. 

1145
01:01:42,399 --> 01:01:46,358
I believe, yes, hallucinations 
still 
 happen, but if you build

1146
01:01:46,358 --> 01:01:49,862
this correctness layer, right, 
your spec, your TDD, the 
 

1147
01:01:49,870 --> 01:01:53,485
things that can verify the 
behaviors accurately, I think, 

1148
01:01:53,485 --> 01:01:58,497
yeah, this 
 thing tends to 
produce accurate… And I am like 

1149
01:01:58,497 --> 01:02:00,954
very thankful to 
 all my users 
of ClawStation. 

1150
01:02:00,954 --> 01:02:03,970
I not-- I don't have hundreds 
and 
 thousands, but whatever I 

1151
01:02:03,970 --> 01:02:07,119
have, whenever they ask me 
something, I'm 
 able to 

1152
01:02:07,119 --> 01:02:09,616
actually respond to them and 
finish fix their stuff, whatever

1153
01:02:09,616 --> 01:02:11,758
it is. 
And it's pretty interesting 
 

1154
01:02:11,766 --> 01:02:14,844
because I have exact replica of 
production running with me. 

1155
01:02:15,744 --> 01:02:17,488
So that's where I am. 
Yeah. 

1156
01:02:17,729 --> 01:02:20,058
Hats off to you. 
Again, like I don't know how you

1157
01:02:20,058 --> 01:02:21,348

 do it, one single person 
building. 

1158
01:02:21,428 --> 01:02:23,765
I think it's quite complex to 
 
build something like a 

1159
01:02:23,765 --> 01:02:26,148
ClawStation. 
And yeah, I've been playing 

1160
01:02:26,148 --> 01:02:28,028
around 
 with it, so thanks for 
creating that. 

1161
01:02:29,128 --> 01:02:32,628
One thing also you mentioned, I 
mean, not 
 just software 

1162
01:02:32,628 --> 01:02:34,890
developers, organizations are 
going through this moment as 

1163
01:02:34,890 --> 01:02:38,008
well, 
 and they need to think 
how to implement, apply AI in 

1164
01:02:38,008 --> 01:02:40,328
their organization. 
One thing very interesting 
 you

1165
01:02:40,328 --> 01:02:43,011
mentioned earlier, right? 
So many people just think of 

1166
01:02:43,011 --> 01:02:45,427
automating, 
 you know, maybe 
something on the local or maybe 

1167
01:02:45,427 --> 01:02:47,508
tasks that they need to do, 
right? 

1168
01:02:47,708 --> 01:02:50,428
But never thinking holistically 
 in terms of maybe putting 

1169
01:02:50,528 --> 01:02:54,122
agents centrally and all that. 
And statistically, I've read 

1170
01:02:54,122 --> 01:02:58,280
also 
 in, you know, some 
articles, maybe 95% of 

1171
01:02:58,280 --> 01:03:01,488
organization that apply AI 
 
doesn't actually reap the 

1172
01:03:01,488 --> 01:03:03,256
benefits. 
Like there's no ROI coming out 

1173
01:03:03,256 --> 01:03:05,868
of this. 
So tell us what's wrong. 

1174
01:03:06,238 --> 01:03:08,808
Like…how come? 
I'll tell you what's wrong. 

1175
01:03:11,508 --> 01:03:18,609
Around 2000, year 2000, a lot 
 
of organizations started moving 

1176
01:03:18,668 --> 01:03:22,668
towards electronic forms. 
And I'll give a simple example. 

1177
01:03:23,468 --> 01:03:28,168
So electronic forms getting… 
 
I'm your customer, right? 

1178
01:03:29,648 --> 01:03:31,928
Earlier you had a physical 
 
form to get complaint. 

1179
01:03:32,009 --> 01:03:33,448
So what will a physical form 
will have? 

1180
01:03:33,769 --> 01:03:37,166
My name, I had to fill my name, 
 my date of birth, where I 

1181
01:03:37,166 --> 01:03:38,928
think, and I put a complaint 
over there. 

1182
01:03:39,828 --> 01:03:44,112
Now, what happened in early 
days, 
 those forms became 

1183
01:03:44,112 --> 01:03:47,479
digitized. 
But I'm your customer, so 

1184
01:03:47,479 --> 01:03:50,217
whoever designed 
 that form 
forgot that they don't have to 

1185
01:03:50,217 --> 01:03:54,628
ask me where I live every time. 
Because when I log in and set 
 

1186
01:03:54,636 --> 01:03:57,328
up my profile, that information 
already there with you. 

1187
01:03:57,708 --> 01:04:01,728
So if you look at earlier 
 
forms, a lot of information were

1188
01:04:01,808 --> 01:04:05,336
again repeated again and again. 
And I'm like, "Why?" 

1189
01:04:05,336 --> 01:04:08,556
And this 
 happens in a lot of 
forms still when you get bank 

1190
01:04:08,556 --> 01:04:12,364
forms and stuff. 
Those bank guys know my name. 

1191
01:04:13,308 --> 01:04:14,808
But why I have to fill it every 
time? 

1192
01:04:14,848 --> 01:04:16,208
Why I have to do DocuSign every 
time? 

1193
01:04:16,468 --> 01:04:19,764
Why can't they do something? 
So that digitization took time. 

1194
01:04:20,248 --> 01:04:23,975
Like I'm s- like we are living 

in Singapore, and I actually see

1195
01:04:23,975 --> 01:04:26,103
GovTech has done some 
 
fantastic job around that, 

1196
01:04:26,103 --> 01:04:28,769
right? 
Like whole Singpass, whole ICA 

1197
01:04:28,769 --> 01:04:30,488
and all. 
So it's amazing, right? 

1198
01:04:31,068 --> 01:04:34,049
So given that, but that 
 
thought process came later. 

1199
01:04:34,089 --> 01:04:37,489
So earlier what was, "Oh, it is 
a physical 
 form. 

1200
01:04:37,489 --> 01:04:40,598
Let's capture this 
electronically as it is without 

1201
01:04:40,598 --> 01:04:44,060
re-rethinking the workflow." 
That's exactly what is 
 

1202
01:04:44,068 --> 01:04:49,621
happening right now. 
We had-- we have, now we have a 

1203
01:04:49,621 --> 01:04:52,765
 smart, reasonable reasoning and
thinking agents which can 

1204
01:04:52,765 --> 01:04:54,921
produce. 
But let's not change anything 

1205
01:04:54,921 --> 01:04:59,413
else. 
So we have a trick or technique 

1206
01:04:59,413 --> 01:05:01,621
called 
 value stream mapping, 
which you know. 

1207
01:05:01,621 --> 01:05:03,941
A lot of organizations are not 
doing that. 

1208
01:05:04,821 --> 01:05:06,201
They are not doing value stream 
mapping. 

1209
01:05:06,801 --> 01:05:10,211
They are not thinking, 
rethinking, 
 saying, "If we 

1210
01:05:10,211 --> 01:05:15,361
keep AI in the center, then how 
would we behave?" 

1211
01:05:15,361 --> 01:05:19,145
And that's 
 where we start 
thinking about like how does-- 

1212
01:05:19,145 --> 01:05:22,641
how do you behave smartly? 
What can you offload? 

1213
01:05:22,981 --> 01:05:25,981
What you shouldn't offload? 
So people are not doing that. 

1214
01:05:26,071 --> 01:05:30,946
Till the time people will keep 

using AI as a, what to say, 

1215
01:05:30,946 --> 01:05:32,421
periph-- I'll give you a simple 
example. 

1216
01:05:32,501 --> 01:05:35,301
One more example. 
You have microwave at home. 

1217
01:05:35,440 --> 01:05:37,799
Yeah. 
How many buttons you use on 

1218
01:05:37,799 --> 01:05:39,812
that? 
Maybe just one, two. 

1219
01:05:40,032 --> 01:05:41,981
Two, maybe max. 
How many buttons it has? 

1220
01:05:42,532 --> 01:05:44,052
A lot. 
A lot more. 

1221
01:05:44,113 --> 01:05:47,041
I mean, like maybe 10 or 12. 
And now they're on internet. 

1222
01:05:47,041 --> 01:05:49,696
Like, oh, yeah, like you can 
 
start microwave using your 

1223
01:05:49,696 --> 01:05:51,459
phone. 
Yeah, but I have to go still put

1224
01:05:51,459 --> 01:05:52,551
the 
 thing inside the 
microwave, right? 

1225
01:05:53,161 --> 01:05:55,993
There's a joke by one Indian, 
one Indian 
 comedian, and I 

1226
01:05:55,993 --> 01:05:59,601
love that joke, right? 
And, same like this. 

1227
01:06:00,241 --> 01:06:03,327
Like if you can build, you can 
have 
 all the features, but if 

1228
01:06:03,327 --> 01:06:06,859
you don't keep people in mind, 
then you can 
 keep building 

1229
01:06:06,859 --> 01:06:10,626
useless features. 
So sometimes it is about 
 

1230
01:06:10,634 --> 01:06:15,273
reducing what you already doing.
For example, the physical form 


1231
01:06:15,281 --> 01:06:18,165
always had the profile or user 
ID, name, address, and all 

1232
01:06:18,165 --> 01:06:19,431
stuff. 
You have to reduce that 
 

1233
01:06:19,439 --> 01:06:21,421
because you don't need that. 
Same with AI. 

1234
01:06:22,361 --> 01:06:23,621
You have to reduce bunch of 
things. 

1235
01:06:23,821 --> 01:06:27,217
You don't need those many. 
So now you have person one, he 

1236
01:06:27,217 --> 01:06:30,434
does 
 something, then person 
two does something, he, she does

1237
01:06:30,434 --> 01:06:32,691
something, and then person 
 
three does something, she does 

1238
01:06:32,691 --> 01:06:34,535
something. 
You actually have person one, 
 

1239
01:06:34,543 --> 01:06:36,665
two, three sitting over there 
and doing the same thing, but 

1240
01:06:36,665 --> 01:06:37,881
they 
 are doing with the use of
AI. 

1241
01:06:38,741 --> 01:06:41,071
But they're still passing 
 the 
buck from here to here. 

1242
01:06:41,481 --> 01:06:45,561
While actually person one can 
 
do this thing with three agents.

1243
01:06:47,001 --> 01:06:49,381
Why we are not doing that? 
So that is where they're not. 

1244
01:06:49,391 --> 01:06:54,267
They're not understanding how to

 keep the AI in center and 

1245
01:06:54,267 --> 01:06:57,541
organize organizations, 
workflows around it. 

1246
01:06:58,571 --> 01:07:02,201
So one of the most important 
thing 
 people have to think 

1247
01:07:02,201 --> 01:07:05,587
through, any organizations has--
only thing they have 
 is 

1248
01:07:05,587 --> 01:07:08,085
workflow, workflow, workflow 
across the things when they're 

1249
01:07:08,085 --> 01:07:10,461
serving somebody. 
They just need to think that. 

1250
01:07:11,181 --> 01:07:13,921
And if they can do that, then 
 
things will become better. 

1251
01:07:13,971 --> 01:07:16,691
For example, our early, like 
most popular 
 software 

1252
01:07:16,691 --> 01:07:21,076
engineering workflow is this. 
Ajey works with Henry, Henry 
 

1253
01:07:21,084 --> 01:07:24,051
is Ajey's tech lead, Ajey is 
Henry's junior developer. 

1254
01:07:24,651 --> 01:07:28,415
Ajey creates a code, Henry gives
a 
 PR, Henry reviews the PR as 

1255
01:07:28,415 --> 01:07:31,428
if Henry has a compiler fit in 
his eyes. 

1256
01:07:32,132 --> 01:07:35,112
Never checks out, says, "Okay, 

it is good," or puts a comment. 

1257
01:07:35,972 --> 01:07:39,352
And then, Henry merges the 
 
code and then it goes to CI. 

1258
01:07:40,533 --> 01:07:44,592
Now, if I am a agent, why 
 you 
need to do that again? 

1259
01:07:45,992 --> 01:07:49,562
What you can do is you can tell 
me 
 as Ajey, as agent saying, 

1260
01:07:49,562 --> 01:07:53,149
"Ajey, do this, but once you do 
this, and once 
 you run this, 

1261
01:07:53,149 --> 01:07:55,107
make sure this, this, this 
happens." 

1262
01:07:55,107 --> 01:07:58,932
Or tell somebody else. 
Two agents, they can run. 

1263
01:07:58,972 --> 01:08:00,232
You can have five agents 
 
running on your Mac. 

1264
01:08:00,812 --> 01:08:03,060
But we have not done that. 
We're still creating branches 

1265
01:08:03,060 --> 01:08:04,132
with agents. 
Can you believe that? 

1266
01:08:04,912 --> 01:08:06,292
We're still creating PRs with 
agents. 

1267
01:08:06,612 --> 01:08:07,812
Can you believe that? 
Why? 

1268
01:08:09,452 --> 01:08:13,076
If I am the verifier and if I am
the 
 executor, then why I'm 

1269
01:08:13,076 --> 01:08:15,494
doing this? 
So they have to-- Like, they're 

1270
01:08:15,494 --> 01:08:18,812
 still using it as a minion 
rather than the builder. 

1271
01:08:20,442 --> 01:08:24,915
Like execution is gone. 
So let it go completely, you 

1272
01:08:24,915 --> 01:08:26,147
know. 
So that's what is happening 
 

1273
01:08:26,154 --> 01:08:28,015
all over the place. 
Yeah. 

1274
01:08:28,616 --> 01:08:31,540
So I think definitely this 
requires like 
 maybe, I don't 

1275
01:08:31,540 --> 01:08:33,122
know, paradigm shift. 
Yeah, yeah. 

1276
01:08:33,122 --> 01:08:34,696
Thinking differently. 
Yeah. 

1277
01:08:35,057 --> 01:08:36,926
Coming back to like value 
 
stream mapping and all that. 

1278
01:08:37,406 --> 01:08:41,992
I think the intention is not to 
 automate what you are doing 

1279
01:08:41,992 --> 01:08:45,426
now, but also holistically 
thinking how can you 
 let 

1280
01:08:45,426 --> 01:08:48,999
agents run by itself- natively, 
like, as if like- AI is the 

1281
01:08:48,999 --> 01:08:52,295
center. 
For leaders out there, I think 


1282
01:08:52,303 --> 01:08:57,456
many are thinking AI means I can
reduce a lot of people, layoff. 

1283
01:08:59,796 --> 01:09:02,285
Any thoughts about that? 
Because I'm sure, yeah, maybe 
 

1284
01:09:02,292 --> 01:09:05,310
in the beginning you can do a 
lot of efficiency and all that, 

1285
01:09:05,310 --> 01:09:08,309
 but what are, what is the risk?
What are the dangers of doing it

1286
01:09:08,309 --> 01:09:12,416
this way? 
I think if you apply our way of 

1287
01:09:12,416 --> 01:09:14,399
doing 
 things in software 
engineering, you can always 

1288
01:09:14,399 --> 01:09:21,997
already reduce a lot of people. 
Yeah. it's not about people 

1289
01:09:21,997 --> 01:09:25,526
reduction, 
 it's about basic 
job going away. 

1290
01:09:26,286 --> 01:09:29,591
So automation can certainly 
reduce 
 people, but they will 

1291
01:09:29,591 --> 01:09:31,786
have to certainly hire more 
people as well. 

1292
01:09:32,984 --> 01:09:36,564
Like how much ever people are 
doing hoopla 
 about this, there

1293
01:09:36,564 --> 01:09:40,350
are still people hiring. 
Why they are hiring if the 

1294
01:09:40,350 --> 01:09:41,404
people 
 don't need software 
developers? 

1295
01:09:42,365 --> 01:09:45,983
Like big companies are still 
hiring 
 for different reasons, 

1296
01:09:45,983 --> 01:09:48,510
right? 
So before reducing the people, 


1297
01:09:48,518 --> 01:09:51,941
it is not like that AI is going 
to do five people's job. 

1298
01:09:52,001 --> 01:09:54,081
It's not. 
Again, that's my thing. 

1299
01:09:54,220 --> 01:09:55,341
Management has to think through 
this. 

1300
01:09:55,361 --> 01:09:57,061
Like, it is flipping on its 
head. 

1301
01:09:57,341 --> 01:09:59,562
Yes, you don't need one-- five 
developers. 

1302
01:09:59,921 --> 01:10:03,402
You need only two maybe. 
But now you need, instead of 

1303
01:10:03,402 --> 01:10:05,907
having 
 two BAs or two product 
managers or two product 

1304
01:10:05,907 --> 01:10:09,661
managers, you need four or five.
Because now you can execute 
 

1305
01:10:09,669 --> 01:10:12,201
five things in parallel. 
So what happens? 

1306
01:10:12,241 --> 01:10:16,691
Like, how are you gonna do that?
What will happen is you will 
 

1307
01:10:16,699 --> 01:10:19,537
actually reduce your output, 
that this person is still not 

1308
01:10:19,537 --> 01:10:21,961
able to 
 produce more things 
for this person. 

1309
01:10:23,371 --> 01:10:27,688
So organizations, yes, you can 
certainly 
 reduce people on the

1310
01:10:27,688 --> 01:10:30,541
very basic structure. 
See, the thing is very simple. 

1311
01:10:30,961 --> 01:10:35,116
What is getting replaced? 
The most structured things 
 are

1312
01:10:35,116 --> 01:10:37,935
getting replaced. 
Programming language 
 are the 

1313
01:10:37,935 --> 01:10:40,681
most structured. 
So that's why LLMs can 
 

1314
01:10:40,689 --> 01:10:44,161
reproduce them very much better.
That is first most structured. 

1315
01:10:44,381 --> 01:10:47,701
Second most structured, maybe 
finance. 

1316
01:10:48,881 --> 01:10:51,161
Third most structured, maybe 
insurance. 

1317
01:10:51,801 --> 01:10:54,241
Then fourth, little bit 
 
subjectivity, comes the law. 

1318
01:10:54,881 --> 01:10:57,541
But still there is some-- So 
 
anywhere the structure is-- 

1319
01:10:57,841 --> 01:10:59,981
structured work is getting 
replaced. 

1320
01:11:00,781 --> 01:11:04,486
Cognitive work is not. 
The organizations should, if 

1321
01:11:04,486 --> 01:11:07,978
they 
 are reducing it, it's 
fine, but they are creating 

1322
01:11:07,978 --> 01:11:10,206
cognitive gap. 
They can't live in that vacuum, 

1323
01:11:10,206 --> 01:11:11,561
so 
 they need to think through 
that. 

1324
01:11:11,762 --> 01:11:14,421
They are not gonna go faster. 
They're gonna go same speed. 

1325
01:11:14,461 --> 01:11:16,698
That's all. 
If they are okay with that 

1326
01:11:16,698 --> 01:11:18,581
speed, 
 yeah, let people go, 
it's fine. 

1327
01:11:18,902 --> 01:11:21,573
But if you want to accelerate 
something, 
 then instead of 

1328
01:11:21,573 --> 01:11:24,530
letting people go, think through
and enable them with more 

1329
01:11:24,530 --> 01:11:25,781
skills. 
Yeah. 

1330
01:11:26,445 --> 01:11:28,125
So I think that's a 
 very good 
point, right? 

1331
01:11:28,205 --> 01:11:30,285
Instead of efficiency only, 
right? 

1332
01:11:30,405 --> 01:11:33,441
Think about how many innovations
can you 
 do if you upskill 

1333
01:11:33,441 --> 01:11:36,765
these people, let them build 
more, you know, innovate more. 

1334
01:11:37,285 --> 01:11:40,885
And I think, yeah, especially 
now 
 with AI, many other 

1335
01:11:40,885 --> 01:11:43,414
companies, small teams can 
actually disrupt, 
 you know, 

1336
01:11:43,414 --> 01:11:44,329
the incumbents. 
Yeah. 

1337
01:11:44,950 --> 01:11:47,595
And tokens are expensive, man. 
Yeah. 

1338
01:11:48,809 --> 01:11:50,469
People don't understand 
 tokens
are expensive. 

1339
01:11:51,249 --> 01:11:52,509
People are burning a lot of 
tokens. 

1340
01:11:52,749 --> 01:11:54,059
They're expensive and it's not a
pride. 

1341
01:11:54,329 --> 01:11:57,629
It's not a pride. 
And sometimes you don't have to 

1342
01:11:57,629 --> 01:12:00,619
 have tokens burning machines. 
And it's okay. 

1343
01:12:01,019 --> 01:12:03,850
So organization have to 
 think 
a balanced approach. 

1344
01:12:04,730 --> 01:12:06,813
Yeah. 
And especially now we know that 

1345
01:12:06,813 --> 01:12:09,849
 tokens are still subsidized 
because of a lot of investments.

1346
01:12:09,901 --> 01:12:12,854
Yes. 
And, oh, like people are still 

1347
01:12:12,854 --> 01:12:17,007
using-- 
 people want to use 
Opus 4.6 max effort for like 

1348
01:12:17,007 --> 01:12:19,046
writing a getter and setter. 
Why? 

1349
01:12:19,961 --> 01:12:21,025
So yeah, man. 
Yeah. 

1350
01:12:21,705 --> 01:12:24,211
So what do you think the 
implications 
 of AI also for 

1351
01:12:24,211 --> 01:12:27,345
Southeast Asian? 
Because we know the models all 


1352
01:12:27,353 --> 01:12:31,134
are coming mostly from Western. 
There are some open models from 

1353
01:12:31,134 --> 01:12:34,286
China. 
But for Southeast Asia, what do 

1354
01:12:34,286 --> 01:12:36,826
you think? 
The tech scene is gonna… We have

1355
01:12:36,826 --> 01:12:37,950
always been consumer, by the 
way. 

1356
01:12:39,870 --> 01:12:43,175
We have not been, in terms of 
 
software, we have always been 

1357
01:12:43,175 --> 01:12:45,372
consumer. 
Like not only AI models, 
 like 

1358
01:12:45,372 --> 01:12:47,820
my watch, your watch. 
My phone, your phone. 

1359
01:12:49,690 --> 01:12:54,040
Everything around us is either 

Western or Chinese, right? 

1360
01:12:55,130 --> 01:12:57,420
Because those are the hubs, and 
 we have been always a consumer.

1361
01:12:57,620 --> 01:12:59,650
We are doing some good stuff 
over here. 

1362
01:13:01,150 --> 01:13:06,524
And that will change. 
But our-- Like, if you look 
 at

1363
01:13:06,524 --> 01:13:09,670
Southeast Asia, the economy is 
mostly non-tech economy. 

1364
01:13:10,010 --> 01:13:11,290
It's mostly financial and other 
things. 

1365
01:13:11,350 --> 01:13:14,504
So there will be some 
automation, 
 but not a lot will

1366
01:13:14,504 --> 01:13:17,630
happen. 
And the non-tech economy, the 

1367
01:13:17,630 --> 01:13:21,857
economy 
 with more human 
dexterity will remain a little 

1368
01:13:21,857 --> 01:13:26,275
bit unchanged for now. 
If you have jobs which has human

1369
01:13:26,275 --> 01:13:28,390

 dexterity, your job is okay. 
You're okay. 

1370
01:13:28,500 --> 01:13:30,370
You're like, if you're a 
 
physiotherapist, you're fine. 

1371
01:13:32,461 --> 01:13:34,430
The robots coming. 
Yeah, no. 

1372
01:13:34,690 --> 01:13:37,513
If you're-- Like, I'm not gonna 
 put my head in like a robot 

1373
01:13:37,513 --> 01:13:39,472
barber. 
Like, they have, like, it's like

1374
01:13:39,472 --> 01:13:41,130

 bloody, like scissors over 
there, man. 

1375
01:13:41,410 --> 01:13:43,133
What are you talking about? 
I'm not gonna put my head over 

1376
01:13:43,133 --> 01:13:44,550
there. 
No, no, no, no, no. 

1377
01:13:44,860 --> 01:13:47,476
One scissor goes wrong, like I 
cut 
 your ear, I'm sorry, I'll 

1378
01:13:47,476 --> 01:13:49,283
do it better. 
I'll cut-- Next time, I'll cut 

1379
01:13:49,283 --> 01:13:49,901
both. 
Yeah. 

1380
01:13:51,420 --> 01:13:56,204
No, I think our industries 
 
will get more modern. 

1381
01:13:56,970 --> 01:14:00,828
I think it's good for us. 
I again, say, I see it as a boon

1382
01:14:00,828 --> 01:14:03,197
for us. 
It is very good boon for a lot 

1383
01:14:03,197 --> 01:14:08,160
of entrepreneurs and builders. 
My view is that people who have 

1384
01:14:08,160 --> 01:14:10,808
 been thinking about building, 
this is the best time to build 

1385
01:14:10,808 --> 01:14:12,722
software. 
It is best time to build 

1386
01:14:12,722 --> 01:14:14,385
automation. 
It's the best time to put-- 

1387
01:14:14,385 --> 01:14:17,480
Like, you 
 want to do a drop 
shipping and you did not know 

1388
01:14:17,480 --> 01:14:19,836
how to build a website and 
 
payment system, it's the best 

1389
01:14:19,836 --> 01:14:21,250
time. 
Just go ahead and do it. 

1390
01:14:21,310 --> 01:14:23,930
Like put fifteen drop shipping 

websites on the thing. 

1391
01:14:24,270 --> 01:14:27,250
Something will sell. 
It's a good time, man. 

1392
01:14:27,370 --> 01:14:31,982
Like, I see it and, we are-- 
since 
 we are consumer economy,

1393
01:14:31,982 --> 01:14:38,790
we will get benefit out of it, 
because the AI adoption and all 

1394
01:14:38,790 --> 01:14:41,030
this tech 
 will make our 
experiences better. 

1395
01:14:41,950 --> 01:14:44,330
There will be many people who 
 
will actually work on building 

1396
01:14:44,350 --> 01:14:48,042
those experiences for us. 
And there will be better 

1397
01:14:48,042 --> 01:14:50,046
products. 
So I think it is a good chance 

1398
01:14:50,046 --> 01:14:53,160
for us. 
It's a good chance for all the 


1399
01:14:53,168 --> 01:14:58,363
people who have been thinking 
that West was or China or Japan 

1400
01:14:58,363 --> 01:15:02,804
or 
 Australia or UK was the 
advantageous place and no level 

1401
01:15:02,804 --> 01:15:04,426
playing field. 
It's a level playing field for 

1402
01:15:04,426 --> 01:15:07,106
us now. 
I'll tell you fifteen, twenty 

1403
01:15:07,106 --> 01:15:09,736
years back 
 when I used to go 
to US, people used to ask me: 

1404
01:15:09,736 --> 01:15:11,120
"Can you bring this? 
Can you bring 
 that? 

1405
01:15:11,120 --> 01:15:13,065
Can you bring that?" 
Now nobody asks me that because 

1406
01:15:13,065 --> 01:15:15,004
everything is available. 
Same thing is gonna happen 
 

1407
01:15:15,012 --> 01:15:18,770
with the technology. 
Now we have very good chance and

1408
01:15:18,770 --> 01:15:22,248

 advantage to create things 
over here and give it to our 

1409
01:15:22,248 --> 01:15:24,390
consumers, 
 and go to global 
scale as well. 

1410
01:15:24,650 --> 01:15:27,169
Like if you want to use 
software, you can 
 use my 

1411
01:15:27,169 --> 01:15:28,490
software anywhere in the world. 
Yeah. 

1412
01:15:28,730 --> 01:15:29,677
Why not? 
Right. 

1413
01:15:30,737 --> 01:15:35,078
So one thing that other, like 
 
some other people also think 

1414
01:15:35,158 --> 01:15:37,611
about using AI a lot more is 
that 
 we're actually not 

1415
01:15:37,611 --> 01:15:40,017
learning a lot. 
Because almost everything now 

1416
01:15:40,017 --> 01:15:42,417
you 
 can ask and you get the 
answer. 

1417
01:15:42,957 --> 01:15:46,444
Like fundamentally, you know, in
terms 
 of learning, in terms of

1418
01:15:46,444 --> 01:15:50,467
going through that hardship to 
actually, you know, 
 understand

1419
01:15:50,467 --> 01:15:54,627
principally what is going on, I 
think it's getting reduced 

1420
01:15:54,627 --> 01:15:57,357
because it's 
 so easy to just 
do something, you know. 

1421
01:15:57,797 --> 01:16:00,737
We're talking about creating 
 
stuff, let the agent do that. 

1422
01:16:02,018 --> 01:16:05,733
What do you think people should 
do 
 in order not to lose this 

1423
01:16:05,733 --> 01:16:08,705
aspect? 
They should not do that. 

1424
01:16:09,305 --> 01:16:11,936
Basically, I have found very few
people 
 who actually learned 

1425
01:16:11,936 --> 01:16:14,885
piano using YouTube. 
You still need an instructor. 

1426
01:16:16,705 --> 01:16:21,594
You still need a coach. 
So I have found if you are 

1427
01:16:21,594 --> 01:16:26,182
trying to 
 do things and the 
more far-fetched away you are, 

1428
01:16:26,182 --> 01:16:31,515
at some point of time 
 you will
hit your limit, and that point 

1429
01:16:31,515 --> 01:16:34,623
of time you will be redundant. 
Especially in cognitive 
 

1430
01:16:34,631 --> 01:16:37,640
industries, especially in 
knowledge workers- not anybody 

1431
01:16:37,640 --> 01:16:39,866
else. 
So I think people should not. 

1432
01:16:40,036 --> 01:16:42,026
That's what is happening. 
Actually, if you look what's 

1433
01:16:42,026 --> 01:16:43,865
happening 
 to software 
developers, because they stopped

1434
01:16:43,865 --> 01:16:46,113
learning somewhere. 
They're like, "Oh, let's do 

1435
01:16:46,113 --> 01:16:48,265
this." 
 
But they did not learn how to 

1436
01:16:48,325 --> 01:16:50,825
understand and learn about 
models 
 and how to use the 

1437
01:16:50,825 --> 01:16:53,005
models, right? 
So that is what it is. 

1438
01:16:54,206 --> 01:16:57,443
Earlier we thought, "No, that is
always 
 gonna stay somewhere 

1439
01:16:57,443 --> 01:17:00,665
else, and AI will do this, 
models will do this." 

1440
01:17:00,665 --> 01:17:03,951
They're 
 doing ETA predictions 
on Google Maps and all this 

1441
01:17:03,951 --> 01:17:08,125
stuff, but now same thing 
 just
came in and came in very fast. 

1442
01:17:08,245 --> 01:17:12,277
It's a very fast change. 
Next era is gonna be about 
 

1443
01:17:12,285 --> 01:17:18,721
back to services industry. 
SaaS will get disrupted, but a 

1444
01:17:18,721 --> 01:17:21,766
small 
 SaaS will get disrupted 
much faster. 

1445
01:17:22,805 --> 01:17:25,065
Large SaaS will also get 
disrupted, right? 

1446
01:17:25,605 --> 01:17:28,225
So it is excellent for services 
industry. 

1447
01:17:28,285 --> 01:17:30,565
If you want to remain 
 
relevant, you have to learn. 

1448
01:17:31,025 --> 01:17:34,641
You have to learn. 
And I think basics, if you learn

1449
01:17:34,641 --> 01:17:38,065
basics, 
 then those basics can 
be applied anywhere. 

1450
01:17:40,417 --> 01:17:42,037
If you learn programming as a 
basic. 

1451
01:17:42,078 --> 01:17:45,157
So I tell people sometimes, you 
 need to learn only four or five

1452
01:17:45,218 --> 01:17:46,497
things in any programming 
language. 

1453
01:17:46,758 --> 01:17:52,115
How to define a variable, how to

 call a function, how to run a 

1454
01:17:52,115 --> 01:17:53,758
if condition, and how to run a 
loop. 

1455
01:17:54,157 --> 01:17:56,927
If you know this, you can learn 
any 
 programming language very 

1456
01:17:56,927 --> 01:17:59,574
quickly. 
But a lot of people start with 

1457
01:17:59,574 --> 01:18:01,258
some other 
 language, but they 
never learn the basics. 

1458
01:18:01,277 --> 01:18:03,747
They know how to do that. 
But then they would switch to 
 

1459
01:18:03,755 --> 01:18:06,457
a different language, and that 
language sounds very strange to 

1460
01:18:06,457 --> 01:18:08,795
them. 
But if you learn the basics, 
 

1461
01:18:08,803 --> 01:18:12,659
then you can switch languages. 
Same way, if you learn basics, 


1462
01:18:12,667 --> 01:18:16,321
and don't forget about reading 
and being curious and being 

1463
01:18:16,321 --> 01:18:19,518
passionate. 
So I always say there are three 

1464
01:18:19,518 --> 01:18:22,557
 things about us as individual 
which makes us very valuable. 

1465
01:18:23,577 --> 01:18:26,357
We have to be responsible. 
That's our job to deliver 

1466
01:18:26,357 --> 01:18:27,637
something. 
We have to be passionate. 

1467
01:18:28,838 --> 01:18:32,457
That's where we contribute to 
 
not only to our job, but our 

1468
01:18:32,497 --> 01:18:35,690
teams and everything else. 
And third, we have to be curious

1469
01:18:35,690 --> 01:18:38,674
because 
 that enables us to go 
to first two, right? 

1470
01:18:39,094 --> 01:18:40,774
So curious, passionate, and 
responsible. 

1471
01:18:40,894 --> 01:18:44,776
If you are, if you have to 
continuously 
 go through this 

1472
01:18:44,776 --> 01:18:48,659
circle, then you're fine. 
People actually, I think people 

1473
01:18:48,659 --> 01:18:50,694
who 
 have been curious, 
passionate, and responsible, 

1474
01:18:50,694 --> 01:18:53,234
they aren't affected. 
I think so. 

1475
01:18:54,049 --> 01:18:56,539
Wow, I love that. 
Being responsible, being 
 

1476
01:18:56,547 --> 01:19:00,290
curious, being passionate. 
So definitely key learnings 
 

1477
01:19:00,298 --> 01:19:02,970
for all of you who listen. 
We have talked a lot. 

1478
01:19:03,190 --> 01:19:05,779
Is there anything else that you 
think… No! 

1479
01:19:05,999 --> 01:19:08,054
… you have to… Okay.. 
Thank. 

1480
01:19:08,074 --> 01:19:10,114
You for hosting me. 
It was good conversation. 

1481
01:19:10,714 --> 01:19:12,294
I had no idea what we're gonna 
talk about. 

1482
01:19:12,374 --> 01:19:16,394
But yeah, looking back, my 
 
perspective has changed a lot. 

1483
01:19:16,634 --> 01:19:20,872
And I wish a lot of people get 
better 
 perspective in life and

1484
01:19:20,872 --> 01:19:23,174
see what goes on. 
It's still learning, man. 

1485
01:19:23,174 --> 01:19:24,494
I'm still learning. 
Yeah. 

1486
01:19:25,310 --> 01:19:26,910
So I don't know whether 
 you 
still remember. 

1487
01:19:28,390 --> 01:19:30,928
Every episode I will ask these 

three technical leadership 

1488
01:19:30,928 --> 01:19:32,831
wisdom. 
So I know you have 
 thrown a 

1489
01:19:32,831 --> 01:19:35,222
bunch of wisdom. 
So maybe if you can just give 

1490
01:19:35,222 --> 01:19:37,691
three 
 version, you know, short
version, what wisdom that you 

1491
01:19:37,691 --> 01:19:39,977
would want to leave 
 the 
listeners today from this 

1492
01:19:39,977 --> 01:19:41,134
episode? 
Aha! 

1493
01:19:43,634 --> 01:19:48,194
I think life perspective 
 
changes with the age, right? 

1494
01:19:49,114 --> 01:19:51,634
So when I was 25, I would 
 say 
something different. 

1495
01:19:51,654 --> 01:19:54,194
When I was 35, I said something 
 different, and now as 50, I'm 

1496
01:19:54,254 --> 01:19:55,314
saying something different, 
right? 

1497
01:19:56,054 --> 01:20:00,370
So first thing is that when 
somebody 
 offers you some help,

1498
01:20:00,370 --> 01:20:03,342
never say no, because their 
intent is always 
 to help you 

1499
01:20:03,342 --> 01:20:05,694
and make you better. 
So that is what I have learned. 

1500
01:20:06,054 --> 01:20:08,218
That is one wisdom. 
I'm talking very different, that

1501
01:20:08,218 --> 01:20:10,455
I would 
 have talked something 
very different, now I'm talking 

1502
01:20:10,455 --> 01:20:12,994
very different things. 
So first one is that. 

1503
01:20:13,694 --> 01:20:18,919
I think second, which is very 
 
important is like, in life, the 

1504
01:20:18,919 --> 01:20:22,851
more you age, the more you 
should and 
 you will value 

1505
01:20:22,851 --> 01:20:26,764
friends and family. 
Because you have very few weeks 

1506
01:20:26,764 --> 01:20:29,426
with them. 
Like for example, if I'm gonna 

1507
01:20:29,426 --> 01:20:31,934
live till 
 like 70 or 80, I 
have only 1,500 weeks. 

1508
01:20:32,084 --> 01:20:33,814
That's all. 
Not much. 

1509
01:20:34,484 --> 01:20:37,122
The weekends. 
If you talk weekends, oh my 
 

1510
01:20:37,130 --> 01:20:41,532
God, only 1,500 weekends. 
And third thing, what I would 

1511
01:20:41,532 --> 01:20:44,274
talk about, 
 be curious, be 
responsible, be passionate. 

1512
01:20:45,354 --> 01:20:50,814
I think that is very important, 
 because that will keep letting 

1513
01:20:50,934 --> 01:20:54,965
you go through the life. 
And then there are a lot of 

1514
01:20:54,965 --> 01:20:57,539
regular 
 stuff, but yeah, I 
think those three things are 

1515
01:20:57,539 --> 01:21:00,634
one, one, one. 
Like if somebody offers 
 you 

1516
01:21:00,634 --> 01:21:03,674
help, don't say no. 
Just take help or ask for help. 

1517
01:21:03,794 --> 01:21:06,574
It's okay. 
Value your friends, families 
 

1518
01:21:06,582 --> 01:21:09,774
more than anything else. 
And third, be curious, be 
 

1519
01:21:09,782 --> 01:21:12,326
passionate, and be responsible. 
Lovely. 

1520
01:21:12,445 --> 01:21:14,325
Thank you again for this 
conversation. 

1521
01:21:14,605 --> 01:21:17,117
Ajey, always excited whenever I 
 talk to you because there's so 

1522
01:21:17,117 --> 01:21:20,331
many wisdom that you always 
gave, right, 
 through your 

1523
01:21:20,331 --> 01:21:24,725
stories, experience, and also 
examples that you also live by. 

1524
01:21:24,785 --> 01:21:27,359
So again, thank you so much, and
I hope 
 people learn a lot from

1525
01:21:27,359 --> 01:21:28,045
this episode. 
Yes. 

1526
01:21:28,145 --> 01:21:29,885
Thank you so much for hosting me
again. 

1527
01:21:29,945 --> 01:21:32,005
It was amazing talking. 
It was great. 

1528
01:21:32,065 --> 01:21:32,625
Thank you so much.
