EP238 - AI & PPC

1
00:00:00,100 --> 00:00:03,000
Hello, and welcome to this
week's episode of the PPC chat

2
00:00:03,000 --> 00:00:05,200
Roundup podcast in today's
episode.

3
00:00:05,200 --> 00:00:08,400
We have Julie taken us on a very
timely topic.

4
00:00:08,400 --> 00:00:10,700
And when I say this many times,
because that's what we're always

5
00:00:10,700 --> 00:00:12,800
doing, because the industry is
always evolving.

6
00:00:13,000 --> 00:00:15,300
There's always something new to
talk about and we're going to be

7
00:00:15,300 --> 00:00:20,000
talking about Ai and PPC.
This time, I'm sure that you

8
00:00:20,000 --> 00:00:23,700
probably have seen on your
LinkedIn or Twitter timeline.

9
00:00:23,700 --> 00:00:27,200
A lot of people have been
talking about chat GPT, so

10
00:00:27,200 --> 00:00:31,000
that's an AI tool that if you
ask it to To deliver you some

11
00:00:31,000 --> 00:00:34,900
content in some way of like an
ad copy or something to write on

12
00:00:34,900 --> 00:00:37,500
your social media or some like
bit of text to, right?

13
00:00:37,500 --> 00:00:41,400
Or you're asking a question for
its answer you, it gives you a

14
00:00:41,400 --> 00:00:44,600
bit of its knowledge of the
thing that you've asked for it

15
00:00:44,600 --> 00:00:48,100
based on, you know, it's in
trained trained.

16
00:00:48,100 --> 00:00:51,800
I think it's been developed to
have analyzed.

17
00:00:51,800 --> 00:00:55,300
A lot of the content that is on
the line all over the world, to

18
00:00:55,300 --> 00:00:57,800
then give you an appropriate
answer to your question.

19
00:00:58,200 --> 00:01:00,500
So, yeah, we're talking about
how that's going to That things

20
00:01:00,500 --> 00:01:04,500
like ad copy keyword creation
and just Automation in digital

21
00:01:04,500 --> 00:01:07,400
advertising in paid search and
I'll just what summation Ai and

22
00:01:07,700 --> 00:01:11,800
how a I could take over the
menial day-to-day tasks that

23
00:01:11,800 --> 00:01:13,300
we're doing.
So yeah.

24
00:01:13,400 --> 00:01:24,600
Hope you enjoy.
Hello and a warm, welcome to

25
00:01:24,600 --> 00:01:28,000
this episode of the PPC chat
Roundup a podcast where I round

26
00:01:28,000 --> 00:01:31,200
up the Twitter discussion that
happens on Tuesdays at 5 p.m.

27
00:01:31,400 --> 00:01:33,600
GMT.
I'm your host, Andrew, I'm not

28
00:01:33,607 --> 00:01:36,400
recording from my studio here in
London UK, and I use this

29
00:01:36,400 --> 00:01:40,000
platform to share not just
expert, but also my ideas and

30
00:01:40,000 --> 00:01:43,200
consider best practices about
paid media and the direction and

31
00:01:43,208 --> 00:01:45,500
the digital industry is going
in.

32
00:01:45,900 --> 00:01:49,800
So if you want to get the
up-to-date tips and tricks in

33
00:01:49,800 --> 00:01:52,800
the industry and get the latest
on the podcast that has been

34
00:01:52,800 --> 00:01:55,900
called, Your go to even if you
don't make it to the PPC chat

35
00:01:55,900 --> 00:01:58,400
live discussions.
And even if you do then like,

36
00:01:58,400 --> 00:02:01,800
follow share and retweets on
Twitter with my handle the

37
00:02:01,800 --> 00:02:04,800
marketing, a new ordering us on
our LinkedIn Group, which is the

38
00:02:04,800 --> 00:02:07,900
PPC chat Roundup podcast.
Or if you're on Instagram, why

39
00:02:07,900 --> 00:02:11,900
don't you check out the latest
Snippets and quotes from

40
00:02:11,900 --> 00:02:15,300
previous episodes?
So as I said, we have Julie

41
00:02:15,300 --> 00:02:20,000
taken us on the topic of AI and
PPC and I'm wondering what have

42
00:02:20,000 --> 00:02:22,700
you guys found it?
Have you guys experienced it,

43
00:02:22,700 --> 00:02:25,000
have you?
The testing that out, I tried to

44
00:02:25,008 --> 00:02:28,100
test it this week but every time
I try to test it, I keep getting

45
00:02:28,100 --> 00:02:30,400
the alert that there's an
overload.

46
00:02:30,500 --> 00:02:32,800
And yeah, it's not available at
the moment.

47
00:02:32,800 --> 00:02:36,900
So I haven't had the chance to
test it yet but it's not a want

48
00:02:36,900 --> 00:02:39,200
of trying.
It's not of, you know, sticking

49
00:02:39,200 --> 00:02:41,500
my head in the sand and not want
to try new things.

50
00:02:41,500 --> 00:02:44,600
I do love to try new things.
But yeah, right now, loads of

51
00:02:44,600 --> 00:02:47,500
people are trying it out and
there's no capacity for me.

52
00:02:47,900 --> 00:02:50,900
Anyway, before we get into the
questions, we've got Jilly's PSA

53
00:02:50,900 --> 00:02:52,800
going.
Yeah, this chat is intended for

54
00:02:52,800 --> 00:02:54,200
people.
I see Professionals of all

55
00:02:54,200 --> 00:02:56,000
levels.
So, if you have something you

56
00:02:56,000 --> 00:03:00,300
want to share, please, do we are
a friendly and very supportive

57
00:03:00,300 --> 00:03:02,100
community.
So please don't be afraid to

58
00:03:02,100 --> 00:03:05,100
share but lurkers and just the
listeners of the podcast are

59
00:03:05,100 --> 00:03:08,200
always, welcome to.
So let's get straight into the

60
00:03:08,200 --> 00:03:09,700
questions.
We have Julie going.

61
00:03:09,700 --> 00:03:12,900
Yeah let's get into it question.
1 are you currently utilizing

62
00:03:13,200 --> 00:03:16,200
any AI artificial intelligence
in your PPC efforts?

63
00:03:16,200 --> 00:03:19,800
If so what are you using?
If not, why not?

64
00:03:20,200 --> 00:03:21,800
And we got serious.
Teaming and responding saying

65
00:03:21,800 --> 00:03:25,300
yes, and sometimes it is Plane
in communication to the client

66
00:03:25,300 --> 00:03:29,300
AKA sentence rewrites and other
times ad copy.

67
00:03:29,300 --> 00:03:30,900
Wow.
And we have, you're gonna

68
00:03:30,900 --> 00:03:33,900
responding saying I'd say we
utilize machine learning, but

69
00:03:33,900 --> 00:03:36,700
not AI.
Artificial intelligence, lots of

70
00:03:36,700 --> 00:03:40,300
Google's automated, approaches
our machine learning, pmax,

71
00:03:40,300 --> 00:03:45,100
bidding excetera, but we don't
use anything outside of that and

72
00:03:45,100 --> 00:03:46,900
Yep, they're now using air as
well.

73
00:03:47,300 --> 00:03:50,000
And actor.
Also asks, Gill saying, do you

74
00:03:50,000 --> 00:03:52,800
also use machine learning apart
from what is already being

75
00:03:52,800 --> 00:03:54,900
provided?
Guided by the platform girl

76
00:03:54,900 --> 00:03:57,300
responds.
No, we've not found that any

77
00:03:57,300 --> 00:04:00,800
third-party tools are really
high quality enough to use on a

78
00:04:00,800 --> 00:04:03,900
day-to-day basis.
Interesting Kristine, Lazar, and

79
00:04:03,900 --> 00:04:06,700
held response to question one
saying we use machine learning,

80
00:04:06,700 --> 00:04:10,900
smart, bidding P Max but not a
lot of a I am still of the

81
00:04:10,900 --> 00:04:13,900
mindset that a human can write
better than a computer.

82
00:04:13,900 --> 00:04:19,600
I wouldn't be against using chat
GPT to get copy or keyword

83
00:04:19,600 --> 00:04:21,399
ideas, though.
Debbie noted response to

84
00:04:21,399 --> 00:04:24,700
question one saying, gee.
Campaigns with some of the

85
00:04:24,700 --> 00:04:27,800
budgets lack of transparency.
With these accounts drives me

86
00:04:27,800 --> 00:04:30,300
nuts.
Runs: response saying, like

87
00:04:30,300 --> 00:04:33,500
most, we have been using machine
learning within Google ads for a

88
00:04:33,500 --> 00:04:35,400
while.
I've had a play around with

89
00:04:35,400 --> 00:04:40,100
chat, GPT for creating ads, Etc,
but mostly experiments and see

90
00:04:40,100 --> 00:04:43,800
how good it really is.
Rory response, question once and

91
00:04:43,800 --> 00:04:47,400
looking forward to testing this
tool that uses Chad GPT to

92
00:04:47,400 --> 00:04:49,600
screen Search terms.
Yes, that's me.

93
00:04:49,600 --> 00:04:51,700
I'm Michael Schroeder that
shared that about.

94
00:04:51,700 --> 00:04:55,500
It's like an Excel file.
While the checks the quality of

95
00:04:55,500 --> 00:04:58,700
the keywords that can t BT has
actually given you.

96
00:04:59,000 --> 00:05:00,700
So, yeah.
And it continues at this stage.

97
00:05:00,700 --> 00:05:04,100
I wanted to test things and see
what sticks and this holds

98
00:05:04,100 --> 00:05:07,300
response saying easy win on the
initial, keyword research via

99
00:05:07,600 --> 00:05:11,700
chat, GPT online advertising.
Academy goes, yes, but for

100
00:05:11,700 --> 00:05:15,500
brainstorming and getting ad
copy ideas from chat GPT, but

101
00:05:15,500 --> 00:05:17,300
the ad copy, still needs
revision.

102
00:05:17,300 --> 00:05:20,700
It's not a good cut and paste
tool for Google ads or blogs

103
00:05:21,200 --> 00:05:25,100
Rabe acid response saying I have
Doubled with some AI to draft

104
00:05:25,100 --> 00:05:28,200
copy ideas about a year ago.
I didn't find it useful.

105
00:05:28,200 --> 00:05:32,600
So abandoned the approach
haven't tried to chat GPT yets

106
00:05:32,700 --> 00:05:34,300
and we have Julia voice
responded.

107
00:05:34,300 --> 00:05:37,500
Question once a not sure what
exactly counts as a.

108
00:05:37,500 --> 00:05:40,600
I but I have a lot of automation
enabled in platforms.

109
00:05:40,600 --> 00:05:44,300
And several platforms are using
lists to look for look-alike

110
00:05:44,300 --> 00:05:48,400
audience has ad, still written
and approved by people keywords,

111
00:05:48,400 --> 00:05:51,600
still researched by people in
the tools.

112
00:05:51,800 --> 00:05:54,400
And I respond to that saying,
Yeah, I froze on this question

113
00:05:54,400 --> 00:05:57,600
as well, because from seen a few
answers, I thought, wait, what

114
00:05:57,600 --> 00:06:00,600
is a ippc because people were
given answers around.

115
00:06:00,700 --> 00:06:04,200
Yes, they still use Automation
in with your Google ads while

116
00:06:04,200 --> 00:06:06,000
I'm like, that's not exactly
what AI is.

117
00:06:06,000 --> 00:06:09,900
But then I thought probably chat
GPT and the like I assume is

118
00:06:09,900 --> 00:06:12,200
what they're talking about in
terms of AI.

119
00:06:12,200 --> 00:06:14,100
So they don't mean like
automation like bidding

120
00:06:14,100 --> 00:06:17,500
strategies or our essays and
yeah, Harrison response to me

121
00:06:17,500 --> 00:06:19,700
saying, yeah, the
differentiation between same

122
00:06:19,700 --> 00:06:24,000
machine learning and artificial
intelligence is, Maybe machine

123
00:06:24,000 --> 00:06:27,100
learning is part of a i-i'm, not
sure neither am.

124
00:06:27,100 --> 00:06:31,700
I actor goes to Julia by saying.
Have you ever tried Facebook API

125
00:06:31,700 --> 00:06:35,100
for audience, segments,
research, Julia, Bosco's no, not

126
00:06:35,100 --> 00:06:37,800
the API, just audiences within
the platform.

127
00:06:38,000 --> 00:06:41,100
Then first party lists actor.
Goes try it.

128
00:06:41,100 --> 00:06:43,200
You will definitely like it
interesting.

129
00:06:43,700 --> 00:06:46,200
And we have Julie responding
saying I'm going to agree with

130
00:06:46,200 --> 00:06:50,300
gildner here in that we use some
machine learning functions in

131
00:06:50,300 --> 00:06:53,900
Google.
I have not used any a I like GPT

132
00:06:53,900 --> 00:06:58,300
for any aspect of PPC to date.
I am interested in testing what

133
00:06:58,300 --> 00:07:01,300
it can do.
But behind the scenes not auto

134
00:07:01,300 --> 00:07:04,900
used or applied and we have
Sophie Logan responding to

135
00:07:04,900 --> 00:07:06,900
question.
Once the surfing Logan, who is

136
00:07:06,900 --> 00:07:10,300
going to be speaking at the next
PPC live UK on the second

137
00:07:10,300 --> 00:07:12,300
offense.
So, if you're in the UK or in,

138
00:07:12,300 --> 00:07:16,400
in London, specifically, please
go to PC live UK.com for your

139
00:07:16,400 --> 00:07:18,800
tickets.
Anyway, continuing her answer

140
00:07:18,800 --> 00:07:22,200
without answer to question 1.
She goes utilizing some in

141
00:07:22,200 --> 00:07:25,000
platform.
And features, but staying away

142
00:07:25,000 --> 00:07:29,400
from anything to do with out
copy and keywords at the moment.

143
00:07:29,400 --> 00:07:31,000
Julia Vice gives another answer
to question.

144
00:07:31,000 --> 00:07:34,900
One saying I did run a test with
Albert AI several years ago,

145
00:07:34,900 --> 00:07:36,700
remember them?
No I don't.

146
00:07:37,100 --> 00:07:40,600
It was a quasi success.
Not perfect but very

147
00:07:40,600 --> 00:07:43,300
interesting.
Yeah but a I don't know and

148
00:07:43,300 --> 00:07:46,600
really responds to that saying I
sometimes wonder what IBM's

149
00:07:46,600 --> 00:07:50,900
Watson thinks about all the
current AI hoopla.

150
00:07:52,600 --> 00:07:54,700
Yeah.
I don't know about, I am IBM's

151
00:07:54,700 --> 00:07:56,800
Watson as well.
I hope it has something to do

152
00:07:56,800 --> 00:08:00,800
with Sherlock Holmes and Watson
doing Brown response to

153
00:08:00,808 --> 00:08:02,400
question.
One thing we have looked at some

154
00:08:02,400 --> 00:08:06,600
AI tools and test stuff, nothing
comes to mind where we use AI in

155
00:08:06,600 --> 00:08:10,300
our day-to-day most use
automation, which is more

156
00:08:10,300 --> 00:08:14,100
machine learning and that jazz
Mark suitable response saying

157
00:08:14,100 --> 00:08:17,500
yes for ad copies.
It's been really helpful when

158
00:08:17,500 --> 00:08:20,900
you're just not feeling creative
and then actor herself response

159
00:08:20,900 --> 00:08:23,600
to question one saying, yes, I
would say I use a Out of it

160
00:08:23,600 --> 00:08:27,900
copies topic, search brand
research, USPS and many more.

161
00:08:27,900 --> 00:08:33,200
I literally put a prompt in chat
GPT to write USPS for clothing

162
00:08:33,200 --> 00:08:36,299
brand and it did.
And I tweaked a bit showed to

163
00:08:36,299 --> 00:08:38,700
clients.
And he said, we didn't notice

164
00:08:38,700 --> 00:08:40,700
these points.
Oh, interesting.

165
00:08:41,200 --> 00:08:43,799
And then we have a system clock
responding to question once and

166
00:08:43,799 --> 00:08:47,400
we test automated bidding.
If that counts as a, i as a

167
00:08:47,400 --> 00:08:51,200
thing, I don't think it does in
various accounts and keep it

168
00:08:51,200 --> 00:08:52,900
active.
If it performs well, other than

169
00:08:53,100 --> 00:08:56,600
Than that, all of our processes
for keyword research ad copy

170
00:08:56,600 --> 00:08:59,500
excetera is still done by
people.

171
00:09:00,100 --> 00:09:03,300
And then we have Sam responded
to question once, and I think

172
00:09:03,300 --> 00:09:06,400
almost everyone is using some
form of machine learning in

173
00:09:06,400 --> 00:09:09,900
their PPC campaigns.
If you use small to automated

174
00:09:09,900 --> 00:09:13,100
anything you use machine
learning, in my opinion, one of

175
00:09:13,100 --> 00:09:16,500
the biggest issues in the
industry is the conflation

176
00:09:16,700 --> 00:09:21,100
between ML and AI to provide
some context on how I think

177
00:09:21,100 --> 00:09:25,100
about this AI is machine.
The mimic or exceed human

178
00:09:25,100 --> 00:09:28,700
capacities without human
intervention or assistance

179
00:09:29,000 --> 00:09:31,900
machine learning algorithms that
I'm able programs are platforms

180
00:09:31,900 --> 00:09:36,600
or systems to identify patterns,
make decisions plus improve via

181
00:09:36,600 --> 00:09:38,900
iteration Julie response to
something.

182
00:09:38,900 --> 00:09:42,400
I agreed, I think we have been
immersed in machine learning in

183
00:09:42,400 --> 00:09:46,700
platforms for quite a few years.
This topic seems timely with all

184
00:09:46,700 --> 00:09:50,000
the talk about Chatterjee PT,
which is a eye.

185
00:09:50,300 --> 00:09:53,400
And yeah, some response saying,
uh, at DBT is just a Addictive

186
00:09:53,400 --> 00:09:57,300
text algorithm, it doesn't
understand language or concepts.

187
00:09:57,500 --> 00:10:01,300
It's simply predicting the next
word in a series based on an

188
00:10:01,300 --> 00:10:05,200
initial data set plus using
methods that have been in use

189
00:10:05,200 --> 00:10:09,400
for years in for example, Google
search just in loads of people

190
00:10:09,400 --> 00:10:12,900
have answers to the fact and,
and this goes, yeah.

191
00:10:12,900 --> 00:10:15,900
I also 30 year-old industry
professional from the cement

192
00:10:15,900 --> 00:10:17,600
industry to tell the biggest
challenge.

193
00:10:17,600 --> 00:10:22,000
They had chat GPT provided a
10/10 answer in 15 seconds.

194
00:10:22,300 --> 00:10:25,100
The effect is More than just
adding words because we have

195
00:10:25,100 --> 00:10:27,800
given the intimate, all our
knowledge and some girls.

196
00:10:27,800 --> 00:10:30,400
Yeah, I have seen the same thing
on the polar opposite, asking

197
00:10:30,400 --> 00:10:34,700
GPT basic information, about
specific, legal case types, and

198
00:10:34,700 --> 00:10:37,600
the financial products, and
receiving information that is

199
00:10:37,600 --> 00:10:42,400
entirely outdated Incorrect.
And in one instance illegal wow

200
00:10:42,500 --> 00:10:44,900
illegal.
So yeah, don't trust everything

201
00:10:44,900 --> 00:10:47,300
chat GPT tells you some continue
saying?

202
00:10:47,300 --> 00:10:50,200
I think there's even been a
significant underestimation of

203
00:10:50,200 --> 00:10:52,700
the importance of trust
reliability because consistency

204
00:10:52,700 --> 00:10:56,600
and Such there is a reason China
is already moving aggressively

205
00:10:56,700 --> 00:11:00,600
to regulate algorithmically
generated content.

206
00:11:00,600 --> 00:11:03,500
Yeah.
Regulation of algorithms and go

207
00:11:03,500 --> 00:11:06,700
as well goes agreed to Sam's
Point saying, yeah, I can't

208
00:11:06,700 --> 00:11:09,500
understand.
Abstractions, console, riddles,

209
00:11:09,900 --> 00:11:12,400
and biases towards vague
responses.

210
00:11:12,600 --> 00:11:16,700
If there isn't exact copy or
paste data available, it's an

211
00:11:16,700 --> 00:11:20,300
impressive linguistic tool, but
I'd hardly called it a.

212
00:11:20,300 --> 00:11:23,500
I cooked also joins the
conversation Let's cook

213
00:11:23,500 --> 00:11:25,800
Williamson.
I was talking to a friend about

214
00:11:25,800 --> 00:11:29,800
Chad TPT and trying to describe
how we had been working with

215
00:11:29,800 --> 00:11:32,300
similar.
Albeit, simpler Tech within

216
00:11:32,300 --> 00:11:35,200
Google ads for years, EG their
headline description.

217
00:11:35,200 --> 00:11:38,500
Suggestions her singles.
This is one reason why I don't

218
00:11:38,500 --> 00:11:41,500
understand the people saying,
Google is terrified of child

219
00:11:41,500 --> 00:11:44,600
GPT.
I mean, Google has to have Tech

220
00:11:44,600 --> 00:11:48,100
on this level at a minimum.
Absolutely, and some goes,

221
00:11:48,100 --> 00:11:51,900
Google's publicly disclosed
alternatives to chat GPT

222
00:11:51,900 --> 00:11:54,700
exponentially more, Oh
impressive, for instance, they

223
00:11:54,700 --> 00:11:57,400
contain significant
reinforcements learning

224
00:11:57,600 --> 00:12:02,100
protocols, which result in more
accurate and properly cited.

225
00:12:02,300 --> 00:12:04,500
Information Harrison was
interesting.

226
00:12:04,500 --> 00:12:06,500
I'll have to research that a bit
personally.

227
00:12:06,500 --> 00:12:09,900
I haven't been that impressed
with Chad GPT it's interesting.

228
00:12:09,900 --> 00:12:12,300
And I can see where it's going,
but all it does.

229
00:12:12,300 --> 00:12:15,800
Now is produce information.
You can find through search.

230
00:12:15,800 --> 00:12:18,900
It's just faster and Kirk
response to that saying, I had

231
00:12:18,900 --> 00:12:21,900
to develop a friend, who showed
me a complex question.

232
00:12:21,900 --> 00:12:25,600
He asked, I, Her to even
describe it cause he lost me and

233
00:12:25,600 --> 00:12:28,600
chat GPT returned.
An accurate answer of what he

234
00:12:28,600 --> 00:12:31,300
was saying with an example line
of the code.

235
00:12:31,500 --> 00:12:34,800
It wasn't a finalized version
but it was enough to impress

236
00:12:34,800 --> 00:12:36,600
him.
So yeah, some people are

237
00:12:36,600 --> 00:12:38,500
impressed.
Some people like Sam are

238
00:12:38,500 --> 00:12:41,500
certainly not.
So after that delving, deep into

239
00:12:41,500 --> 00:12:43,800
the difference between machine
learning and AI.

240
00:12:43,800 --> 00:12:48,000
We go on to question, 2.
And Julia asked us, what type of

241
00:12:48,000 --> 00:12:52,800
AI do you think has the most
promise for use in PPC and why?

242
00:12:53,000 --> 00:12:55,800
A service team and response
saying I think imagery but I

243
00:12:55,808 --> 00:12:58,300
don't know.
It's almost hard to perceive

244
00:12:58,400 --> 00:13:00,900
where we will be with regards to
this.

245
00:13:00,900 --> 00:13:04,500
I do think Google will lean in
and promise less data more

246
00:13:04,500 --> 00:13:07,500
Automation and then Ryan scholar
response to question 2 saying

247
00:13:07,500 --> 00:13:11,600
the most obvious AI tools would
be around keyword research and

248
00:13:11,600 --> 00:13:14,900
AI copy creation.
As I think those are quite

249
00:13:14,900 --> 00:13:18,300
simple but I wouldn't be
surprised if we use some really

250
00:13:18,300 --> 00:13:22,600
cool and advanced tools out by
the end of next year, Debbie not

251
00:13:22,600 --> 00:13:23,700
in response.
It's question, Tuesday.

252
00:13:23,700 --> 00:13:26,400
And programmatic, paid search is
gaining much traction.

253
00:13:26,800 --> 00:13:31,100
I'm seeing High exposure, medium
ctrs, and high cpcs.

254
00:13:31,400 --> 00:13:35,200
If it leads to a sale, then it's
good Gil response to question 2

255
00:13:35,200 --> 00:13:39,000
saying, I think there would be a
place for a tool that optimizes

256
00:13:39,000 --> 00:13:41,300
shopping feeds in a less
annoying manner.

257
00:13:41,600 --> 00:13:46,000
But I've seen too many weird
edge cases to be super confident

258
00:13:46,000 --> 00:13:49,500
in how well it would do.
And this response to question to

259
00:13:49,500 --> 00:13:53,400
say and I think we will see tons
of AI Solutions pop up, Up this

260
00:13:53,400 --> 00:13:57,100
year, initial usage for PPC, is
on keyword research

261
00:13:57,100 --> 00:14:01,300
categorization, then it will
move towards add analysis and

262
00:14:01,300 --> 00:14:03,700
copy text.
And finally landing page

263
00:14:03,700 --> 00:14:07,300
drafting calculations response.
And I think, Chad GPT like

264
00:14:07,300 --> 00:14:10,200
applications as an assist to a
good.

265
00:14:10,200 --> 00:14:14,700
PPC, are are great for instant
in thinking up new headline

266
00:14:14,700 --> 00:14:17,400
ideas, or even in creating
landing page content.

267
00:14:17,500 --> 00:14:20,700
But it has limitations, and the
devil is in the details and

268
00:14:20,700 --> 00:14:24,500
still need someone to finalize.
Results and a mixed

269
00:14:24,500 --> 00:14:28,500
extraordinary response saying
100% in agreement, Rory

270
00:14:28,500 --> 00:14:30,900
response, same products, like
copied AI.

271
00:14:30,900 --> 00:14:35,500
That supplement tools in the PPC
is built anything that will

272
00:14:35,500 --> 00:14:40,100
automate compiled creates
eliminate or manual work a few

273
00:14:40,100 --> 00:14:43,000
more levers for us to pull.
But end of the day, the value to

274
00:14:43,000 --> 00:14:46,600
be in the advertisers brain and
got Christian response to

275
00:14:46,600 --> 00:14:48,100
question.
Teasing one of my clients

276
00:14:48,100 --> 00:14:51,300
recently sent me a list of
keywords ideas that were

277
00:14:51,300 --> 00:14:53,300
generated by chat, you Pete?
ET.

278
00:14:53,600 --> 00:14:56,000
And I thought that was an
interesting idea.

279
00:14:56,000 --> 00:14:59,900
I would hesitate to rely on AI
tools to write my ad copy.

280
00:14:59,900 --> 00:15:03,800
But using it during the reset
phase seems like a safer

281
00:15:03,800 --> 00:15:06,900
approach.
Chris Connor wall restaurants to

282
00:15:06,900 --> 00:15:09,400
question Tuesday.
And I think tools like pencil A,

283
00:15:09,400 --> 00:15:12,900
I can reduce the development
time for ad creative and

284
00:15:12,900 --> 00:15:16,900
increase the speed with which
campaigns go live and iterate

285
00:15:16,900 --> 00:15:19,700
testing.
This can also help Brands test

286
00:15:19,700 --> 00:15:22,800
into new channels who have
previously faced cost.

287
00:15:23,000 --> 00:15:26,300
Constraints and then we have
Harrison responding to question

288
00:15:26,300 --> 00:15:29,800
2 cm as I've been sitting here
thinking about it, I wonder if a

289
00:15:29,800 --> 00:15:33,600
I could work well for initial
keyword research at least to

290
00:15:33,600 --> 00:15:36,500
generate an initial list for
review.

291
00:15:36,500 --> 00:15:40,800
I think it has potential for
Creative assistance to over time

292
00:15:40,800 --> 00:15:45,100
I imagine the use cases will
grown phrase, Andrews response,

293
00:15:45,100 --> 00:15:48,900
a my guess ideally it's how a i
Sports things that we haven't

294
00:15:48,900 --> 00:15:53,200
obviously as humans we have the
context that AI doesn't So the

295
00:15:53,200 --> 00:15:56,100
combination between the two is
where I can see a.

296
00:15:56,100 --> 00:16:01,200
I come into its own and then I
respond in say yeah the AI that

297
00:16:01,200 --> 00:16:03,900
helps with ad copy.
I definitely think does have

298
00:16:03,900 --> 00:16:07,100
promise.
I do think it gives a clear help

299
00:16:07,100 --> 00:16:11,400
for those who Embrace Automation
and want to cut coming up with

300
00:16:11,400 --> 00:16:15,400
creative Time by a lot.
So yeah if you want to reduce

301
00:16:15,400 --> 00:16:18,300
the amount of time it takes to
come up with a creative, I

302
00:16:18,300 --> 00:16:19,900
think.
Yeah, definitely use these AI

303
00:16:19,900 --> 00:16:22,800
tools, Julie response, question
Tuesday, and idea of something.

304
00:16:23,000 --> 00:16:26,800
Chad GPT that can generate a ton
of ideas for ad copy landing

305
00:16:26,800 --> 00:16:29,100
pages.
Etc is intriguing.

306
00:16:29,200 --> 00:16:32,000
The models being up-to-date is a
concern though.

307
00:16:32,300 --> 00:16:36,900
Chad GPT data only goes up onto
2020, or maybe it's 2021.

308
00:16:37,300 --> 00:16:40,000
That's a huge gap.
Yes, definitely, that is a huge

309
00:16:40,000 --> 00:16:42,100
gap.
Julia voice response to question

310
00:16:42,100 --> 00:16:45,100
2 saying hot take.
I don't know if the type of AI

311
00:16:45,400 --> 00:16:47,400
with the most promise has come
out yet.

312
00:16:47,400 --> 00:16:50,200
Huh.
The way people search or deal

313
00:16:50,200 --> 00:16:52,800
with potential for a our results
tells?

314
00:16:52,900 --> 00:16:56,500
Me that the most promising, AI
is the deepest, surveillance a

315
00:16:56,500 --> 00:17:00,400
hum learning tools.
That platforms can devise

316
00:17:01,000 --> 00:17:03,100
action, cock response to
question Tuesday, and I've been

317
00:17:03,100 --> 00:17:06,099
playing around with Todd GPT,
and it's pretty impressive.

318
00:17:06,200 --> 00:17:10,000
Definitely couldn't be used.
As the end-all-be-all IE.

319
00:17:10,000 --> 00:17:13,700
I wouldn't just copy and paste
the output and deliver it to a

320
00:17:13,700 --> 00:17:16,099
client.
But it could be a great tool to

321
00:17:16,099 --> 00:17:19,400
start brainstorming copy and
keyword ideas.

322
00:17:19,900 --> 00:17:22,000
So if you log in response to
question to say, I'd probably

323
00:17:22,000 --> 00:17:25,599
say keyword Search.
But, with how loose keywords are

324
00:17:25,599 --> 00:17:29,500
getting nowadays, I'd still say
it would need a fair amount of

325
00:17:29,500 --> 00:17:30,600
human guidance.
Yeah.

326
00:17:30,600 --> 00:17:32,400
Definitely hearts and goes.
Yeah.

327
00:17:32,400 --> 00:17:36,100
Today's keyword matching type
seems to be more like this exact

328
00:17:36,100 --> 00:17:38,900
is equals to phrase.
Phrase is actually broad.

329
00:17:38,900 --> 00:17:41,200
Broad whatever.
Google feels like yeah.

330
00:17:41,400 --> 00:17:43,800
And so he goes so glad my
current clients don't have

331
00:17:43,800 --> 00:17:47,300
interest in seeing their Search
terms reports that will take

332
00:17:47,300 --> 00:17:50,000
some explaining.
Yeah, that's no kidding Harrison

333
00:17:50,000 --> 00:17:52,800
goes sometimes I have to explain
it to myself.

334
00:17:52,900 --> 00:17:54,500
Self.
Yeah, looking at some of the

335
00:17:54,508 --> 00:17:57,300
things that come out and said,
query reports can be shocking

336
00:17:57,600 --> 00:18:00,900
and in response to Julie where
she answered question to saying,

337
00:18:00,900 --> 00:18:04,800
that being up to date, is a
concern for the fact that GPT

338
00:18:04,800 --> 00:18:08,700
data only goes up until 2020 or
2021 Harrison goes.

339
00:18:08,700 --> 00:18:11,700
Yeah, I like that.
It can generate a lot of ideas

340
00:18:11,700 --> 00:18:13,900
really quickly.
They may not be unique but it

341
00:18:13,900 --> 00:18:16,700
gives you a great head start and
you can work from there.

342
00:18:17,000 --> 00:18:19,500
Julie goes for me.
I want AI or machine learning to

343
00:18:19,500 --> 00:18:21,800
make things easier or more
efficient.

344
00:18:21,800 --> 00:18:25,200
So if it can generate Two ton of
ideas that I can use to write

345
00:18:25,200 --> 00:18:29,200
ideas or landing page copy.
That's a great tool and

346
00:18:29,200 --> 00:18:31,800
time-saver.
Yeah, in this industry, we just

347
00:18:31,800 --> 00:18:34,800
want to save time.
Mark, Suba response to question

348
00:18:34,800 --> 00:18:38,000
2, saying the ad copy of landing
page suggestion from Chad, GPT

349
00:18:38,000 --> 00:18:40,000
have been.
Okay, but not great.

350
00:18:40,000 --> 00:18:42,200
We use it more as an assist
mechanism.

351
00:18:42,200 --> 00:18:45,700
Right now, to give us ideas, we
haven't thought of and then

352
00:18:45,700 --> 00:18:50,000
tweak it a bit to make more
sense and then we go on to

353
00:18:50,000 --> 00:18:53,000
question 3 from Julia.
She asked what type of a A, I

354
00:18:53,000 --> 00:18:56,600
do, you think is potentially
most problematic for PPC?

355
00:18:56,600 --> 00:19:01,300
And why rounds: response says, I
as a few people have already

356
00:19:01,300 --> 00:19:04,900
mentioned, I think it's more
problematic when people take the

357
00:19:04,900 --> 00:19:08,100
output from Ai and apply it
without checking it over first,

358
00:19:08,100 --> 00:19:12,200
these AI tools can be great to
give you a head start, but it

359
00:19:12,200 --> 00:19:14,500
still needs a human to check
things over.

360
00:19:14,500 --> 00:19:16,300
And I responded to that saying
that.

361
00:19:16,300 --> 00:19:17,900
Yeah, I was going to say it's
the similar thing.

362
00:19:17,900 --> 00:19:20,800
It's not necessarily about the
eyes of other people but then I

363
00:19:20,800 --> 00:19:22,800
thought this is more to do with
problem.

364
00:19:22,900 --> 00:19:27,300
And people not problematic Ai.
And he goes, He agrees, girl,

365
00:19:27,300 --> 00:19:29,400
response him.
Anything that involves Google

366
00:19:29,400 --> 00:19:32,900
and budgets.
And this goes AI is a tool usage

367
00:19:32,900 --> 00:19:36,600
of AI gets problematic when
someone else is using it and you

368
00:19:36,600 --> 00:19:39,700
are not in control, in my
opinion, robbery response to

369
00:19:39,700 --> 00:19:44,300
question three, same faulty data
sets containing false data, I am

370
00:19:44,300 --> 00:19:48,200
skeptical of types of AI or
machine learning that claims to

371
00:19:48,200 --> 00:19:52,200
scrape data in XYZ and provides
you with Niche audience

372
00:19:52,200 --> 00:19:54,200
information.
Ian for better targeting.

373
00:19:54,400 --> 00:19:57,500
I think there will be an
opportunity for tools to be

374
00:19:57,500 --> 00:20:01,600
created that audit the quality
of data and serious team

375
00:20:01,600 --> 00:20:05,100
response to question. 3 saying
well if you have used chat GPT,

376
00:20:05,100 --> 00:20:07,800
some answers are just factually
incorrect.

377
00:20:07,800 --> 00:20:11,800
So incorrect copy claims and
then the legal aspects could

378
00:20:11,800 --> 00:20:15,700
definitely be a concern and
factor and then we have Julie

379
00:20:15,700 --> 00:20:18,300
virus responding to question 3
saying, I mean, replacing

380
00:20:18,300 --> 00:20:21,700
keywords with audiences plus
predictive Behavior Niche

381
00:20:21,700 --> 00:20:26,600
searches and Built-in Nuance are
going the way of the dodo,

382
00:20:27,000 --> 00:20:29,200
interesting.
And then we have Ashton clock

383
00:20:29,200 --> 00:20:33,000
responding to question 3 saying,
I think creative could be hugely

384
00:20:33,000 --> 00:20:35,600
problematic.
A I can pull images from other

385
00:20:35,600 --> 00:20:39,300
artists, use it in their
outputs, but they at anyone

386
00:20:39,300 --> 00:20:43,300
using that generated image don't
actually have the rights to use

387
00:20:43,300 --> 00:20:45,600
the artists work.
Interesting.

388
00:20:45,700 --> 00:20:48,700
And then Robert, Brady response
to question 3, saying, I think a

389
00:20:48,700 --> 00:20:51,100
online will always have a bias
problem.

390
00:20:51,100 --> 00:20:52,700
It could be coming from the data
set.

391
00:20:52,800 --> 00:20:56,100
Set or from assumptions made in
development, Etc.

392
00:20:56,100 --> 00:20:59,700
They will work to minimize it
but it will always be there

393
00:21:00,200 --> 00:21:01,600
during Brown response to
question 2.

394
00:21:01,600 --> 00:21:06,000
Saying, if a I can understand
humans and context data analysis

395
00:21:06,100 --> 00:21:09,000
could be super interesting
person meets machine.

396
00:21:09,500 --> 00:21:11,600
So if you look in response to
question 3, say and I think the

397
00:21:11,608 --> 00:21:15,200
concept in general could be
problematic for agencies if a

398
00:21:15,208 --> 00:21:18,600
business feels comfortable,
using all these AI tools and

399
00:21:18,600 --> 00:21:21,400
automation options then they
might feel.

400
00:21:21,400 --> 00:21:22,700
It's best to just do it
themselves.

401
00:21:23,500 --> 00:21:25,500
Dwayne Brown response to
question 3 saying, see my

402
00:21:25,500 --> 00:21:29,200
responses to it's a question to
Tech is bad at understanding

403
00:21:29,500 --> 00:21:33,400
humans and context which is why
it struggles right now.

404
00:21:33,800 --> 00:21:36,300
Harrison response question, 3
saying anything within a black

405
00:21:36,300 --> 00:21:40,600
box operating without oversight
or without us being able to

406
00:21:40,600 --> 00:21:43,100
understand processes is
concerning.

407
00:21:43,100 --> 00:21:45,600
That is why I don't like
Google's automated creative

408
00:21:45,600 --> 00:21:48,100
option.
You can get some really

409
00:21:48,100 --> 00:21:52,300
problematic combination if it's
not watched Julie, Dan response

410
00:21:52,300 --> 00:21:53,400
to question.
Three saying.

411
00:21:53,400 --> 00:21:56,900
I think the biggest issue will
always be that, what is best for

412
00:21:56,900 --> 00:22:00,600
the platforms, and what is best
for its advertisers can never

413
00:22:00,600 --> 00:22:04,300
aligned 100%.
So when you bring a tie or even

414
00:22:04,300 --> 00:22:08,000
machine learning into the
equation, that bias is very much

415
00:22:08,000 --> 00:22:11,400
baked in and it all sucks at
Nuance correctly.

416
00:22:12,000 --> 00:22:13,800
And then we have actor
responding to question three

417
00:22:13,800 --> 00:22:15,400
saying.
Anything that can get into

418
00:22:15,400 --> 00:22:19,600
competitors insights or AI that
can degrade other brands done

419
00:22:19,600 --> 00:22:25,300
out of Market competition and M
is Programmatic then with copies

420
00:22:25,400 --> 00:22:28,800
or creative as well.
It will need human touch until

421
00:22:28,800 --> 00:22:32,700
it becomes Ai and machine
learning and human emotions.

422
00:22:32,700 --> 00:22:34,800
Are taking all of those things
into consideration.

423
00:22:35,400 --> 00:22:37,200
And yeah, jelly goes.
I'm loving the conversation

424
00:22:37,200 --> 00:22:39,100
today as she goes on to question
for.

425
00:22:39,100 --> 00:22:43,000
She asks, what kind of AI would
you like to see developed or

426
00:22:43,000 --> 00:22:47,900
deployed and PPC rural response?
I'd love to develop a tool that

427
00:22:47,900 --> 00:22:51,600
allows small businesses local
services to create their own

428
00:22:51,600 --> 00:22:54,000
campaigns?
For those who can't afford a

429
00:22:54,000 --> 00:22:59,000
hiring PPC is so think to add
keywords, budgets few pieces of

430
00:22:59,000 --> 00:23:02,400
AD copy discussing businesses to
compiles.

431
00:23:02,400 --> 00:23:06,200
Plus launches tool provides
weekly reports girl responds in

432
00:23:06,200 --> 00:23:10,300
AI which helps us identify
granularly, which keywords or

433
00:23:10,300 --> 00:23:15,000
campaigns show up for a nice
start and so basically better

434
00:23:15,000 --> 00:23:17,700
sqrl.
So basically one that brings

435
00:23:17,700 --> 00:23:19,900
back.
Transparency in cars and goes

436
00:23:20,300 --> 00:23:22,600
Debbie response.
I'd particularly like to see

437
00:23:22,800 --> 00:23:24,900
Same stats or manual
intervention.

438
00:23:24,900 --> 00:23:29,300
We G smart campaigns as we do
with regular G search campaigns.

439
00:23:29,700 --> 00:23:32,000
And to this question, I
answered, yes, to be fair.

440
00:23:32,000 --> 00:23:36,100
I'm more in the process of work
on Automation and AI that has

441
00:23:36,100 --> 00:23:38,700
been currently dispatch.
That still needs a lot of

442
00:23:38,700 --> 00:23:42,900
improvement first before looking
to deploy, anything new.

443
00:23:42,900 --> 00:23:46,200
So work on, you know, giving
them more transparency in the AI

444
00:23:46,200 --> 00:23:49,000
and automation has already been
worked on and Tim Halloran

445
00:23:49,000 --> 00:23:52,600
response saying yes.
How about polish up P Max?

446
00:23:52,700 --> 00:23:55,100
Before working on anything new.
Yeah.

447
00:23:55,100 --> 00:23:57,900
And do that.
Well before launching it, and

448
00:23:57,900 --> 00:24:00,100
then we have Julie responding to
question forcing.

449
00:24:00,100 --> 00:24:03,800
I would love an AI for PPC,
competitor, research input, your

450
00:24:03,800 --> 00:24:07,100
client or brand, and then have
options to input known

451
00:24:07,100 --> 00:24:10,600
competitors or have it find
competitors in sites or social

452
00:24:10,800 --> 00:24:14,700
some tools have parts of this
but not across everything.

453
00:24:14,700 --> 00:24:17,800
Interesting, and we have actor
responding saying you have some

454
00:24:17,800 --> 00:24:20,900
apis like this Julie where you
can pull this out.

455
00:24:20,900 --> 00:24:24,200
Still need some facelift, but
Good for the start.

456
00:24:24,500 --> 00:24:26,800
What kind of apis are doing this
with tools?

457
00:24:26,800 --> 00:24:30,300
Are doing this at Target, are
some resources and we have

458
00:24:30,300 --> 00:24:32,300
Kristen responding to question
forcing.

459
00:24:32,300 --> 00:24:36,300
I would like to see AI in smart
reporting where it can quickly

460
00:24:36,300 --> 00:24:41,500
pull up analysis on any major
metric in PPC, at a given

461
00:24:41,600 --> 00:24:44,100
moment.
And Tim Herron response saying,

462
00:24:44,100 --> 00:24:46,500
I'd love placement
recommendations based on

463
00:24:46,500 --> 00:24:50,500
industry.
Oh, this is an HR our saas

464
00:24:50,500 --> 00:24:54,600
product, consider adding bid.
Fires to these websites YouTube

465
00:24:54,600 --> 00:24:58,500
channels apps that have the
highest conversion rate for your

466
00:24:58,500 --> 00:25:00,800
industry.
All right, so if you look in

467
00:25:00,800 --> 00:25:03,800
response, question for saying,
competitor research, although at

468
00:25:03,800 --> 00:25:07,300
the same time, I wouldn't like
my clients competitors, having

469
00:25:07,300 --> 00:25:10,500
more insights about them,
absolutely?

470
00:25:10,500 --> 00:25:11,600
Right.
Yeah.

471
00:25:11,600 --> 00:25:14,400
It's a double-edged sword.
That one and Harrison response

472
00:25:14,400 --> 00:25:17,500
to question for saying research
takes a lot of time and can be

473
00:25:17,500 --> 00:25:19,700
expensive whether that's
competitor.

474
00:25:19,700 --> 00:25:23,800
Keyword creative, excetera, AI
back and In those areas.

475
00:25:24,100 --> 00:25:28,000
And then provide information for
a human to analyze would be

476
00:25:28,000 --> 00:25:30,300
great.
And then Ashton clock response

477
00:25:30,300 --> 00:25:32,600
to question, forcing
competition, and keyword

478
00:25:32,600 --> 00:25:37,200
research would be great to use
cases, for AI, chat GPT can help

479
00:25:37,200 --> 00:25:41,900
generate keyword ideas but the
volume and cpc's aren't accurate

480
00:25:41,900 --> 00:25:47,000
because the date they use only
goes up until 2021 asking about

481
00:25:47,000 --> 00:25:50,900
competitors can also give
inaccurate answers for the same

482
00:25:50,900 --> 00:25:52,600
reason.
So Julie.

483
00:25:52,700 --> 00:25:56,000
Dan asks, do you have any
ethical concerns about the use

484
00:25:56,000 --> 00:26:00,500
of AI in PPC or even in AI in
general?

485
00:26:00,800 --> 00:26:03,400
Gil goes no.
I will be the first person to

486
00:26:03,400 --> 00:26:08,100
embrace the power of true Ai and
set up a kingdom of servers hard

487
00:26:08,100 --> 00:26:11,900
coded to my DNA in order to
dominate the internet.

488
00:26:11,900 --> 00:26:16,200
Ooh, that sounds a bit painful.
Debbie, no time response in Ai

489
00:26:16,200 --> 00:26:20,500
and PPC can be complacency with
agencies with the set it and

490
00:26:20,500 --> 00:26:23,900
forget it approached in PPC.
AI in general?

491
00:26:23,900 --> 00:26:29,900
Like the Matrix on steroids and
mind-boggling and this response

492
00:26:29,900 --> 00:26:31,800
AI place is a lot of ethical
problems.

493
00:26:31,800 --> 00:26:36,200
It is bringing us into the post.
Truth error answers are no

494
00:26:36,200 --> 00:26:39,300
longer scientific.
They are probabilistic copyright

495
00:26:39,300 --> 00:26:42,300
has gone, powerful tools can and
will be misused.

496
00:26:42,300 --> 00:26:47,100
People will get lazy idiocracy
and conspiracy will rise.

497
00:26:47,200 --> 00:26:49,400
Oh, that sounds very conspiracy
theorist.

498
00:26:49,600 --> 00:26:51,900
We then have Julie answering
this question, five, same

499
00:26:51,900 --> 00:26:54,500
history has shown On that all
models include bias.

500
00:26:54,700 --> 00:26:57,600
So that is concerning.
And if you're ever running a

501
00:26:57,608 --> 00:27:00,600
campaign on the Google Display
Network the idea of the internet

502
00:27:00,600 --> 00:27:04,200
being the data set is
terrifying, learn skills.

503
00:27:04,200 --> 00:27:06,100
Yes.
Google or any other company, who

504
00:27:06,100 --> 00:27:09,400
owns the AI will inherently
wants to increase their own

505
00:27:09,400 --> 00:27:13,100
Revenue, which will bias.
The machine AI is cool, but our

506
00:27:13,100 --> 00:27:17,100
society is still very driven by
money and what can it do for me?

507
00:27:17,300 --> 00:27:20,100
And it goes, yes, this is
important to understand.

508
00:27:20,200 --> 00:27:22,500
It's been fed with a lot of
Wikipedia bid.

509
00:27:22,800 --> 00:27:26,400
Also some reddits, right?
And then we have device

510
00:27:26,400 --> 00:27:28,500
responding saying, why?
Yes, I do.

511
00:27:28,800 --> 00:27:32,500
I outlined a few hair and she
puts it in her post.

512
00:27:32,800 --> 00:27:35,800
Yeah, I'll give a link to.
It is a very long post and that

513
00:27:35,800 --> 00:27:38,900
will make the timing of this
episode.

514
00:27:38,900 --> 00:27:41,500
Double.
If I decide to read it out, I'll

515
00:27:41,500 --> 00:27:44,300
put a link to her opinions on
the show notes to make sure you

516
00:27:44,300 --> 00:27:48,100
check that out on the marketing.
The a noob forward slash podcast

517
00:27:48,100 --> 00:27:51,000
site Tim Halloran response
saying, I think.

518
00:27:51,000 --> 00:27:52,500
So sing will be important at
all.

519
00:27:52,700 --> 00:27:56,800
Also think combining AI with
auctions could serve a board of

520
00:27:56,800 --> 00:27:59,200
directors.
Well, if tampered with, so

521
00:27:59,200 --> 00:28:04,000
regulating that may be important
in the future to Ashton clock

522
00:28:04,000 --> 00:28:07,300
response in, yes, it can be a
helpful tool if used carefully,

523
00:28:07,300 --> 00:28:10,800
but people are biased, which
means that the data API uses

524
00:28:10,800 --> 00:28:14,100
will be biased as well
plagiarism and using other

525
00:28:14,100 --> 00:28:16,300
people's work without their
consent.

526
00:28:16,300 --> 00:28:20,300
Is also an issue Lawrence.
Chase respond saying it is just

527
00:28:20,300 --> 00:28:22,500
a matter of time before Rai
overloads.

528
00:28:22,700 --> 00:28:26,200
Over Halal.
The truth is AI is going to be

529
00:28:26,200 --> 00:28:29,400
huge and probably quicker than
anyone wants to hear.

530
00:28:29,400 --> 00:28:32,600
The writing is on the wall.
The investment companies are

531
00:28:32,600 --> 00:28:36,600
making are for a reason, change
is coming.

532
00:28:36,600 --> 00:28:39,200
So yeah, important message for
all of us.

533
00:28:39,400 --> 00:28:41,200
So, if you Logan goes, my
concern is that.

534
00:28:41,200 --> 00:28:44,300
Some agencies are going to, put
everything through Ai, and do

535
00:28:44,300 --> 00:28:47,700
very little, in terms of
bringing added value strategy

536
00:28:47,700 --> 00:28:50,400
and consultation for their
clients, especially, they are

537
00:28:50,400 --> 00:28:54,400
smaller contacts.
Lee I've sadly already seen this

538
00:28:54,400 --> 00:28:57,900
on a lot of odets.
I do the time they spend on the

539
00:28:57,900 --> 00:29:01,200
account strategy is a lot lower
than the time.

540
00:29:01,200 --> 00:29:04,400
They actually charge for rather
than using it to free up.

541
00:29:04,400 --> 00:29:09,700
Time to use for other things it
helps them make more profits and

542
00:29:09,700 --> 00:29:11,100
actor response saying you are
now.

543
00:29:11,100 --> 00:29:14,500
I guess I would rather be some
of the first ones to run towards

544
00:29:14,500 --> 00:29:16,800
it.
Like you run towards water on a

545
00:29:16,800 --> 00:29:20,400
beach and then our final
question from Julie is what AI

546
00:29:20,400 --> 00:29:24,200
tools are out there that you've
Either used or wanted to try

547
00:29:24,200 --> 00:29:28,100
being trying to capture them as
you've shared them in the chat

548
00:29:28,100 --> 00:29:30,500
to go response.
I've experimented a lot with

549
00:29:30,500 --> 00:29:34,300
Chad TPT, but to be honest, it
started to feel like I'm talking

550
00:29:34,300 --> 00:29:36,900
to a low-bottom mised
politician.

551
00:29:37,200 --> 00:29:40,900
My grinder is a rather new
add-on for Google Sheets is

552
00:29:40,900 --> 00:29:44,200
simple ml for Sheets.
If you go to work spaces, don't

553
00:29:44,200 --> 00:29:47,500
google.com, I'll put that link
in the show notes as well.

554
00:29:47,500 --> 00:29:50,900
Loads of links shared hair and
yeah, and there's manchus chat

555
00:29:50,900 --> 00:29:54,000
GPT as well.
Well, and he goes, I think tons

556
00:29:54,000 --> 00:29:57,000
of AI applications will emerge
in 2023.

557
00:29:57,000 --> 00:30:01,800
So yeah, hold on to your hats,
we got AI coming out this year

558
00:30:01,800 --> 00:30:04,400
as well.
So, yeah, on that note, we've

559
00:30:04,400 --> 00:30:08,000
come to the end of the chat.
So that's our third chart of

560
00:30:08,000 --> 00:30:10,000
2023.
We've gotten through talking

561
00:30:10,000 --> 00:30:13,100
about Ai and PPC.
I using some of it are there.

562
00:30:13,100 --> 00:30:16,200
Some that you're seeing that we
have mentioned that you think

563
00:30:16,200 --> 00:30:18,300
that we should check out.
I think there's one called

564
00:30:18,300 --> 00:30:21,800
Jasper that AI that also helps
with writing of content.

565
00:30:21,800 --> 00:30:25,500
I think something like I will
also help writing ad copy ideas

566
00:30:25,800 --> 00:30:27,800
as well.
But yeah, most of the AI tools

567
00:30:27,800 --> 00:30:29,000
that people are talking about
here.

568
00:30:29,000 --> 00:30:32,200
It's pretty much chat GPT I
think.

569
00:30:32,200 --> 00:30:35,900
So if he goes Shay, she's likes
testing the images based tools.

570
00:30:36,200 --> 00:30:39,000
That's mainly, just for personal
interest.

571
00:30:39,000 --> 00:30:42,000
She's going to try anything.
That seems interesting and if

572
00:30:42,000 --> 00:30:45,400
it's free.
So, yeah, I know we're all about

573
00:30:45,400 --> 00:30:48,600
using free tools as well here.
I hope you've taken some great

574
00:30:48,600 --> 00:30:51,600
takeaways from this chat as
well, and if you want to

575
00:30:51,600 --> 00:30:55,000
continue the chatter If you have
more ideas about Ai and PPC,

576
00:30:55,200 --> 00:30:57,600
please join us on Thursday at
5:00 p.m.

577
00:30:57,600 --> 00:31:01,700
GMT. 12 p.m. eastern time as we
continue this chat, on Twitter

578
00:31:01,700 --> 00:31:05,100
spaces, for any feedback about
the podcast, or even maybe a

579
00:31:05,100 --> 00:31:07,000
correction on anything that has
been shared.

580
00:31:07,000 --> 00:31:09,100
Do get in.
Touch with me on Twitter, my

581
00:31:09,100 --> 00:31:13,300
DM's are open on my handle the
marketing, a new or LinkedIn.

582
00:31:13,300 --> 00:31:16,600
Find me on a neurotic Bola, I'll
be happy to chat with you.

583
00:31:16,900 --> 00:31:19,700
It also be great to for any
long-standing members or brand

584
00:31:19,700 --> 00:31:22,500
new listeners to give the
podcast a lovely rate.

585
00:31:22,800 --> 00:31:26,800
It's very quickly, you can do 45
out of Star Racing, I would be

586
00:31:26,800 --> 00:31:29,500
nice on Spotify, maybe just go
to link.

587
00:31:29,500 --> 00:31:34,200
Tree.com, /p P, c-- chat,
underscore Roundup.

588
00:31:34,300 --> 00:31:38,400
And also just before I leave you
with our outro, I we've also got

589
00:31:38,400 --> 00:31:41,800
the PVC live UK event.
The number four, the fourth one,

590
00:31:42,200 --> 00:31:45,000
first, one of the 2023 happening
on February.

591
00:31:45,000 --> 00:31:47,900
The 2nd tickets are only 16
pounds.

592
00:31:47,900 --> 00:31:50,300
As I mentioned earlier on
Sophie, Logan, one of our

593
00:31:50,300 --> 00:31:54,100
amazing PPC chat contributors
He's going to be talking about

594
00:31:54,100 --> 00:31:57,600
Microsoft ad Network.
We've also got Izzy mammal.

595
00:31:57,600 --> 00:32:01,200
Who going to be talking about
how PPC Works within your

596
00:32:01,200 --> 00:32:03,800
marketing strategy, we're going
to be have Maddie Crowfoot

597
00:32:03,800 --> 00:32:07,000
talking about, ad copy, and
we've also got Tiffany shares.

598
00:32:07,000 --> 00:32:10,100
Also joins the PPC chat
conversation sometimes talking

599
00:32:10,100 --> 00:32:13,200
about GTM.
So that's Google tag manager.

600
00:32:13,200 --> 00:32:17,200
So if you want to learn all
about that, go to PC live UK.com

601
00:32:17,400 --> 00:32:21,400
for your 16-pound tickets where
you'll get drinks and food and

602
00:32:21,400 --> 00:32:24,300
pizza and snacks.
Some swag that I've got some

603
00:32:24,300 --> 00:32:28,500
nice tote bags.
So nice charges so nice pens.

604
00:32:28,600 --> 00:32:32,000
And some mugs, I'm going to give
away as a prize for many

605
00:32:32,000 --> 00:32:33,300
competition.
I'm going to do.

606
00:32:33,300 --> 00:32:36,600
So yeah, if any of those things
sound interesting to you, please

607
00:32:36,600 --> 00:32:39,900
go to PC live UK.com for your
ticket.

608
00:32:40,000 --> 00:32:43,900
So remember your can pay for
your campaigns and businesses to

609
00:32:43,900 --> 00:32:45,600
Glide smoothly.
There's a lot of hard work

610
00:32:45,600 --> 00:32:48,900
needed beneath the surface.
So keep your son's kicking by

611
00:32:48,900 --> 00:32:50,000
and speak to you next week.