1
00:00:00,120 --> 00:00:06,040
A junior developer stares at 
their screen, fingers posed over

2
00:00:06,040 --> 00:00:10,120
the keyboard, and then does 
absolutely nothing. 

3
00:00:10,880 --> 00:00:19,120
Instead, they type a prompt into
AI build me a full stack SAS app

4
00:00:19,600 --> 00:00:25,240
that lets users write office 
coffee using sentiment analysis 

5
00:00:25,640 --> 00:00:28,680
and a crypto wallet. 
Seconds later. 

6
00:00:28,760 --> 00:00:33,000
Poof, like code appears. 
It's not perfect, it's not even 

7
00:00:33,000 --> 00:00:36,160
going to work, but it runs, sort
of. 

8
00:00:36,920 --> 00:00:41,200
Welcome to the age of vibe 
coding We're writing. 

9
00:00:41,200 --> 00:00:45,040
Software isn't about syntax or 
architecture anymore. 

10
00:00:45,240 --> 00:00:49,920
It's about the vibe, man. 
Think of it as coding by gut 

11
00:00:49,920 --> 00:00:53,280
feeling. 
Except your gut is now a chat 

12
00:00:53,280 --> 00:00:57,600
bot with ACPU and a questionable
taste in logic. 

13
00:00:57,800 --> 00:01:01,880
In this episode, we'll be diving
deep into what vibe coding 

14
00:01:02,000 --> 00:01:04,760
actually is, why it's gaining 
traction, You'll see it on 

15
00:01:04,760 --> 00:01:08,920
LinkedIn every 5 minutes, where 
it shines, and more importantly,

16
00:01:09,200 --> 00:01:13,640
where it crashes hard. 
We'll explore use cases, real 

17
00:01:13,640 --> 00:01:18,600
life case studies, and why this 
trend might just be the fastest 

18
00:01:18,600 --> 00:01:22,480
route to technical debt. 
You can't refinance. 

19
00:01:23,440 --> 00:01:26,120
If you're a business analyst, a 
product owner, a developer, a 

20
00:01:26,120 --> 00:01:30,520
business leader, or someone just
trying to stay one buzzword 

21
00:01:30,520 --> 00:01:33,760
ahead of your next meeting, this
one's for you. 

22
00:01:35,040 --> 00:01:39,080
The Better Business Analysis 
Institute presence, the Better 

23
00:01:39,080 --> 00:01:42,280
Business analysis podcast with 
Kingsman Walsh. 

24
00:01:46,920 --> 00:01:50,120
Let's get into it. 
Here is 10 things that you need 

25
00:01:50,120 --> 00:01:54,600
to know about vibe coding before
it's vibe puts you into 

26
00:01:54,600 --> 00:01:56,520
technical debt. 
That's right. 

27
00:01:57,200 --> 00:02:01,080
So number one, we really need to
talk about what is this thing 

28
00:02:01,080 --> 00:02:02,480
called vibe coding? 
It sounds cool. 

29
00:02:03,400 --> 00:02:07,400
Vibe coding is an AI driven 
programming approach where 

30
00:02:07,400 --> 00:02:12,480
developers provide high level 
prompts to large language model 

31
00:02:12,480 --> 00:02:15,440
models. 
So AI, think of ChatGPT as an 

32
00:02:15,440 --> 00:02:19,200
example, which then generates 
the corresponding code. 

33
00:02:19,600 --> 00:02:22,360
And with things like MCP that 
we've talked about recently, 

34
00:02:22,560 --> 00:02:26,280
this could actually start 
building multi layer 

35
00:02:26,520 --> 00:02:30,280
architecture. 
This was coined in February 

36
00:02:30,280 --> 00:02:33,320
2025. 
So it's really new and it shifts

37
00:02:33,320 --> 00:02:37,360
the developer's role from the 
manual Kona to guiding and 

38
00:02:37,360 --> 00:02:41,000
refining AI code. 
And while we're on this number 

39
00:02:41,000 --> 00:02:46,200
one point, I do this, I do vibe 
coding all the time because I'm 

40
00:02:46,200 --> 00:02:49,200
not a programmer anymore. 
I don't know the latest 

41
00:02:49,200 --> 00:02:53,320
languages and I feed in prompts 
to AI, more complicated prompts 

42
00:02:53,320 --> 00:02:56,640
and areas of code and it allows 
me to prototype quickly. 

43
00:02:56,640 --> 00:02:58,000
So we'll get on to that in a 
minute. 

44
00:02:58,000 --> 00:03:01,520
But you might be one of these 
developers out there today. 

45
00:03:01,840 --> 00:03:05,000
But this really vibe coding is 
when developers are doing this. 

46
00:03:05,000 --> 00:03:08,680
This is instead of them having 
to writing lines of code, they 

47
00:03:08,680 --> 00:03:11,200
can be, some of them can be 
notoriously lazy anyway. 

48
00:03:11,440 --> 00:03:15,400
They're actually getting AI to 
do the work for them #2 there's 

49
00:03:15,400 --> 00:03:17,800
an appeal about wide coding. 
It sounds cool. 

50
00:03:18,360 --> 00:03:21,720
It accelerates development by 
automating routine coding tasks 

51
00:03:21,720 --> 00:03:24,120
that you might have to do again,
and it's boring. 

52
00:03:24,600 --> 00:03:27,320
It's accessible. 
Even individuals with limited 

53
00:03:27,320 --> 00:03:31,160
coding experience like myself, 
we can develop functional 

54
00:03:31,160 --> 00:03:33,920
software. 
And you see a lot of promotion 

55
00:03:33,920 --> 00:03:36,640
about this online and 
innovation. 

56
00:03:36,720 --> 00:03:40,400
It fosters rapid prototyping and
experimentation. 

57
00:03:40,440 --> 00:03:45,240
And that's a really good point. 
So #3 what are the practical 

58
00:03:45,240 --> 00:03:48,680
applications here? 
Well, in start-ups, if you have 

59
00:03:48,680 --> 00:03:53,440
small teams, you can achieve 
outputs that are on scale with 

60
00:03:53,440 --> 00:03:57,080
larger teams by leveraging AI 
tools, especially for some of 

61
00:03:57,080 --> 00:04:00,120
the manual work, but not all of 
the workflow steps. 

62
00:04:00,680 --> 00:04:04,480
And enterprise solutions 
companies like Amazon utilise AI

63
00:04:04,480 --> 00:04:06,520
assistance to enhance 
development. 

64
00:04:07,080 --> 00:04:11,120
They have like a a code plug in,
Visual Studio has it, the GitHub

65
00:04:11,400 --> 00:04:14,920
plug in for Visual Studio I've 
used and it helps you for 

66
00:04:14,920 --> 00:04:18,480
looking up classes and what not 
and educational tools. 

67
00:04:18,480 --> 00:04:23,520
So facilities can allow students
to focus on problem solving 

68
00:04:23,920 --> 00:04:26,480
rather than syntax. 
So they can kind of think about 

69
00:04:26,480 --> 00:04:30,440
how they might program a Python 
script and get it about right. 

70
00:04:30,440 --> 00:04:33,000
And they're like, well, don't 
worry, you didn't get the the 

71
00:04:33,000 --> 00:04:34,440
syntax right. 
We'll move on. 

72
00:04:34,720 --> 00:04:38,960
And syntax, to be honest, is 
just the like grammar of 

73
00:04:38,960 --> 00:04:42,240
programming. 
So I use AI to fix my grammar 

74
00:04:42,240 --> 00:04:45,880
all the time. 
So programmers can use AI to 

75
00:04:45,880 --> 00:04:47,840
check their syntax and their 
syntax checkers. 

76
00:04:48,120 --> 00:04:53,000
So why wouldn't you use AI for 
that part of it #4 OK. 

77
00:04:53,000 --> 00:04:59,720
And there's a company called RE 
Pets AI generation game. 

78
00:04:59,840 --> 00:05:04,720
So there's a, a developer uses 
Repat's AI to generate a racing 

79
00:05:04,720 --> 00:05:11,800
game featuring actually Elon 
Musk and Sam Altman from Open 

80
00:05:11,800 --> 00:05:14,640
AI. 
And despite initial bugs, the 

81
00:05:14,640 --> 00:05:17,640
project actually demonstrated 
the potential of AI in a 

82
00:05:17,640 --> 00:05:20,920
simplified app development. 
It, it worked and people were 

83
00:05:20,920 --> 00:05:23,400
able to play it. 
And this gave people the idea 

84
00:05:23,680 --> 00:05:28,160
of, well, why don't we utilise 
that model for our business? 

85
00:05:28,440 --> 00:05:31,640
And that leads us to #5 which 
are the potential risks here? 

86
00:05:32,080 --> 00:05:34,120
Number one, it's, it's technical
get. 

87
00:05:34,520 --> 00:05:38,320
So the architects and people 
that are really hot on good 

88
00:05:39,040 --> 00:05:43,440
effective code. 
I I I do not like this concept 

89
00:05:43,440 --> 00:05:46,120
when applied to production 
environments. 

90
00:05:47,080 --> 00:05:51,400
Rapid development without proper
structure leads to maintenance 

91
00:05:51,400 --> 00:05:54,320
challenges and bad architectural
decisions. 

92
00:05:54,880 --> 00:05:57,400
All the code is usually a couple
of files. 

93
00:05:57,640 --> 00:06:01,200
It's not spread out how do you 
maintain it and this makes it 

94
00:06:01,200 --> 00:06:04,120
worse. 
The other issue is around 

95
00:06:04,120 --> 00:06:08,800
security, and this opens up 
security holes. 

96
00:06:09,080 --> 00:06:11,560
The reason we have programmers 
and the reason we have people 

97
00:06:11,560 --> 00:06:15,160
that know about security, 
security architects, is that AI 

98
00:06:15,160 --> 00:06:19,080
driven code may overlook some of
these practices if you haven't 

99
00:06:19,080 --> 00:06:22,080
fully explained what they are. 
And then there's an over 

100
00:06:22,080 --> 00:06:24,480
reliance on AI. 
So developers might become 

101
00:06:24,480 --> 00:06:27,200
detached from the understanding 
of the code and then when 

102
00:06:27,200 --> 00:06:29,080
something breaks, they don't 
know when to fix it. 

103
00:06:30,080 --> 00:06:33,320
And so let's talk about 
mitigating that risk. 

104
00:06:33,320 --> 00:06:36,320
So number six is mitigating the 
technical debt problem, and this

105
00:06:36,320 --> 00:06:41,040
is an argument against the 
technical weights Det starts and

106
00:06:41,040 --> 00:06:44,240
that is that you can have code 
reviews to ensure that the code 

107
00:06:44,240 --> 00:06:48,520
quality is there. 
You can ask AI to generate clean

108
00:06:48,520 --> 00:06:51,920
documentation and to provide 
it's thinking about what it's 

109
00:06:51,920 --> 00:06:56,400
done and include that in the 
code itself as comments and 

110
00:06:56,400 --> 00:06:58,160
testing. 
You can make sure that your 

111
00:06:58,160 --> 00:07:01,560
testing is way more robust now 
and it incorporates rigorous 

112
00:07:01,560 --> 00:07:03,840
testing to identify and fix the 
issues early. 

113
00:07:04,080 --> 00:07:05,800
So you can actually use this 
technique. 

114
00:07:07,920 --> 00:07:09,840
There is a comparative study 
here. 

115
00:07:09,840 --> 00:07:13,680
So #7 here is around Airbnb. 
They use this concept of 

116
00:07:13,920 --> 00:07:17,520
modernization, right? 
And the challenge was that 

117
00:07:17,520 --> 00:07:21,920
Airbnb had this monolithic 
architecture and it led to 

118
00:07:21,920 --> 00:07:24,480
increased dependencies and slow 
releases. 

119
00:07:24,720 --> 00:07:27,720
And we talked about that in the 
last episode around architecture

120
00:07:27,720 --> 00:07:29,640
and the putting the bits and 
pieces together. 

121
00:07:30,280 --> 00:07:33,520
And the solution was to 
transition to micro services, 

122
00:07:33,520 --> 00:07:39,440
which is the other architecture 
we talked about using various, I

123
00:07:39,440 --> 00:07:43,440
won't go into technical ways 
they did that, but they did that

124
00:07:43,440 --> 00:07:45,760
to improve scalability and 
reduce technical debt. 

125
00:07:46,320 --> 00:07:49,320
And that is a kind of a better 
way of doing things is breaking 

126
00:07:49,320 --> 00:07:51,400
things up. 
And then maybe if you had a 

127
00:07:51,400 --> 00:07:56,720
small micro service, maybe you 
could apply AI to that micro 

128
00:07:56,720 --> 00:08:01,320
service to make that better, but
it's kind of not applying AI and

129
00:08:01,320 --> 00:08:03,440
vibe coding across the whole 
structure. 

130
00:08:03,440 --> 00:08:05,080
So get your architecture right 
first. 

131
00:08:06,040 --> 00:08:08,600
Some of the organizational 
considerations we need to think 

132
00:08:08,600 --> 00:08:11,640
about, which is number 8 is 
making sure that you are 

133
00:08:11,640 --> 00:08:14,560
training your teams effectively.
You need to equip your teams 

134
00:08:14,560 --> 00:08:18,560
with the skills to effectively 
effectively the keyword here use

135
00:08:18,560 --> 00:08:20,640
AI tools and not replace 
themselves. 

136
00:08:21,120 --> 00:08:23,400
There's governance. 
So what are the policies how we 

137
00:08:23,400 --> 00:08:28,120
manage AI generated code 
responsibly, but allow AI to be 

138
00:08:28,120 --> 00:08:30,880
the true copilot as opposed to 
the writer. 

139
00:08:31,800 --> 00:08:34,559
And if you I can tell you that 
if you're using AOI to write 

140
00:08:34,559 --> 00:08:37,559
code for you as a developer, 
then you are not going to be 

141
00:08:37,559 --> 00:08:40,280
needed soon enough. 
And integration. 

142
00:08:40,280 --> 00:08:44,840
So ensure these AI tools align 
with your existing development 

143
00:08:44,840 --> 00:08:48,240
workflows, checking in code, 
code reviews and what not. 

144
00:08:49,160 --> 00:08:54,000
And any of your best practices 
need to be built into the vibe 

145
00:08:54,000 --> 00:08:57,560
code ring experience, unless 
you're going to be using that 

146
00:08:57,840 --> 00:09:02,200
for prototypes, which I'm going 
to finish off with #9 is 

147
00:09:02,200 --> 00:09:05,920
strategic implementations. 
So if you are want to try vibe 

148
00:09:05,920 --> 00:09:09,040
coding and I I suggest you do, 
I'm not saying it's all bad. 

149
00:09:09,280 --> 00:09:12,880
Start with a small start scale 
project or a micro service to 

150
00:09:12,880 --> 00:09:16,680
assess the effectiveness of that
in your organization and whether

151
00:09:16,680 --> 00:09:20,480
or not by allowing this option, 
you know all the rules go out 

152
00:09:20,480 --> 00:09:24,320
the window. 
Have feedback loops, look at 

153
00:09:24,320 --> 00:09:29,320
feedback, provide AI prompts and
outputs and look at reasoning 

154
00:09:29,320 --> 00:09:33,160
behind the AI prompts which are 
available now and look and see 

155
00:09:33,160 --> 00:09:37,600
if that's correct reasoning. 
I would say that in my 

156
00:09:37,600 --> 00:09:41,920
experience, the tools out there 
today are not that great at 

157
00:09:41,920 --> 00:09:44,960
programming, to be honest. 
I don't think they're full. 

158
00:09:45,160 --> 00:09:49,640
They're good at creating good 
structured architecture and look

159
00:09:49,640 --> 00:09:53,000
at scalability. 
Can this AI tool that you've 

160
00:09:53,000 --> 00:09:56,400
just create actually be scaled? 
And if it can't be scaled, then 

161
00:09:56,400 --> 00:09:59,320
obviously it's just a peripheral
concept and can be thrown away 

162
00:09:59,640 --> 00:10:01,800
at the end of your vibe coding 
session. 

163
00:10:03,000 --> 00:10:06,760
Before I drop into the final 
thoughts, the best use case for 

164
00:10:06,760 --> 00:10:09,880
vibe coding in my mind is you're
an entrepreneur, you're working 

165
00:10:09,880 --> 00:10:12,160
with a developer or you're 
working with people who know a 

166
00:10:12,160 --> 00:10:15,000
bit about architecture and you 
want to rapidly produce a 

167
00:10:15,000 --> 00:10:19,120
prototype to show the market to 
get funding to do it properly. 

168
00:10:19,360 --> 00:10:24,080
OK, now that doesn't mean you 
don't use AI in the generation 

169
00:10:24,080 --> 00:10:28,200
of production materials and 
documentation, but you do it in 

170
00:10:28,200 --> 00:10:31,000
a non vibe way. 
You do it in a very structured, 

171
00:10:31,640 --> 00:10:35,200
methodical way and you use the 
vibe, the vibe to get the 

172
00:10:35,200 --> 00:10:38,600
funding and to get people on 
board by, you know, showing them

173
00:10:38,600 --> 00:10:42,120
something cool fast. 
My final thoughts are vibe 

174
00:10:42,120 --> 00:10:46,600
coding represents a significant 
shift in software development, 

175
00:10:46,840 --> 00:10:48,800
offering both opportunities and 
challenging. 

176
00:10:49,000 --> 00:10:52,480
And I think it will evolve into 
just code coding as opposed to 

177
00:10:52,480 --> 00:10:55,880
vibe coding, apart from the use 
case I just mentioned which is 

178
00:10:55,880 --> 00:11:00,240
prototyping. 
And by embracing AI thoughtfully

179
00:11:00,280 --> 00:11:04,520
and implementing best practices,
then organizations can harness 

180
00:11:04,520 --> 00:11:06,640
it's benefits while mitigating 
the risk. 

181
00:11:08,000 --> 00:11:11,640
I will see you next week and I 
hope you know what vibe coding 

182
00:11:11,640 --> 00:11:14,360
is. 
See you later and Happy Easter.