Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:00,000 --> 00:00:02,940
Thank you Mark, and thank you,
um, everybody for being here.
2
00:00:03,000 --> 00:00:07,470
Um, but first I gotta say this, it's great
to see so many of my friends on here.
3
00:00:07,620 --> 00:00:08,310
Hey everybody.
4
00:00:08,315 --> 00:00:09,360
It's good to see you guys.
5
00:00:09,360 --> 00:00:10,050
Nice to see y'all.
6
00:00:10,050 --> 00:00:10,800
Nice to see y'all.
7
00:00:10,950 --> 00:00:12,480
Thanks for being, being here.
8
00:00:12,810 --> 00:00:18,060
Um, as we dive into this and also,
man, I've gone through the content that
9
00:00:18,065 --> 00:00:25,650
y'all have gone through and it has been
amazing, like, Just amazing all this
10
00:00:25,650 --> 00:00:29,610
stuff that you've learned so far and,
and so I'm, I'm super excited about
11
00:00:29,610 --> 00:00:33,330
that, and I'm super excited to be a
part of it and take it the next level.
12
00:00:33,690 --> 00:00:34,080
Okay.
13
00:00:34,440 --> 00:00:37,110
Uh, let me figure out
sharing my screen here.
14
00:00:39,210 --> 00:00:43,200
Now, I, I want to be completely
honest with you guys.
15
00:00:43,230 --> 00:00:44,280
Can y'all see my screen?
16
00:00:45,030 --> 00:00:46,050
No, probably not.
17
00:00:48,480 --> 00:00:48,870
Hold on.
18
00:00:52,334 --> 00:00:56,714
So one of the beauties here is when you
are an old school geek or even a marketer,
19
00:00:56,775 --> 00:00:59,655
you might be good at those things, but
you're not always good at things like
20
00:00:59,655 --> 00:01:03,915
hitting the share button when it's time
to share screen, which is what I just did.
21
00:01:10,095 --> 00:01:11,295
All right, can y'all see that?
22
00:01:11,295 --> 00:01:12,795
Can I get a thumbs up if you all can see?
23
00:01:12,795 --> 00:01:13,335
See the screen?
24
00:01:13,335 --> 00:01:14,175
Fantastic.
25
00:01:15,375 --> 00:01:15,675
Okay.
26
00:01:16,035 --> 00:01:16,515
Um.
27
00:01:17,130 --> 00:01:20,910
So I'd like to start, start here by
saying, um, my team knows, I don't usually
28
00:01:20,910 --> 00:01:24,539
talk a ton about myself, but I do wanna
give a little bit of a background today
29
00:01:24,539 --> 00:01:27,360
as I start to kind of frame some of this.
30
00:01:27,810 --> 00:01:31,440
Um, I started coding when
I was in seventh grade.
31
00:01:32,730 --> 00:01:38,670
In ninth grade I took a class called,
uh, analytics and Data Structures.
32
00:01:39,435 --> 00:01:40,905
And in that class, Mr.
33
00:01:40,905 --> 00:01:44,805
Yee, over the course of a year took
us through a process by the end
34
00:01:44,805 --> 00:01:48,855
of the year where we had a test
and had to produce a program that
35
00:01:48,860 --> 00:01:51,825
he would give us a set of data.
36
00:01:52,245 --> 00:01:52,665
Right.
37
00:01:52,815 --> 00:01:56,895
It was just this big set of data and, and
we call it big cuz we were 14 years old.
38
00:01:56,895 --> 00:01:58,605
It was like 50 terms.
39
00:01:58,755 --> 00:02:02,955
It was this big set of data
and he then would say, okay.
40
00:02:03,675 --> 00:02:08,985
With this data, I am going to give
you a term and I want you to write
41
00:02:08,985 --> 00:02:13,335
a computer program that would
predict which term I'd say next.
42
00:02:14,895 --> 00:02:19,905
Now, we did that by then also having
other data that said, Hey, oftentimes
43
00:02:20,385 --> 00:02:27,945
dog followed cat, but if you said
bird and then dog or bird and
44
00:02:27,945 --> 00:02:30,495
then cat, then fish came, not cat.
45
00:02:32,459 --> 00:02:37,530
What we didn't know back then was
he was teaching us literally the
46
00:02:37,530 --> 00:02:45,510
super, super, super, super, super
junior version of what AI does and
47
00:02:45,510 --> 00:02:47,370
teaching us to think in that way.
48
00:02:47,370 --> 00:02:49,980
So I, I start today by
saying thank you to Mr.
49
00:02:49,980 --> 00:02:50,040
Y.
50
00:02:51,105 --> 00:02:54,315
For being able to do this
presentation based on something
51
00:02:54,315 --> 00:02:56,984
that he taught me God, years ago.
52
00:02:57,405 --> 00:02:57,825
Okay.
53
00:02:58,155 --> 00:02:59,864
Um, won't give up how long that was.
54
00:03:00,225 --> 00:03:02,325
Um, and we're.
55
00:03:04,425 --> 00:03:06,435
Prompt engineering today.
56
00:03:06,435 --> 00:03:07,965
But first, lemme tell you what my goal is.
57
00:03:07,965 --> 00:03:11,205
My goal is to arm you with some
prompt fag fragments and modifiers
58
00:03:11,595 --> 00:03:17,955
that will help you enhance your G P
T output and answer your questions.
59
00:03:17,955 --> 00:03:22,665
So I wanna answer some of the questions
that you ha have had, um, about.
60
00:03:23,180 --> 00:03:27,980
Prompting about chat g p t about
the responses that you get about
61
00:03:27,980 --> 00:03:33,320
even why that wonderful thing of it
forgets why I do say certain things.
62
00:03:33,320 --> 00:03:34,130
Why is it forget?
63
00:03:34,850 --> 00:03:37,790
I wanna answer some of those questions
to help you understand what's actually
64
00:03:37,790 --> 00:03:42,410
going on so that you can write
better prompts to get better answers.
65
00:03:42,410 --> 00:03:43,370
Is that fair, everybody?
66
00:03:43,640 --> 00:03:48,019
If that's fair And if that's cool,
do me a favor, put, um, in the chat.
67
00:03:48,019 --> 00:03:48,800
Put it's fair.
68
00:03:48,920 --> 00:03:51,950
Let talk to me here if
that's cool, everybody.
69
00:03:54,315 --> 00:03:55,065
Fantastic.
70
00:03:55,125 --> 00:03:55,485
Thank you.
71
00:03:55,605 --> 00:03:55,935
Thank you.
72
00:03:55,935 --> 00:03:56,355
Thank you.
73
00:03:56,415 --> 00:03:57,195
Appreciate that.
74
00:03:57,195 --> 00:03:57,675
Super fair.
75
00:03:57,705 --> 00:03:58,335
I like that.
76
00:03:58,665 --> 00:03:59,475
Super fair.
77
00:03:59,535 --> 00:03:59,955
Thank you.
78
00:04:01,485 --> 00:04:08,235
Okay, now I got a bunch of slides and
I'll be honest with you and tell you
79
00:04:08,235 --> 00:04:12,285
that, that, um, Jeff Hunter last week
touched on quite a bit of this, and so I'm
80
00:04:12,285 --> 00:04:17,204
gonna run through some of what he touched
on, but if you have a question about
81
00:04:17,204 --> 00:04:21,435
something that he said or something that
I say, please stop me, ask the question.
82
00:04:22,859 --> 00:04:25,440
I wanna get to some more advanced
stuff, but I have to make sure that
83
00:04:25,440 --> 00:04:29,760
everybody's on the same page with
some of the baseline stuff that,
84
00:04:29,760 --> 00:04:32,099
that Jeff covered last week as well.
85
00:04:32,429 --> 00:04:32,789
Okay?
86
00:04:33,780 --> 00:04:38,880
All right, so, oh, next thing, before
I forget, I want you to drop some
87
00:04:38,880 --> 00:04:42,000
prompts, so I do want to help you
if we have time and if Mark will
88
00:04:42,000 --> 00:04:45,539
allow us at the end of this all.
89
00:04:46,275 --> 00:04:49,665
Okay, at the end of this all, and
Mark, I think I sent you the link
90
00:04:49,755 --> 00:04:52,575
to the, um, the Google sheet.
91
00:04:53,325 --> 00:04:56,325
Um, if you could drop that in the chat
for me cuz I forgot to copy the link.
92
00:04:56,715 --> 00:05:02,685
Um, I want you to drop the prompts
that have given you challenges.
93
00:05:03,375 --> 00:05:06,435
If you got a prompt you've been working
on, it's not doing what you wanted to do.
94
00:05:07,530 --> 00:05:13,020
Drop that prompt in there for me so
that I can, um, thank you, Michelle.
95
00:05:13,110 --> 00:05:16,950
See, that's why Michelle wears
the Superman hat like me.
96
00:05:17,460 --> 00:05:17,850
Okay.
97
00:05:17,910 --> 00:05:18,960
Um, thank you, Michelle.
98
00:05:19,380 --> 00:05:19,590
All right.
99
00:05:19,590 --> 00:05:20,190
Drop the prompts.
100
00:05:20,190 --> 00:05:21,420
Like, it's like, it's hot for me.
101
00:05:23,220 --> 00:05:27,240
All right, let's start off
with understanding the model.
102
00:05:28,230 --> 00:05:29,970
Um, and I want you to kinda understand.
103
00:05:30,420 --> 00:05:34,290
And take you back so you understand
what is actually happening.
104
00:05:34,290 --> 00:05:40,410
When you type something in to chat g
p t, you type something in and it's
105
00:05:40,410 --> 00:05:47,040
called the prompt and the computer quote
unquote, the computer takes that prompt
106
00:05:47,040 --> 00:05:49,950
and it tries to encode that prompt.
107
00:05:50,160 --> 00:05:54,630
That's a fancy word for saying get
rid of as many words that I don't
108
00:05:54,720 --> 00:05:56,760
need to pay attention to as possible.
109
00:05:58,320 --> 00:05:59,159
That's what it means.
110
00:05:59,520 --> 00:06:03,870
So you put in, how do I
start learning to code?
111
00:06:05,370 --> 00:06:07,200
And the computer encodes that too.
112
00:06:07,320 --> 00:06:08,940
How Learn code,
113
00:06:12,000 --> 00:06:16,050
it cuts all those other characters
out, all those other words out.
114
00:06:17,210 --> 00:06:18,480
It then takes.
115
00:06:18,870 --> 00:06:23,700
What it encoded and it decodes it and
tries to figure out, okay, now what order
116
00:06:23,700 --> 00:06:29,010
do we need to put this in so that we can
actually understand what they were asking?
117
00:06:31,650 --> 00:06:36,210
And then it performs the operation of,
in this case, answering your question.
118
00:06:36,590 --> 00:06:43,440
It gets that, and then it takes that
answer and then it encodes it back
119
00:06:43,440 --> 00:06:47,849
and generates an output for you and
sends that back for you to read.
120
00:06:49,200 --> 00:06:50,310
Does that make sense to everybody?
121
00:06:50,310 --> 00:06:51,570
Anybody have a question about this?
122
00:06:51,570 --> 00:06:55,380
Cause I want you to wanna make sure
you understand just this, we're not
123
00:06:55,380 --> 00:06:58,230
gonna do too much theory today, but
understand just this piece of theory.
124
00:06:58,710 --> 00:07:01,110
If you understand, gimme a
thumbs up in the chat, please.
125
00:07:01,110 --> 00:07:03,780
If you don't understand or you have
a question, gimme a thumbs down.
126
00:07:04,050 --> 00:07:04,690
Ask me the question.
127
00:07:10,410 --> 00:07:10,800
All right?
128
00:07:10,830 --> 00:07:11,490
Fantastic.
129
00:07:11,490 --> 00:07:12,300
Getting thumbs up.
130
00:07:12,300 --> 00:07:13,230
Appreciate that.
131
00:07:14,190 --> 00:07:15,000
Appreciate that.
132
00:07:17,850 --> 00:07:18,420
All right,
133
00:07:21,750 --> 00:07:27,030
so let's do on a couple
frustrations that people have.
134
00:07:27,035 --> 00:07:27,570
I hear it.
135
00:07:27,570 --> 00:07:29,070
People say it all the time.
136
00:07:29,520 --> 00:07:34,350
Um, one of them is, it forgets,
the darn thing forgets.
137
00:07:34,650 --> 00:07:37,590
Um, I was watching Jeff, and Jeff
was talking about how it forgot
138
00:07:37,590 --> 00:07:39,270
his proprietary framework on it.
139
00:07:39,270 --> 00:07:39,930
How do you forget?
140
00:07:40,020 --> 00:07:43,170
Then it asked him again, or he asked
it again, Hey, use my framework,
141
00:07:43,175 --> 00:07:45,300
and it was like, Still forgot.
142
00:07:46,080 --> 00:07:47,370
How did that happen?
143
00:07:47,850 --> 00:07:49,440
Why did that happen?
144
00:07:50,130 --> 00:07:54,120
Well, it happened because in the
model, especially when you're
145
00:07:54,120 --> 00:07:58,830
talking G P T three and 3.54 is
going to be a little bit different.
146
00:07:59,460 --> 00:08:04,860
But in the model, what's going on is,
number one, you put in a prompt, right?
147
00:08:05,340 --> 00:08:09,570
And it chopped out a whole bunch of words
like I just explained it, figured out what
148
00:08:09,570 --> 00:08:11,370
you wanted, and it gave you a response.
149
00:08:11,969 --> 00:08:13,080
Cool, we got that.
150
00:08:13,679 --> 00:08:19,620
But then it, it also looks at it and
says, Hey, I've got a maximum of 4,000
151
00:08:19,620 --> 00:08:23,460
words that I can remember or respond with.
152
00:08:23,729 --> 00:08:27,120
Combination, how much it remembers
and how much it responds.
153
00:08:27,989 --> 00:08:33,419
And so every time you send in another
prompt in what we call chain of thought
154
00:08:33,419 --> 00:08:38,520
prompting, which is what we do in
chat G P T, it has to make a decision.
155
00:08:39,510 --> 00:08:40,679
What am I gonna remember?
156
00:08:41,564 --> 00:08:43,184
And what am I going to forget?
157
00:08:44,625 --> 00:08:47,385
Every time you prompt it
has to make that choice.
158
00:08:49,005 --> 00:08:52,214
And so when you start to say, Hey, it,
it forgot, it's because it's reached
159
00:08:52,214 --> 00:08:54,704
the point where it's got too much
data and it's gotta cut something.
160
00:08:56,655 --> 00:09:01,155
And unfortunately it cut the thing
that you thought was super important.
161
00:09:01,155 --> 00:09:05,055
And so by the end of today, I'm
gonna show you how to get over that.
162
00:09:06,725 --> 00:09:09,795
The very last prompt that I'm gonna
give you today is going to show you
163
00:09:09,795 --> 00:09:15,194
exactly how to, it happened to you today,
Vicky, I'm sorry, but by the end of
164
00:09:15,200 --> 00:09:17,745
today, you'll know how to overcome that.
165
00:09:17,745 --> 00:09:19,995
Does anybody want to know
how to overcome that?
166
00:09:20,235 --> 00:09:23,175
If you do wanna know how to overcome
that, put a heck yes in there for me.
167
00:09:25,245 --> 00:09:28,215
Put a heck yes in the chat if you
wanna know how to overcome it.
168
00:09:28,220 --> 00:09:30,495
Forgetting, yeah, everybody wants that.
169
00:09:30,975 --> 00:09:31,995
Right now.
170
00:09:31,995 --> 00:09:32,925
What about it ignores?
171
00:09:32,930 --> 00:09:34,155
Well, guess what?
172
00:09:34,454 --> 00:09:37,395
The fact that it cuts the majority
of the words that you say,
173
00:09:37,605 --> 00:09:39,255
that's how you know it ignores.
174
00:09:40,965 --> 00:09:48,465
It's designed to ignore, unfortunately,
but it has to in order to, to actually
175
00:09:48,465 --> 00:09:49,845
be able to perform its function.
176
00:09:50,444 --> 00:09:52,725
And we'll talk a little bit about
that and how you can help that,
177
00:09:52,725 --> 00:09:56,145
because there's, there are a few
ways around it, but not that many.
178
00:09:56,204 --> 00:09:57,194
We will talk about that.
179
00:09:58,005 --> 00:10:01,094
Um, what about inconsistent
output structure?
180
00:10:03,464 --> 00:10:05,655
Anybody run into inconsistent
output structure?
181
00:10:05,655 --> 00:10:10,035
Like, you asked something and it gave
you something that was beautiful in
182
00:10:10,035 --> 00:10:11,625
terms of the, the response, right.
183
00:10:11,865 --> 00:10:16,275
Um, sometimes you'll ask, Hey,
give me a, um, an I idea for
184
00:10:16,425 --> 00:10:18,255
I, um, we do video marketing.
185
00:10:18,314 --> 00:10:19,485
That's what my agency does.
186
00:10:20,295 --> 00:10:23,055
And so we may ask, what's it,
give me a video script and it
187
00:10:23,055 --> 00:10:26,925
will give you scene selection
and voiceover and image thoughts.
188
00:10:27,194 --> 00:10:31,814
And then other times it just gives
you a paragraph of pros, right?
189
00:10:31,814 --> 00:10:33,464
We'll talk about that as well.
190
00:10:33,704 --> 00:10:39,314
And then we'll also deal a bit with
the robotic and generic context, right?
191
00:10:39,319 --> 00:10:40,694
Um, as well.
192
00:10:41,834 --> 00:10:45,915
All right, so I got a
prognostication for you.
193
00:10:47,295 --> 00:10:49,844
And all of you fit in
the first line of this.
194
00:10:49,844 --> 00:10:51,765
Smart people are going
to get smarter and dumb.
195
00:10:51,765 --> 00:10:53,925
People are going to get dumber with ai.
196
00:10:54,645 --> 00:10:56,925
And smart people are gonna get
smarter because you're going to
197
00:10:56,930 --> 00:10:58,995
learn to ask better question.
198
00:10:59,204 --> 00:11:00,885
That's all prompts are.
199
00:11:01,425 --> 00:11:05,295
The better we ask questions, is the
smart you'll get and the more you'll
200
00:11:05,295 --> 00:11:08,025
be able to have the system do for you.
201
00:11:08,025 --> 00:11:08,415
No, no.
202
00:11:08,444 --> 00:11:10,935
Michelle, you're one of the
smartest here, so you're fine.
203
00:11:11,385 --> 00:11:11,714
Okay.
204
00:11:12,074 --> 00:11:12,765
Um.
205
00:11:13,485 --> 00:11:15,915
So what the heck is this thing
called prompt engineering?
206
00:11:16,905 --> 00:11:21,105
Well, prompt engineering is the process
of designing and crafting the text.
207
00:11:21,105 --> 00:11:21,525
Yeah, yeah, yeah.
208
00:11:21,555 --> 00:11:21,765
Okay.
209
00:11:21,770 --> 00:11:23,535
That's a whole lot of speak.
210
00:11:23,835 --> 00:11:24,765
It's asking questions.
211
00:11:24,765 --> 00:11:31,515
Y'all Prompt engineering is
the art of asking questions.
212
00:11:33,360 --> 00:11:38,010
Now, it isn't just the, and especially
if, if you're married, it isn't just the
213
00:11:38,010 --> 00:11:42,120
learning how or just asking a question
to your spouse, although technically
214
00:11:42,120 --> 00:11:46,980
it kind of is because Mark, you know,
if you, there are certain ways that
215
00:11:46,980 --> 00:11:49,320
you ask a question to your wife,
you're gonna get a certain response.
216
00:11:50,880 --> 00:11:51,390
True or true.
217
00:11:51,390 --> 00:11:51,450
Yeah.
218
00:11:52,020 --> 00:11:54,420
Very, very, very specific way to ask.
219
00:11:55,290 --> 00:11:55,710
Right?
220
00:11:56,430 --> 00:11:57,740
And so true.
221
00:12:00,885 --> 00:12:03,615
And that's prompt engineering is
understanding and which is why
222
00:12:03,615 --> 00:12:04,844
I started with that first slide.
223
00:12:05,145 --> 00:12:10,035
Once you start to understand that the
computer wants to, it wants to cut out
224
00:12:10,040 --> 00:12:14,204
most of what you say, then you have
to start to, you start to understand,
225
00:12:14,204 --> 00:12:16,035
okay, what's really important to it?
226
00:12:18,495 --> 00:12:20,985
And as you start to understand
that, then you start to get better.
227
00:12:21,435 --> 00:12:25,965
Responses or in, in our case, for
those of us men who are married.
228
00:12:25,965 --> 00:12:28,785
Cause women never have this problem
when, uh, with their husbands,
229
00:12:29,085 --> 00:12:31,395
men who are married, we learn
how to ask questions better.
230
00:12:31,395 --> 00:12:33,675
Women ask perfect questions
every single time.
231
00:12:33,975 --> 00:12:34,485
Right Mark?
232
00:12:39,855 --> 00:12:40,215
All right.
233
00:12:40,575 --> 00:12:41,295
Um.
234
00:12:42,194 --> 00:12:44,204
We're master destruction,
uh, the first thing.
235
00:12:44,204 --> 00:12:48,015
And, and so again, here in this
first set of slides that we're
236
00:12:48,015 --> 00:12:53,175
gonna go through, um, we're going,
yes, we asked Mark, right Mark.
237
00:12:53,535 --> 00:12:56,355
Um, we're gonna go through
some of the basics here.
238
00:12:56,745 --> 00:13:00,585
Now, as I go through these basics, I'm
going to kind of go through some quickly.
239
00:13:00,585 --> 00:13:04,485
If you've got a question, please,
please, please raise your hand,
240
00:13:04,635 --> 00:13:07,395
get my attention, um, explain it.
241
00:13:07,725 --> 00:13:11,730
Um, Actually, we're gonna talk about
explaining it a different way, by the way.
242
00:13:12,030 --> 00:13:13,500
Um, raise your hand.
243
00:13:13,500 --> 00:13:14,520
Get, get my attention.
244
00:13:14,525 --> 00:13:14,700
Mark.
245
00:13:14,700 --> 00:13:16,800
Is it okay if they come
off mute and interrupt me?
246
00:13:17,100 --> 00:13:17,670
I don't mind.
247
00:13:17,810 --> 00:13:18,690
Do, do you mind?
248
00:13:20,840 --> 00:13:21,840
No, no, I can do that.
249
00:13:21,840 --> 00:13:21,990
Fine.
250
00:13:21,990 --> 00:13:22,110
Great.
251
00:13:22,770 --> 00:13:23,790
So y'all have permission.
252
00:13:23,790 --> 00:13:25,920
You can come off mute and ask
me a question if you need to.
253
00:13:25,980 --> 00:13:26,310
Okay.
254
00:13:26,520 --> 00:13:27,510
Um, to get my attention.
255
00:13:27,540 --> 00:13:28,964
So, Number one.
256
00:13:28,964 --> 00:13:34,305
First thing is what is GT's
role in this conversation?
257
00:13:34,454 --> 00:13:38,535
So we are gonna tell it act
as a, or I am a what are you?
258
00:13:38,985 --> 00:13:39,344
Okay.
259
00:13:40,214 --> 00:13:43,995
Act as a YouTube copywriter
that knows video seo.
260
00:13:43,995 --> 00:13:45,615
So here we even took it
a little bit further.
261
00:13:46,005 --> 00:13:48,105
Sometimes people just say,
Hey, act as a copywriter.
262
00:13:48,285 --> 00:13:48,615
No, no, no.
263
00:13:48,620 --> 00:13:49,275
I was specific.
264
00:13:49,635 --> 00:13:52,694
I am a YouTube copywriter and guess what?
265
00:13:52,875 --> 00:13:54,915
This is what I know.
266
00:13:55,800 --> 00:13:56,970
Why do we do that?
267
00:13:57,270 --> 00:14:01,230
We do that because as these models
get bigger, they know and know
268
00:14:01,230 --> 00:14:03,480
and know and know more stuff.
269
00:14:03,960 --> 00:14:09,030
Some of us here are old enough to
remember going to the library, and
270
00:14:09,030 --> 00:14:13,050
when you went to the library and you
walked into the library, you asked
271
00:14:13,055 --> 00:14:17,610
the library and you, you said, Hey,
I'm writing a book report on x.
272
00:14:18,405 --> 00:14:21,824
And the librarian said
that's upstairs to the left.
273
00:14:21,824 --> 00:14:26,324
In this, this section would be
the reference books on that topic.
274
00:14:26,655 --> 00:14:31,214
So what we're telling G P T
here is, Hey, this is who I am.
275
00:14:31,395 --> 00:14:35,535
Go reference that data
that's up there on the left.
276
00:14:35,834 --> 00:14:36,975
Ignore everything else for now.
277
00:14:37,605 --> 00:14:38,954
Go reference that.
278
00:14:39,285 --> 00:14:42,255
That's why act as A is so important.
279
00:14:43,355 --> 00:14:43,645
Okay.
280
00:14:45,885 --> 00:14:47,474
Had a librarian, but they had computer.
281
00:14:47,474 --> 00:14:47,834
Yes.
282
00:14:48,045 --> 00:14:48,615
Yes.
283
00:14:48,704 --> 00:14:49,875
Uh, you, you had a librarian.
284
00:14:49,875 --> 00:14:51,464
My librarians didn't have computers.
285
00:14:51,464 --> 00:14:54,645
They, they just had the um,
um, Dewey decimal system,
286
00:14:54,650 --> 00:14:55,964
which I still don't understand.
287
00:14:56,625 --> 00:15:01,665
Um, number two, who are you speaking
to, especially in marketing?
288
00:15:01,905 --> 00:15:02,925
So we're creating content.
289
00:15:03,015 --> 00:15:05,265
A lot of what we do here is.
290
00:15:05,360 --> 00:15:07,340
Content creation of some sort.
291
00:15:07,610 --> 00:15:08,780
We do video content.
292
00:15:08,840 --> 00:15:10,160
We're always ideating.
293
00:15:10,160 --> 00:15:16,040
Yes, microfiche, we're always
ideating through, uh, video
294
00:15:16,045 --> 00:15:17,330
content for our clients.
295
00:15:18,285 --> 00:15:21,945
Now when we're doing that, we are
sitting here thinking about our
296
00:15:21,945 --> 00:15:23,835
client and their ideal customer.
297
00:15:24,225 --> 00:15:27,675
And so it's important that you understand
the perspective of the content that
298
00:15:27,675 --> 00:15:30,705
you're creating, as we all know,
and, and we had some early sessions,
299
00:15:31,005 --> 00:15:36,285
um, in this James at a, an amazing
session on, on having chat G p t even
300
00:15:36,290 --> 00:15:37,875
help you figure out your customer.
301
00:15:38,175 --> 00:15:39,045
Um, but.
302
00:15:39,690 --> 00:15:43,800
You gotta tell it who your
ideal customer is, right?
303
00:15:43,800 --> 00:15:48,630
My ideal viewer is business owners
who want to hire virtual assistants.
304
00:15:48,750 --> 00:15:49,290
Okay, great.
305
00:15:50,670 --> 00:15:52,110
And then what do you want it to produce?
306
00:15:53,400 --> 00:15:55,080
A blog, a social post, an email.
307
00:15:55,350 --> 00:15:57,720
A video, an ad, a, a video script.
308
00:15:58,110 --> 00:16:02,070
These is the basics which I
think most people, given all that
309
00:16:02,070 --> 00:16:06,510
we've covered so far, most of us
understand most of these already.
310
00:16:07,605 --> 00:16:09,345
Fair point, true or true?
311
00:16:09,405 --> 00:16:10,275
Drop it in the track.
312
00:16:10,515 --> 00:16:11,505
True or true?
313
00:16:12,675 --> 00:16:14,325
Then obviously, what's the topic?
314
00:16:14,685 --> 00:16:15,255
What's the topic?
315
00:16:15,260 --> 00:16:15,945
What's your question?
316
00:16:17,505 --> 00:16:18,615
Okay, what's the question?
317
00:16:18,615 --> 00:16:21,555
So these are all the basics
that you understand so far.
318
00:16:21,975 --> 00:16:26,505
All right, now let's start
talking about desired outputs.
319
00:16:27,555 --> 00:16:33,405
Um, this one here I love because
we, we also do use this for emails.
320
00:16:33,795 --> 00:16:37,395
Um, I don't know about y'all,
but email is one of those things,
321
00:16:37,395 --> 00:16:38,895
and Mark, cover your ears.
322
00:16:39,435 --> 00:16:42,885
Email is one of those things that we
struggle with, like we can ideate through
323
00:16:42,885 --> 00:16:47,594
video content like nothing but man to
come up with what to write in an email.
324
00:16:49,155 --> 00:16:50,084
We're like toast.
325
00:16:51,435 --> 00:16:52,574
Absolute toast.
326
00:16:53,295 --> 00:16:58,425
And so we've used Chat, G P T,
and OpenAI to help us with that.
327
00:16:58,574 --> 00:17:00,885
But now you may write a beautiful
email, but then you still
328
00:17:00,885 --> 00:17:03,074
need a subject line for it.
329
00:17:03,630 --> 00:17:07,260
Right, and so here we would
give it examples, so we have
330
00:17:07,260 --> 00:17:11,069
a prompt that produces subject
lines and we give it examples.
331
00:17:11,460 --> 00:17:12,390
Where did these come from?
332
00:17:12,390 --> 00:17:15,329
Yeah, these are digital marketer
subject lines because, well, why digital
333
00:17:15,329 --> 00:17:17,190
marketers creates great subject lines.
334
00:17:17,190 --> 00:17:22,170
I know because I'm always opening their
emails for products I already owned,
335
00:17:23,280 --> 00:17:24,630
and it's like I already have that.
336
00:17:24,630 --> 00:17:25,530
I don't need to buy it again.
337
00:17:25,650 --> 00:17:30,210
But the subject line was so great, I just
wanted to read it and so you can give it.
338
00:17:30,855 --> 00:17:35,355
Examples of what your
desired output would be.
339
00:17:35,385 --> 00:17:36,555
Now what about structure?
340
00:17:37,305 --> 00:17:38,685
You've gotta dictate your structure.
341
00:17:39,195 --> 00:17:40,365
If you want a list, tell it.
342
00:17:40,395 --> 00:17:41,175
You want a list?
343
00:17:41,415 --> 00:17:42,885
If you want a nested list, tell it.
344
00:17:42,885 --> 00:17:43,845
You want a nest, a list.
345
00:17:44,055 --> 00:17:45,735
If you want to create a table, tell it.
346
00:17:45,735 --> 00:17:51,735
You want to A, okay, one of the
ones, we use a ton for social
347
00:17:51,735 --> 00:17:57,135
posts because nobody wants to sit
and read a two page social post.
348
00:17:57,140 --> 00:17:58,605
That's all in one paragraph.
349
00:17:58,635 --> 00:17:59,565
No, you don't want that.
350
00:18:01,034 --> 00:18:04,215
You want a social post,
especially on LinkedIn usually
351
00:18:04,215 --> 00:18:06,554
are one sentence per paragraph.
352
00:18:07,004 --> 00:18:10,905
Well, if it produced great content
and I still had to go in and hit enter
353
00:18:10,909 --> 00:18:13,845
a bunch of times, okay, I'm lazy.
354
00:18:14,115 --> 00:18:15,225
Don't want to do that either.
355
00:18:16,544 --> 00:18:17,625
So you can tell it.
356
00:18:18,284 --> 00:18:22,875
Each sentence is a paragraph
and it will format that for you.
357
00:18:24,465 --> 00:18:29,055
Going back, um, two slides ago,
if you wanted to give it structure
358
00:18:29,055 --> 00:18:30,915
examples, you can do that as well.
359
00:18:31,305 --> 00:18:34,455
You can say, structure it in
the example of, and give it the
360
00:18:34,455 --> 00:18:38,115
example that of how you want the
structure and it will return that.
361
00:18:38,745 --> 00:18:41,775
Um, anybody have any specific structure
questions before I move on this?
362
00:18:41,775 --> 00:18:44,475
Cause I saw somebody ask
about it a little bit earlier.
363
00:18:45,795 --> 00:18:49,065
Actually, we did have a couple
questions if you wanna do that now.
364
00:18:49,095 --> 00:18:49,125
Okay.
365
00:18:50,025 --> 00:18:50,595
Uh, let's see.
366
00:18:50,600 --> 00:18:53,865
We have nd saying, uh, what
sections changed between your
367
00:18:53,865 --> 00:18:55,455
first talk and this talk?
368
00:18:55,815 --> 00:18:59,145
Maybe just wanna, you know, kind of
go over where kind of you started.
369
00:18:59,145 --> 00:19:01,605
So, So that's a great question, Andy.
370
00:19:01,695 --> 00:19:03,585
Um, and, and here's what changed.
371
00:19:03,795 --> 00:19:09,255
The first eight slides are all the same,
and in terms of when I'm numbering here,
372
00:19:09,705 --> 00:19:13,005
um, then I think when we get to about
number 10, it starts to changing and
373
00:19:13,005 --> 00:19:14,475
then the end of this is very different.
374
00:19:16,275 --> 00:19:16,605
Nice.
375
00:19:17,325 --> 00:19:20,895
Uh, and then if you wanna look at,
uh, can you see the Q and As Tiva?
376
00:19:20,895 --> 00:19:20,955
Yeah.
377
00:19:21,075 --> 00:19:23,235
There's a question by, uh, Saprina Boy.
378
00:19:24,885 --> 00:19:27,165
I use prompt now each
time I start in the chat.
379
00:19:30,360 --> 00:19:31,530
Act as a topic.
380
00:19:31,530 --> 00:19:33,360
You have all the information as a topic.
381
00:19:35,010 --> 00:19:35,340
Okay.
382
00:19:35,760 --> 00:19:38,879
Um, so good.
383
00:19:38,885 --> 00:19:41,760
So you're using this to, to kind
of train on who you are, is that
384
00:19:41,760 --> 00:19:42,629
what you're saying, Sabrina?
385
00:19:47,160 --> 00:19:47,399
Yes.
386
00:19:49,370 --> 00:19:49,590
Yes.
387
00:19:50,340 --> 00:19:53,970
So that's perfectly fine, and I'm
gonna give you a better version
388
00:19:53,970 --> 00:19:56,730
of this in the end as well.
389
00:19:57,855 --> 00:19:58,335
Nice.
390
00:19:59,235 --> 00:19:59,565
Okay.
391
00:20:01,305 --> 00:20:03,254
Um, go back to step four.
392
00:20:04,575 --> 00:20:08,445
Uh, mark, did you have Mark se Selzer?
393
00:20:08,655 --> 00:20:10,004
Did you have a question on step four?
394
00:20:19,335 --> 00:20:20,235
Oh, just didn't catch it.
395
00:20:20,325 --> 00:20:20,925
Okay, sure.
396
00:20:21,135 --> 00:20:21,555
Oh, okay.
397
00:20:23,355 --> 00:20:24,135
You probably keep going.
398
00:20:24,135 --> 00:20:26,085
We'll, we'll touch base
on the questions again.
399
00:20:26,564 --> 00:20:27,195
There it is.
400
00:20:27,284 --> 00:20:29,024
Um, what, what is your topic?
401
00:20:29,415 --> 00:20:29,774
Okay.
402
00:20:29,955 --> 00:20:32,385
Um, Jeff spent a ton of time
on this one, so I'm not, I'm
403
00:20:32,385 --> 00:20:33,465
going to zoom through this.
404
00:20:33,465 --> 00:20:35,895
What is the writing model
that you want to follow?
405
00:20:36,165 --> 00:20:40,425
Um, yeah, that was his whole
presentation, so that was awesome.
406
00:20:40,725 --> 00:20:42,554
Um, that, that Jeff did last week.
407
00:20:42,615 --> 00:20:47,564
And then this is one of my favorites,
topical, um, or additional constraints.
408
00:20:47,745 --> 00:20:50,085
So, and we are gonna go a little
bit further than we did in the M
409
00:20:50,085 --> 00:20:51,375
three presentation on this one.
410
00:20:51,375 --> 00:20:55,574
So topical exclusions have you.
411
00:20:56,820 --> 00:21:00,510
Written a prompt to return content.
412
00:21:00,899 --> 00:21:04,649
And you know that there are certain
contents you don't wanna cover in
413
00:21:04,649 --> 00:21:06,330
this prompt or in this response.
414
00:21:07,230 --> 00:21:10,949
There's certain things you don't want
it to talk about, and sometimes those
415
00:21:10,949 --> 00:21:13,679
are the, the most common answers.
416
00:21:14,340 --> 00:21:15,000
Well, guess what?
417
00:21:15,389 --> 00:21:16,649
You can tell it not to.
418
00:21:18,720 --> 00:21:19,020
Okay?
419
00:21:19,020 --> 00:21:22,169
You can literally type in either
one of these bullet points here
420
00:21:22,320 --> 00:21:24,389
of do not include these topics.
421
00:21:25,125 --> 00:21:28,935
Put 'em in parentheses, right?
422
00:21:28,935 --> 00:21:32,685
Because when you put 'em in parentheses,
it lets the, let's the system know that
423
00:21:32,685 --> 00:21:39,165
it's not two separate words, but it
is one word, and comma, separate them.
424
00:21:42,345 --> 00:21:42,735
Okay?
425
00:21:42,975 --> 00:21:47,175
This will help you, especially
when you're dealing with the
426
00:21:48,284 --> 00:21:51,764
issue of, well, my topic's pretty.
427
00:21:52,770 --> 00:21:57,330
Broad or pretty popular, and people say
the same things about it, which leads to
428
00:21:57,540 --> 00:22:03,480
very generic content, topical exclusions
and saying, Hey, I know this is broad,
429
00:22:03,480 --> 00:22:05,040
but don't talk about this, this, or that.
430
00:22:05,970 --> 00:22:12,210
Forces the system to ideate further and
give you new topics, which will take
431
00:22:12,210 --> 00:22:16,140
you out of being just, um, generic.
432
00:22:17,010 --> 00:22:17,400
Okay.
433
00:22:17,970 --> 00:22:21,090
Um, next is, uh, phrase exclusions.
434
00:22:21,719 --> 00:22:28,169
This drives me nuts and I may be the
only one, but please write a blog
435
00:22:28,169 --> 00:22:32,610
post in this blog in this video.
436
00:22:32,610 --> 00:22:34,199
Drives me nuts.
437
00:22:34,320 --> 00:22:37,530
So just tell it not to, if it's doing
something like that to you where
438
00:22:37,590 --> 00:22:41,879
there are certain phrases that you
don't want it to use, you just like
439
00:22:41,885 --> 00:22:44,280
the topical exclusions, you can say.
440
00:22:45,585 --> 00:22:47,535
Don't use this phrase.
441
00:22:47,535 --> 00:22:50,025
Now, what I'll tell you about
phrase, phrase exclusions
442
00:22:50,025 --> 00:22:51,975
is it doesn't always work.
443
00:22:52,725 --> 00:22:56,745
It works, my opinion,
70 to 80% of the time.
444
00:22:58,335 --> 00:22:58,755
Okay?
445
00:22:59,145 --> 00:23:02,985
Um, especially the, in this video, if it,
if you're writing a video script and it's
446
00:23:02,985 --> 00:23:07,905
a long video script, it may not start the
video within this video, but paragraph
447
00:23:07,905 --> 00:23:09,555
number five here comes that in this video.
448
00:23:12,300 --> 00:23:12,659
Okay.
449
00:23:12,960 --> 00:23:15,480
Um, so phrase exclusions,
tonal suggestions.
450
00:23:15,480 --> 00:23:18,690
Jeff, um, talked about about
this, uh, before, and I know
451
00:23:18,690 --> 00:23:20,010
some other people have as well.
452
00:23:20,220 --> 00:23:26,310
Um, that you can get into great tonal
suggestions, uh, with the, the chatbots.
453
00:23:26,490 --> 00:23:31,020
Now, this is one of my favorites as well.
454
00:23:31,620 --> 00:23:33,990
Let's think about.
455
00:23:36,330 --> 00:23:40,110
And you can take each one of
these exactly as they are.
456
00:23:40,680 --> 00:23:41,670
You don't need to change 'em.
457
00:23:41,670 --> 00:23:42,660
This is how we use them.
458
00:23:43,740 --> 00:23:45,180
Okay, so let's say you have a passage.
459
00:23:46,425 --> 00:23:49,514
Let's think about this
passage using an example.
460
00:23:49,635 --> 00:23:50,895
So sometimes we want it to rewrite.
461
00:23:50,895 --> 00:23:53,385
Well, maybe a rewrite
isn't just what you want.
462
00:23:53,715 --> 00:23:56,774
Maybe you want something that's
completely different and or, or
463
00:23:56,774 --> 00:23:58,365
a rephrase isn't what you want.
464
00:23:58,574 --> 00:24:00,135
You want something that's
completely different at a different
465
00:24:00,135 --> 00:24:02,625
perspective about the same thing.
466
00:24:02,774 --> 00:24:03,254
Okay, great.
467
00:24:03,915 --> 00:24:07,754
Let's think about this
passage using an example.
468
00:24:09,795 --> 00:24:17,445
And it will come back and create an
example slash story for your content.
469
00:24:19,605 --> 00:24:24,645
Let's think about this topic
from a reverse perspective.
470
00:24:25,665 --> 00:24:31,395
Again, the thing that we do at our agency
is a ton of ideation, and when you have
471
00:24:31,395 --> 00:24:34,935
a bunch of clients in your ideating
form, it's so easy to get pigeonholed.
472
00:24:34,935 --> 00:24:38,835
And even in your own brand, it's so easy
to get pigeonholed and only see things
473
00:24:39,015 --> 00:24:44,445
from one perspective because that's the
way you've looked at it forever, right?
474
00:24:44,450 --> 00:24:48,225
You did your I C P, you did all
your doctors and, and you only
475
00:24:48,225 --> 00:24:49,815
think about the world in this way.
476
00:24:51,015 --> 00:24:55,845
These give you the opportunity to see
things from a completely different
477
00:24:55,845 --> 00:25:01,485
perspective and, and sometimes
realize things in ideation that you
478
00:25:01,485 --> 00:25:04,275
would've never considered before.
479
00:25:05,055 --> 00:25:05,385
Okay.
480
00:25:05,685 --> 00:25:09,495
Um, another one is, is think
about this from a bigger context.
481
00:25:09,615 --> 00:25:10,305
So
482
00:25:14,565 --> 00:25:16,155
this helps us with clients.
483
00:25:16,245 --> 00:25:19,425
Have you ever talked to a
client and, and Oh my gosh.
484
00:25:20,625 --> 00:25:21,915
I'm thinking of one right now.
485
00:25:22,935 --> 00:25:26,925
Have you ever talked to a client and
you're trying to understand what they do
486
00:25:26,925 --> 00:25:29,895
because you're, you're trying to create
copy for them or a video, what, what have
487
00:25:29,895 --> 00:25:35,115
you, you you're trying to create content
for them and all they can do is give you
488
00:25:35,115 --> 00:25:38,535
the details of how they make their widget.
489
00:25:39,075 --> 00:25:45,254
Like they're so in depth in what they
do that they can't actually see the
490
00:25:45,705 --> 00:25:47,415
theoretical side of what they do.
491
00:25:49,889 --> 00:25:55,980
Right, and they can't picture bigger
because they're so in the weeds.
492
00:25:56,310 --> 00:25:58,350
This helps you get out of the weeds.
493
00:25:58,919 --> 00:26:02,340
You can take their weeds, stick
it in, and it will take you up.
494
00:26:03,870 --> 00:26:04,950
Didn't go up high enough.
495
00:26:05,040 --> 00:26:05,550
Great.
496
00:26:05,909 --> 00:26:06,690
Do it again.
497
00:26:08,639 --> 00:26:12,780
Run it again on whatever it
produced, and just keep going
498
00:26:12,780 --> 00:26:15,720
until you get that bigger concept.
499
00:26:16,710 --> 00:26:18,060
Okay, why?
500
00:26:18,060 --> 00:26:19,110
Where is that helpful?
501
00:26:19,320 --> 00:26:24,000
Well, guess what those bigger concepts
are usually what you use in the awareness
502
00:26:24,000 --> 00:26:29,820
stage of the CV J four content in the
awareness stage of the CV J for content.
503
00:26:29,825 --> 00:26:32,940
You're not deep diving
in how this thing works.
504
00:26:34,230 --> 00:26:37,200
It's big theory of why they need it.
505
00:26:38,340 --> 00:26:43,409
This prompt helps us a ton with that,
or I should say this prompt fragment.
506
00:26:43,740 --> 00:26:45,570
Because you're taking this,
you're taking your regular
507
00:26:45,570 --> 00:26:48,090
prompt and you're adding this on.
508
00:26:50,490 --> 00:26:50,850
Okay?
509
00:26:52,140 --> 00:26:55,980
Um, let's think about this
prompt using analogies.
510
00:26:56,610 --> 00:26:59,670
Let's think about this
prompt from multiple angles.
511
00:27:02,400 --> 00:27:03,660
Got a few more of these.
512
00:27:03,660 --> 00:27:06,090
Let's think about how does, let's
think about, so these, let's
513
00:27:06,090 --> 00:27:07,080
think about working for y'all.
514
00:27:07,080 --> 00:27:08,490
Y'all, y'all feeling with these?
515
00:27:09,120 --> 00:27:12,030
I if, if these are, are doing
something for you, say, writing the,
516
00:27:12,030 --> 00:27:13,260
the chat, doing something for me.
517
00:27:15,945 --> 00:27:16,754
Thank you, Kathy.
518
00:27:18,705 --> 00:27:25,875
Um, let's think about, about this topic
by looking at its impact on, so this
519
00:27:25,875 --> 00:27:29,115
is another great one right now, right?
520
00:27:29,415 --> 00:27:36,375
Um, I have a client who, he works with
kids and helping kids through transitions,
521
00:27:36,675 --> 00:27:39,595
but he sells to who, he doesn't sell
to kids, he sells to the parents.
522
00:27:40,740 --> 00:27:43,889
And all he can do is tell me about the
transformation that he gives to kids.
523
00:27:44,340 --> 00:27:47,699
And he's kind of doesn't
understand the parent side.
524
00:27:48,570 --> 00:27:52,649
And so we're able to use this to
say, Hey, think about how these
525
00:27:52,649 --> 00:27:56,010
sorts of transformation for kids
will have, what kind of impact
526
00:27:56,010 --> 00:27:57,629
that will have on the parents.
527
00:28:00,990 --> 00:28:01,939
Helps us ideate.
528
00:28:03,629 --> 00:28:06,780
Let's think about this topic
from a historical standpoint.
529
00:28:06,929 --> 00:28:08,850
This works great if you're
trying to tell a story.
530
00:28:08,850 --> 00:28:09,270
Sometimes.
531
00:28:09,270 --> 00:28:14,189
Sometimes this topic has existed in
a different way in history and chat.
532
00:28:14,195 --> 00:28:18,929
G P T pulls that out and all of a sudden
I've got this story I can relate to here.
533
00:28:21,149 --> 00:28:21,480
Okay?
534
00:28:21,780 --> 00:28:24,510
Um, in terms of economics
or the negative perspective.
535
00:28:24,510 --> 00:28:26,760
Now someone's gonna ask
negative versus reverse.
536
00:28:27,060 --> 00:28:29,189
Those are not the same.
537
00:28:33,149 --> 00:28:34,169
Those are not the same.
538
00:28:34,169 --> 00:28:36,960
So on the first slide here, it
said, think about 'em from a
539
00:28:36,960 --> 00:28:41,280
reverse perspective, reverse and
negative produce different results.
540
00:28:42,990 --> 00:28:43,290
Okay?
541
00:28:45,330 --> 00:28:45,629
All right.
542
00:28:46,020 --> 00:28:47,010
Um, any questions?
543
00:28:47,040 --> 00:28:48,149
Let, let's jump in here.
544
00:28:48,330 --> 00:28:51,030
Do a lot of long form content,
specifically help with clients.
545
00:28:52,785 --> 00:28:53,175
Okay.
546
00:28:53,235 --> 00:28:56,865
Yeah, if you wanna read the, the whole
question or I can if you want Lati.
547
00:28:57,105 --> 00:28:57,285
Yeah.
548
00:28:57,435 --> 00:28:57,825
Oh, sorry.
549
00:28:57,825 --> 00:28:58,095
Sorry.
550
00:28:58,095 --> 00:29:01,755
I do a lot of long form content,
specifically helping clients write
551
00:29:01,755 --> 00:29:04,905
12,000 word non-fiction books.
552
00:29:05,175 --> 00:29:06,555
GT four is a godan.
553
00:29:06,560 --> 00:29:06,795
Yes.
554
00:29:06,795 --> 00:29:10,635
It would be GT three and 3.5
weren't very helpful in that sense.
555
00:29:10,695 --> 00:29:15,615
Um, any thoughts on how much even better
it's going to be once they release pl?
556
00:29:15,615 --> 00:29:16,185
Oh my gosh.
557
00:29:17,504 --> 00:29:20,595
Um, so memory's not gonna change
once they re release plugins.
558
00:29:20,895 --> 00:29:22,305
Plugins aren't gonna help you with memory.
559
00:29:22,725 --> 00:29:25,605
Um, that, that was, sorry, that
was the end of of his question.
560
00:29:25,875 --> 00:29:29,475
How, um, things are gonna be
different once they release plugins,
561
00:29:29,475 --> 00:29:30,855
especially in terms of memory.
562
00:29:31,785 --> 00:29:33,345
Nothing's gonna change
with memory and plugins.
563
00:29:33,375 --> 00:29:39,285
However, the thing that plugins are
going to allow us to do is to, again,
564
00:29:39,285 --> 00:29:43,275
getting back to that slide where I say
smart people are going to get smarter.
565
00:29:45,450 --> 00:29:53,370
Your ability to be able to consider now
the content that was created and how I
566
00:29:53,370 --> 00:30:00,210
can then marry it with what this plugin
can do is going to allow you to create
567
00:30:00,240 --> 00:30:02,070
even better and bigger and faster.
568
00:30:03,149 --> 00:30:05,250
Paul, I hope that answered,
answered your question there.
569
00:30:05,550 --> 00:30:08,399
Uh, my friend, um, writing a.
570
00:30:09,105 --> 00:30:11,415
You know, I was trying to remember
Bucket do io earlier today, and I
571
00:30:11,415 --> 00:30:12,975
could not, I just came off of mute.
572
00:30:13,275 --> 00:30:15,225
Is it OK if I ask if I
ask a follow up there?
573
00:30:15,435 --> 00:30:16,365
Yeah, yeah, go ahead buddy.
574
00:30:16,815 --> 00:30:17,445
This is Paul.
575
00:30:17,685 --> 00:30:19,395
Thank you for asking me
for answering the question.
576
00:30:20,085 --> 00:30:26,295
Um, do you, do you see, or are you
aware of any, um, improvements, whether
577
00:30:26,295 --> 00:30:30,465
it's within GT four or, um, and maybe
no one knows this yet, but in terms
578
00:30:30,465 --> 00:30:33,885
of that memory and being able to go
back further, more than 4,000 words?
579
00:30:34,910 --> 00:30:36,540
Oh, oh, oh, I understand your question.
580
00:30:36,540 --> 00:30:36,690
Now.
581
00:30:36,690 --> 00:30:37,920
I misunderstood your question.
582
00:30:37,980 --> 00:30:40,530
Um, however, the answer's the same.
583
00:30:41,160 --> 00:30:46,770
Um, so it's gonna to, it's going
to be interesting here with G PT
584
00:30:46,770 --> 00:30:51,060
four because G p T four, everyone's
like, oh my gosh, 32,000 words.
585
00:30:51,065 --> 00:30:54,330
Now, you know, all, all this wonderful
stuff and, and that seems wonderful.
586
00:30:54,470 --> 00:30:59,880
But if you actually read the fine print,
the prompt side of this is still at 4,000.
587
00:31:00,735 --> 00:31:05,745
Now that's greater than it was
in G P T three or 3.5, where the
588
00:31:06,105 --> 00:31:08,205
prompt and response was 4,000.
589
00:31:08,925 --> 00:31:12,675
So it's gonna be able to remember
more prompts in that sense, but
590
00:31:12,675 --> 00:31:14,145
it's still going to have a limit.
591
00:31:15,015 --> 00:31:15,615
That make sense?
592
00:31:15,975 --> 00:31:16,815
It, it does.
593
00:31:16,820 --> 00:31:19,635
And, and, and for example,
I've, I've, I've really enjoyed
594
00:31:19,635 --> 00:31:20,875
this training so far, and I.
595
00:31:21,195 --> 00:31:24,045
Took from, I forget the person's name,
but he showed us how to create the,
596
00:31:24,675 --> 00:31:28,035
um, ideal client profile, including
demographics and psychographics.
597
00:31:29,475 --> 00:31:32,895
And so now when I'm writing a
book, I'll paste that whole thing
598
00:31:32,895 --> 00:31:36,075
in and, you know, who knows, maybe
it's 1,500 to a thousand words.
599
00:31:36,495 --> 00:31:40,905
Um, do, is, is that the best approach
or do you think that that should
600
00:31:40,905 --> 00:31:43,215
be simplified into something more?
601
00:31:43,935 --> 00:31:44,745
Concrete.
602
00:31:45,135 --> 00:31:46,455
In other words, can't, is it?
603
00:31:46,455 --> 00:31:50,145
Is it really if I say, Hey, here's my
deal, client profile, write a paragraph.
604
00:31:50,235 --> 00:31:54,105
Considering this person in mind, is
that too much information for it?
605
00:31:54,975 --> 00:31:58,455
So I'm going to answer that
question with the very last slide.
606
00:31:58,965 --> 00:31:59,655
Okay, perfect.
607
00:31:59,655 --> 00:32:00,015
Thank you.
608
00:32:00,555 --> 00:32:00,825
Cool.
609
00:32:02,205 --> 00:32:02,955
Hold me to that too.
610
00:32:03,750 --> 00:32:04,110
Okay.
611
00:32:05,100 --> 00:32:10,470
Um, man, writing a marketing quiz
for bucket io need a question
612
00:32:10,470 --> 00:32:13,080
for a quiz with four responses.
613
00:32:13,139 --> 00:32:17,639
Each response should be tied to
one of four specific categories,
614
00:32:17,639 --> 00:32:21,360
no duplicate response categories,
but all four categories should
615
00:32:21,360 --> 00:32:23,670
be included in the question.
616
00:32:23,670 --> 00:32:25,350
So you want a prompt for that?
617
00:32:26,070 --> 00:32:27,330
Is that that's what you're saying, man.
618
00:32:37,590 --> 00:32:43,680
Do we have Nan, you do, I, I
thought I just put yes in the,
619
00:32:43,680 --> 00:32:45,090
uh, in the response there.
620
00:32:45,095 --> 00:32:45,900
So I thought you saw that.
621
00:32:46,020 --> 00:32:46,260
Sorry.
622
00:32:46,265 --> 00:32:46,860
Uh, no I didn't.
623
00:32:46,860 --> 00:32:47,370
I'm sorry.
624
00:32:47,370 --> 00:32:47,940
I'm so sorry.
625
00:32:48,240 --> 00:32:49,260
Um, okay.
626
00:32:49,470 --> 00:32:51,180
We can talk about that one at the end.
627
00:32:51,210 --> 00:32:55,410
Hold, let's hold that cuz that's gonna,
um, we have a prompt that does something
628
00:32:55,830 --> 00:32:58,080
similar to this, but not exactly.
629
00:32:58,965 --> 00:33:01,545
Um, to hold that, let, let
me, let's talk about that.
630
00:33:01,545 --> 00:33:03,015
Cause that, that's gonna
take a little work.
631
00:33:03,375 --> 00:33:03,555
Okay.
632
00:33:03,555 --> 00:33:04,105
Which is fine.
633
00:33:04,155 --> 00:33:04,845
Which is fine.
634
00:33:05,265 --> 00:33:05,565
Okay.
635
00:33:05,565 --> 00:33:09,735
Besides commas, how do other
elements like parentheses, brackets,
636
00:33:10,335 --> 00:33:12,405
and impact, input and output?
637
00:33:12,465 --> 00:33:14,985
It depends on how you, what you tell us.
638
00:33:14,985 --> 00:33:21,105
So, and d ask, besides commas, how
do other elements, like parentheses,
639
00:33:21,435 --> 00:33:28,215
brackets, et cetera, impact or, um,
the input slash output, um, Depends
640
00:33:28,215 --> 00:33:30,824
on what you tell it with chat g p t.
641
00:33:31,125 --> 00:33:35,385
Um, y'all gonna get tired of this answer,
but the last slide is also going to answer
642
00:33:35,385 --> 00:33:40,725
that for you there as well, um, because
you'll see how I'm using them in the last
643
00:33:40,725 --> 00:33:46,395
slide and how I tell it in the last slide
that, uh, a curly bracket means this
644
00:33:46,395 --> 00:33:49,004
versus, um, a square bracket means that.
645
00:33:49,935 --> 00:33:52,274
So you can, you can absolutely tell it.
646
00:33:52,695 --> 00:33:56,804
Um, so that's my answer for the moment.
647
00:33:58,080 --> 00:34:00,030
Uh, also this is Paul.
648
00:34:00,030 --> 00:34:04,710
Now also I see that G PT
four just released templates.
649
00:34:05,370 --> 00:34:10,530
Uh, that seems like it's going to
completely undercut companies like Jasper.
650
00:34:10,530 --> 00:34:13,290
Any thoughts on the direction of open ai?
651
00:34:14,130 --> 00:34:18,270
Open AI seems amazing to me, but
would love your thoughts on it.
652
00:34:18,630 --> 00:34:22,470
Um, no.
653
00:34:23,549 --> 00:34:31,650
It's not going to undercut Jasper,
um, would be my short answer to you.
654
00:34:32,909 --> 00:34:37,799
My longer answer would be I haven't seen
the templates, so that's my longer answer.
655
00:34:37,799 --> 00:34:38,340
Go ahead, mark.
656
00:34:39,105 --> 00:34:40,455
Oh, I was gonna say y Yeah.
657
00:34:40,514 --> 00:34:42,915
Uh, Jasper really focuses
on small business.
658
00:34:42,975 --> 00:34:46,395
So in terms of how it generates
answers, it's trying to generate
659
00:34:46,395 --> 00:34:51,884
answers for business use, which is a
big differentiator from, um, chat G P T.
660
00:34:51,884 --> 00:34:54,884
But the whole system itself,
again, is a research platform.
661
00:34:54,889 --> 00:34:59,595
So my theory is that open AI's entire
business model revolves around licensing
662
00:34:59,595 --> 00:35:02,295
agreements with outside businesses.
663
00:35:02,384 --> 00:35:04,185
So they're not planning
on making money with.
664
00:35:04,365 --> 00:35:06,044
Consumers using the product.
665
00:35:06,044 --> 00:35:08,475
They really just wanna
get those APIs out there.
666
00:35:08,774 --> 00:35:12,825
Cuz pretty soon, once everybody has an
app that's driven by OpenAI, they're gonna
667
00:35:12,825 --> 00:35:14,475
start charging those companies for it.
668
00:35:14,475 --> 00:35:15,615
So that's my theory.
669
00:35:16,214 --> 00:35:16,754
Well, guess what?
670
00:35:16,759 --> 00:35:17,714
They're already charging us.
671
00:35:18,104 --> 00:35:18,524
Yes.
672
00:35:20,274 --> 00:35:21,915
I mean, I pay the bill monthly.
673
00:35:22,125 --> 00:35:23,084
They're already charging us.
674
00:35:24,435 --> 00:35:26,475
Um, yeah.
675
00:35:26,504 --> 00:35:30,015
And, and it's going to be to,
to Mark's point, you know, a
676
00:35:30,015 --> 00:35:31,995
company like Jasper uses open ai.
677
00:35:32,654 --> 00:35:33,075
Right.
678
00:35:33,285 --> 00:35:37,064
And so you, you, you're not gonna really
piss off your customer base in that
679
00:35:37,064 --> 00:35:39,825
sense and compete directly with them.
680
00:35:40,544 --> 00:35:40,814
Right?
681
00:35:40,814 --> 00:35:44,654
Like digital marketer is not
gonna start doing marketing for
682
00:35:45,254 --> 00:35:47,205
agent or for small businesses.
683
00:35:47,325 --> 00:35:48,314
They teach us how to do it.
684
00:35:48,314 --> 00:35:49,334
That's what their job is.
685
00:35:49,694 --> 00:35:50,205
That's what they do.
686
00:35:51,254 --> 00:35:51,495
Right.
687
00:35:52,694 --> 00:35:54,584
Um, all right.
688
00:35:54,584 --> 00:35:59,685
Stephanie, at what point in the prompt do
you use, let's think about terminology.
689
00:35:59,685 --> 00:36:00,674
That's, that's a great question.
690
00:36:00,765 --> 00:36:08,174
Um, Because there, let me, let
me think about how I answer this.
691
00:36:08,174 --> 00:36:12,285
Well, there is no direct
order to the prompts.
692
00:36:12,525 --> 00:36:16,634
If we go back to that third slide
where the first thing it's gonna do
693
00:36:16,634 --> 00:36:22,694
is figure out what words to to, to
cut and then reorder what you say it,
694
00:36:23,475 --> 00:36:25,455
what it's reordering, what you say.
695
00:36:27,435 --> 00:36:32,415
Now, logically, It may make sense
to put it in a certain way so that
696
00:36:32,415 --> 00:36:36,104
it's readable to you as a human,
but the computer doesn't care.
697
00:36:38,354 --> 00:36:38,504
Okay.
698
00:36:38,509 --> 00:36:44,714
Stephanie, uh, Kathleen, would you
please clarify your meaning with regard
699
00:36:44,720 --> 00:36:48,464
to the angle of reverse perspective?
700
00:36:49,214 --> 00:36:53,714
Um, that's a good question, Kathy.
701
00:36:55,424 --> 00:36:56,145
So.
702
00:36:59,985 --> 00:37:02,415
Um, what I'm trying to do is
think of a good example for you.
703
00:37:02,715 --> 00:37:04,605
Negative means negative, right?
704
00:37:04,665 --> 00:37:06,075
So, um,
705
00:37:09,945 --> 00:37:10,245
here
706
00:37:14,535 --> 00:37:17,025
is the Samsung watch.
707
00:37:17,235 --> 00:37:17,895
Good?
708
00:37:18,105 --> 00:37:18,465
No.
709
00:37:18,555 --> 00:37:18,945
Okay.
710
00:37:19,035 --> 00:37:20,385
Let me go a little bit further.
711
00:37:20,955 --> 00:37:27,345
Studies have proven that the Samsung watch
is good for children under the age of six.
712
00:37:29,700 --> 00:37:30,450
Is that true?
713
00:37:30,600 --> 00:37:31,590
Let's say that was your prompt.
714
00:37:33,000 --> 00:37:35,040
Think about this from
a negative perspective.
715
00:37:36,450 --> 00:37:37,950
What type, this is the response.
716
00:37:38,130 --> 00:37:42,630
What type of idiot would give a
Samsung watch to a six-year-old
717
00:37:42,634 --> 00:37:45,870
child That's negative.
718
00:37:46,920 --> 00:37:53,790
Reverse would be where studies have
shown that children under the age of six,
719
00:37:54,360 --> 00:37:57,660
um, Would benefit from a Samsung watch.
720
00:37:57,900 --> 00:38:04,320
The concern exists of what the long term
effects of that would be on the child.
721
00:38:06,510 --> 00:38:07,050
Does that make sense?
722
00:38:07,050 --> 00:38:07,470
Kathy,
723
00:38:10,980 --> 00:38:11,820
as an example, I.
724
00:38:17,595 --> 00:38:18,285
Thank you.
725
00:38:18,435 --> 00:38:18,705
All right.
726
00:38:18,735 --> 00:38:19,095
Awesome.
727
00:38:20,355 --> 00:38:26,085
Okay, Michelle, uh, can you provide
suggestions on how to create calculators?
728
00:38:26,115 --> 00:38:31,695
Ooh, that was a great example
of a calculator, uh, by dm, but
729
00:38:31,695 --> 00:38:33,195
want to learn how to create them.
730
00:38:35,295 --> 00:38:37,065
I'll be honest, I've never
created a calculator.
731
00:38:40,590 --> 00:38:41,190
What type of cal?
732
00:38:41,430 --> 00:38:44,910
Um, do me a favor, put it in the
chat, um, in the regular Zoom chat.
733
00:38:44,940 --> 00:38:45,810
Michelle, what type?
734
00:38:46,350 --> 00:38:46,920
I love this.
735
00:38:46,920 --> 00:38:52,530
Like gimme something I've not done
and make me think, um, yeah, just so
736
00:38:52,530 --> 00:38:57,000
everybody knows, Dean Davis, he actually,
he covers that pretty extensively and
737
00:38:57,000 --> 00:39:04,610
his, and that'll be Thursday don't, or
Wednesday, but let me, let me confirm.
738
00:39:06,075 --> 00:39:09,555
So lemme actually get into more app
development is, is how I looked at it.
739
00:39:09,975 --> 00:39:10,425
Yeah, yeah.
740
00:39:10,430 --> 00:39:11,175
No, absolutely.
741
00:39:11,235 --> 00:39:14,715
So let me also say this one of, and
this is one of the things that I, I, I
742
00:39:14,720 --> 00:39:22,365
appreciate with the, um, doing, um, this,
uh, AI bootcamp and more so creating the,
743
00:39:22,635 --> 00:39:25,365
the, um, the, the Slack channel as well.
744
00:39:26,340 --> 00:39:29,400
AI is collaborative, you
know, for human history.
745
00:39:29,400 --> 00:39:32,070
We've learned stuff and we've kind of
siloed and, and I learned something,
746
00:39:32,070 --> 00:39:34,320
but Mark didn't know it, and I'm not
sure I'm gonna share, share with Mark.
747
00:39:34,530 --> 00:39:41,100
But the point of ai, the, the, the, the
genesis of this all is a collaborative
748
00:39:41,100 --> 00:39:45,480
experience of all of us working together,
learning from each other of I learned
749
00:39:45,480 --> 00:39:46,320
something, I shared it with you.
750
00:39:46,320 --> 00:39:47,120
You learned something you shared.
751
00:39:47,130 --> 00:39:51,090
And we go back and forth in on
this process that there isn't now
752
00:39:51,090 --> 00:39:53,310
just this one guru that knows.
753
00:39:53,665 --> 00:39:56,785
Every freaking thing, it
isn't possible anymore.
754
00:39:57,415 --> 00:39:57,805
Right.
755
00:39:57,925 --> 00:40:01,015
That's one of the things I love about
how, of what you've done here, mark
756
00:40:01,165 --> 00:40:05,335
pulling all of us together because yeah,
I've learned a ton from everybody else
757
00:40:05,340 --> 00:40:09,025
and I can't wait for that one cuz I've
never thought about calculators, so yeah.
758
00:40:09,025 --> 00:40:09,865
I wanna learn that too.
759
00:40:11,995 --> 00:40:12,295
Okay.
760
00:40:12,415 --> 00:40:13,435
We all learned together.
761
00:40:15,145 --> 00:40:19,465
All right, moving on, let's.
762
00:40:21,165 --> 00:40:24,375
Variables, and these aren't
true, true variables.
763
00:40:25,155 --> 00:40:27,735
These are more placeholders
than variables.
764
00:40:28,095 --> 00:40:32,174
And these are designed to
one, help you with some of the
765
00:40:32,234 --> 00:40:34,065
input concerns that we have.
766
00:40:34,575 --> 00:40:34,845
Okay.
767
00:40:34,845 --> 00:40:37,004
Especially in, in three, 3.5.
768
00:40:37,004 --> 00:40:40,815
And if heck, if you're writing a book
in four and you're having some input
769
00:40:40,819 --> 00:40:42,705
concerns, could help you there too.
770
00:40:43,154 --> 00:40:48,045
The purpose of the variables
slash placeholders is to
771
00:40:48,045 --> 00:40:49,754
cut down on the amount of.
772
00:40:50,880 --> 00:40:53,850
Text that you're sending
in with the prompt.
773
00:40:54,450 --> 00:41:02,160
Okay, so if you're going to use the same
term multiple times, then a placeholder
774
00:41:02,165 --> 00:41:04,590
slash variable is the way to do it.
775
00:41:05,310 --> 00:41:10,980
Likewise, being an old school programmer,
it just makes life more readable.
776
00:41:10,980 --> 00:41:13,020
It makes things more readable for me.
777
00:41:13,635 --> 00:41:13,785
Okay.
778
00:41:13,785 --> 00:41:17,025
So that's also secondary
reason why I use them.
779
00:41:17,415 --> 00:41:20,745
Now, the way this works and, and
OpenAI is actually pretty smart
780
00:41:20,745 --> 00:41:23,535
with this cuz I screw it up all the
time and it still gets it right.
781
00:41:24,015 --> 00:41:24,375
Okay.
782
00:41:25,095 --> 00:41:29,955
So you can see here I've got ICP
right on a second line, right?
783
00:41:29,955 --> 00:41:30,405
Right.
784
00:41:30,735 --> 00:41:37,125
Uh, YouTube video description to attract
I c P and then down below I have I CCP
785
00:41:37,185 --> 00:41:41,145
equals and I define who the I C P is.
786
00:41:41,685 --> 00:41:46,065
Now, um, I've tried this both ways of
putting it in parentheses cuz you'll say,
787
00:41:46,065 --> 00:41:47,775
well, shouldn't that be in parentheses?
788
00:41:47,775 --> 00:41:51,075
Because we, we wanted to consider that
whole thing of business owners who
789
00:41:51,075 --> 00:41:53,685
want to hire virtual assistants as one.
790
00:41:55,335 --> 00:42:00,705
In our experience so far, using
the variable slash placeholders,
791
00:42:00,915 --> 00:42:04,335
it treats it all as one and it
doesn't return different results.
792
00:42:04,485 --> 00:42:06,555
At least not different enough
results where it was like, oh, you
793
00:42:06,555 --> 00:42:07,755
completely ignored what I said.
794
00:42:08,265 --> 00:42:08,625
Okay.
795
00:42:09,045 --> 00:42:10,424
Um, so.
796
00:42:10,980 --> 00:42:15,810
Variable slash placeholders help you
with readability, but then also saving
797
00:42:15,810 --> 00:42:20,610
characters, which is the main reason,
saving characters on your input prompt.
798
00:42:21,750 --> 00:42:22,050
Okay.
799
00:42:22,680 --> 00:42:23,640
That make sense to everybody?
800
00:42:24,390 --> 00:42:25,740
Any questions on that before I move on?
801
00:42:28,260 --> 00:42:30,420
Carrie, can you just
repeat that one more time?
802
00:42:31,920 --> 00:42:32,340
Sure.
803
00:42:33,090 --> 00:42:35,160
So variables and, and placeholders.
804
00:42:35,730 --> 00:42:35,970
Okay.
805
00:42:35,970 --> 00:42:36,540
Same thing.
806
00:42:37,080 --> 00:42:42,750
They help you to cut down on the
number of characters that your
807
00:42:42,750 --> 00:42:45,060
prompt is sending into the system.
808
00:42:45,060 --> 00:42:48,509
So it saves characters that
are, they're sending in.
809
00:42:48,509 --> 00:42:54,029
Cuz again, that's wh when we start talking
about how much can it taken and return.
810
00:42:54,390 --> 00:42:58,410
Every time I repeat I c p, all of
those words, it counts multiple times.
811
00:42:59,100 --> 00:42:59,910
Here it doesn't.
812
00:42:59,910 --> 00:43:03,509
If I have to repeat i c p
multiple times in my prompt.
813
00:43:04,920 --> 00:43:05,640
That make sense?
814
00:43:10,680 --> 00:43:11,070
Yes.
815
00:43:11,070 --> 00:43:11,640
Thank you.
816
00:43:12,150 --> 00:43:13,140
You're welcome.
817
00:43:15,030 --> 00:43:15,450
Okay.
818
00:43:15,660 --> 00:43:16,710
What about hallucinating?
819
00:43:20,490 --> 00:43:21,930
Anybody have a hallucinate on you?
820
00:43:22,410 --> 00:43:24,210
Anybody wanted to stop
hallucinating on you?
821
00:43:26,070 --> 00:43:27,780
This prompt will make
it stop hallucinating.
822
00:43:32,595 --> 00:43:35,775
Now it's gonna be in the slide, so
you don't have to go copy or try to
823
00:43:35,775 --> 00:43:37,904
write all of this down right now.
824
00:43:38,265 --> 00:43:41,445
Um, but here it is, where I've
actually only taken the, the
825
00:43:41,475 --> 00:43:43,425
very last three lines of it here.
826
00:43:43,455 --> 00:43:47,055
Cause you the last three lines are
really, what are the workhorse of
827
00:43:47,055 --> 00:43:48,765
making this thing stop hallucinating.
828
00:43:49,095 --> 00:43:49,785
But here it is.
829
00:43:50,384 --> 00:43:50,745
Okay?
830
00:43:52,125 --> 00:43:55,185
And so if you can't
answer because you don't.
831
00:43:55,635 --> 00:44:01,845
Have the information, output, a
clarifying question, and allow me to
832
00:44:01,845 --> 00:44:04,935
respond by providing the information.
833
00:44:05,745 --> 00:44:06,495
So we're telling it.
834
00:44:06,615 --> 00:44:08,655
You don't know the
answer, don't make it up.
835
00:44:09,075 --> 00:44:14,835
Figure out something to ask me to help
you answer the question, then ask it
836
00:44:14,835 --> 00:44:20,445
to me and let me give you information
that you can answer the question.
837
00:44:24,885 --> 00:44:25,215
Right.
838
00:44:25,845 --> 00:44:26,985
Then it tells me it understood.
839
00:44:26,985 --> 00:44:27,855
What's my question?
840
00:44:28,275 --> 00:44:29,595
When is Atiba de SU's birthday?
841
00:44:29,595 --> 00:44:31,815
And now I was, I was really hoping
that this wouldn't work, but it
842
00:44:31,820 --> 00:44:33,075
has no idea where my birthday is.
843
00:44:33,375 --> 00:44:34,035
Of course it doesn't.
844
00:44:37,365 --> 00:44:37,725
Right?
845
00:44:37,935 --> 00:44:43,515
And it asks the appropriate question,
who is this famed Atiba de Suza?
846
00:44:43,545 --> 00:44:44,655
And why don't I know him?
847
00:44:46,605 --> 00:44:46,755
Okay.
848
00:44:46,755 --> 00:44:47,535
I didn't say all that.
849
00:44:51,165 --> 00:44:59,325
This will help it stop hallucinating and
then also tell you how it needs help.
850
00:45:01,905 --> 00:45:04,455
How it needs help.
851
00:45:06,615 --> 00:45:06,705
Okay.
852
00:45:10,425 --> 00:45:16,965
If this was mind blowing to you,
put mind blowing in the chat for me.
853
00:45:18,225 --> 00:45:19,905
It was to me when I first learned it.
854
00:45:29,235 --> 00:45:29,565
Okay.
855
00:45:31,335 --> 00:45:32,235
Y'all ready for this?
856
00:45:36,015 --> 00:45:36,615
Now?
857
00:45:36,675 --> 00:45:39,915
This is, I think this is the second
to last slide, so I know I've, I've
858
00:45:39,915 --> 00:45:42,615
promised a lot in, in the, the last slide.
859
00:45:43,045 --> 00:45:43,265
Um,
860
00:45:47,130 --> 00:45:52,650
We talk a lot about training
chat, G P T and one I.
861
00:45:54,300 --> 00:45:59,490
I somewhat hope that you understand
from what I've said, that the concept of
862
00:45:59,490 --> 00:46:04,260
how we've been thinking about training
chat, g P t is a little bit flawed.
863
00:46:06,570 --> 00:46:06,900
Okay.
864
00:46:06,930 --> 00:46:10,560
Because if we're asking it 5, 6, 7
prompts and making it respond to a bunch
865
00:46:10,560 --> 00:46:14,760
of stuff and hoping that it remembered
everything from prompt number five,
866
00:46:15,600 --> 00:46:18,420
because that was the one where it
finally gave us the right answer, and
867
00:46:18,420 --> 00:46:25,710
then we started doing stuff from there,
doesn't really work because it's going to
868
00:46:25,770 --> 00:46:32,250
chop, chop, chop, interpret, interpret,
interpret, and quote unquote, forget.
869
00:46:33,375 --> 00:46:34,125
Forget.
870
00:46:34,395 --> 00:46:36,075
So how do we help it?
871
00:46:37,095 --> 00:46:43,395
How do we help it remember what
we needed to remember and then
872
00:46:43,395 --> 00:46:46,185
produce what we wanted to produce?
873
00:46:47,055 --> 00:46:51,884
Well, the answer to that question is
actually quite simple in the model, and
874
00:46:51,890 --> 00:46:55,634
I, I don't know why more people don't
talk about this, but in the model, the
875
00:46:55,634 --> 00:46:57,015
first prompt is the most important.
876
00:46:59,220 --> 00:47:02,100
What's unfortunate about that
is for most of us, the first
877
00:47:02,100 --> 00:47:03,270
prompt is the most generic.
878
00:47:04,770 --> 00:47:08,880
So if you saw my presentation in in in
Austin, one of the things that I was
879
00:47:08,884 --> 00:47:12,779
talking about is getting to the place
where instead of thinking about this
880
00:47:12,779 --> 00:47:19,140
as chain of thought prompting, how
do you write the perfect prompt that
881
00:47:19,200 --> 00:47:21,900
produces the result that you want?
882
00:47:24,525 --> 00:47:26,625
The first prompt is the most important.
883
00:47:27,075 --> 00:47:31,575
When it has to decide what to
prioritize, it's always going to default
884
00:47:31,695 --> 00:47:33,825
to remembering what you said first.
885
00:47:36,525 --> 00:47:38,565
If your first prompt was the most generic,
886
00:47:42,375 --> 00:47:43,815
that's why I forget sometimes.
887
00:47:44,805 --> 00:47:51,464
So there's a, there's a, a format to
this, to your first prompt format, macros
888
00:47:51,464 --> 00:47:56,145
and style, format, macros and style.
889
00:47:58,125 --> 00:47:58,455
Okay?
890
00:47:58,545 --> 00:48:00,255
So that, this is the last slide.
891
00:48:00,255 --> 00:48:01,335
Doesn't it answer everything?
892
00:48:03,674 --> 00:48:04,424
It's a joke.
893
00:48:04,424 --> 00:48:04,694
Y'all.
894
00:48:05,505 --> 00:48:06,165
Somebody laugh.
895
00:48:06,194 --> 00:48:09,645
Put, put laughter in, in the,
in the chat bot in the, uh,
896
00:48:13,060 --> 00:48:13,439
I love it.
897
00:48:14,839 --> 00:48:16,680
Oh, let me show you what this looks like.
898
00:48:16,919 --> 00:48:22,799
And, um, in the handouts for today, I
have, um, a Google doc that has this
899
00:48:22,805 --> 00:48:27,240
prompt in it because listen, this is
the prompt here, and let me see if
900
00:48:27,240 --> 00:48:28,649
I can make this a little bit bigger.
901
00:48:32,520 --> 00:48:33,419
This is the prompt.
902
00:48:34,649 --> 00:48:36,629
All of this, all of that.
903
00:48:36,930 --> 00:48:38,040
It keeps going.
904
00:48:38,370 --> 00:48:39,540
It's still going.
905
00:48:40,140 --> 00:48:41,399
Still going, y'all.
906
00:48:42,015 --> 00:48:46,995
All of this, this is
your ideal first prompt.
907
00:48:48,705 --> 00:48:52,995
Now let me break it down for
you and you can modify this.
908
00:48:52,995 --> 00:48:58,334
This is a very, very, very, very
generic version that you can modify.
909
00:48:58,515 --> 00:48:58,875
Okay?
910
00:48:59,334 --> 00:49:04,245
Um, you are my general answer device.
911
00:49:06,120 --> 00:49:08,310
Okay, you told the who, what it is.
912
00:49:08,940 --> 00:49:10,440
This is prompt number one.
913
00:49:10,710 --> 00:49:12,600
Your response will be output number one.
914
00:49:12,605 --> 00:49:14,880
Your next response will
be output number two.
915
00:49:16,050 --> 00:49:16,440
Okay.
916
00:49:16,529 --> 00:49:17,040
Number.
917
00:49:17,040 --> 00:49:18,270
Your response is every time.
918
00:49:18,450 --> 00:49:20,310
It does not always like this very well.
919
00:49:20,340 --> 00:49:21,600
Just being honest.
920
00:49:22,170 --> 00:49:22,500
Okay.
921
00:49:22,590 --> 00:49:27,600
Um, here is the format of my inputs.
922
00:49:29,340 --> 00:49:34,710
PP means previous input prompt from me.
923
00:49:35,820 --> 00:49:36,540
You're doing a prompt.
924
00:49:36,540 --> 00:49:40,500
You wanna reference a previous input,
just told you how you can tell it.
925
00:49:40,590 --> 00:49:43,980
That PP means the previous input prompt.
926
00:49:43,980 --> 00:49:45,600
That's what I'm talking about right now.
927
00:49:47,595 --> 00:49:51,045
PO means previous output from you.
928
00:49:51,225 --> 00:49:56,355
So in your prompt you can type
PO and it knows you are talking
929
00:49:56,355 --> 00:49:58,845
about its previous output
930
00:50:02,145 --> 00:50:02,715
expert.
931
00:50:02,745 --> 00:50:05,475
This is just a term you can use
any term that you want here,
932
00:50:05,775 --> 00:50:08,205
expert means, and then you define.
933
00:50:08,445 --> 00:50:11,985
So think, um, ideal customer.
934
00:50:12,450 --> 00:50:16,169
Think, what am I, who, how, what's
my perspective on the world?
935
00:50:16,350 --> 00:50:17,940
Anything like that you can put in here.
936
00:50:17,940 --> 00:50:18,899
We use expert.
937
00:50:18,990 --> 00:50:23,399
Again, this is a generic
prompt that you can customize.
938
00:50:24,180 --> 00:50:27,839
We use expert because most people
understand what an expert is, right?
939
00:50:28,169 --> 00:50:32,460
Relevant, um, 20 years in the
field, multiple PhDs, and then
940
00:50:32,460 --> 00:50:36,450
we even tell how it, they think
they're, um, they're un orthodox,
941
00:50:36,480 --> 00:50:39,060
less known advice in their answer.
942
00:50:39,645 --> 00:50:41,955
Right, um, style.
943
00:50:42,165 --> 00:50:48,345
When I use the word style, it means use
the following style guide in the writing.
944
00:50:48,825 --> 00:50:49,815
I'm not gonna read all this.
945
00:50:50,805 --> 00:50:52,215
Y'all have it in in in the doc.
946
00:50:52,215 --> 00:50:54,495
Y'all can go through and read,
but it it, it kind of goes through
947
00:50:54,495 --> 00:50:59,265
and tells you tonality stuff, all
metaphors, et cetera, et cetera,
948
00:50:59,505 --> 00:51:02,595
of how we want our style response.
949
00:51:02,595 --> 00:51:06,855
So now think of what Jeff was
doing last week in his session.
950
00:51:07,335 --> 00:51:08,865
Right when he had all that.
951
00:51:08,924 --> 00:51:10,634
Jeff can put all of that right here.
952
00:51:11,475 --> 00:51:12,105
First prompt.
953
00:51:12,705 --> 00:51:13,875
It won't forget.
954
00:51:17,505 --> 00:51:20,895
Then you can define different
terms of what you wanted to do.
955
00:51:21,134 --> 00:51:21,645
Critique.
956
00:51:22,095 --> 00:51:25,785
I want to, for whatever reason,
I wanna scan this material and
957
00:51:25,785 --> 00:51:27,015
tell me a list of issues with it.
958
00:51:27,645 --> 00:51:29,475
Tell me things that are
concerning about this.
959
00:51:29,475 --> 00:51:33,075
You know, you, you write a blog
post and, and, um, you wanna know
960
00:51:33,134 --> 00:51:35,805
that I include everything Again,
it will critique what you wrote.
961
00:51:37,230 --> 00:51:42,210
Anything in parentheses is
the perspective from which I
962
00:51:42,210 --> 00:51:44,280
want you to write your answer.
963
00:51:44,610 --> 00:51:48,690
So I just told it what
parentheses mean to me.
964
00:51:49,470 --> 00:51:50,850
You can put anything that you want.
965
00:51:50,850 --> 00:51:54,960
You can change this to make parentheses
mean, whatever you want it to make.
966
00:51:55,800 --> 00:51:57,390
Same thing with curly brackets.
967
00:52:00,660 --> 00:52:05,670
Anything with, uh, between the colon
is the input to use for writing.
968
00:52:07,049 --> 00:52:09,960
And then it goes through three words
or less and some more information here.
969
00:52:10,140 --> 00:52:11,279
Now we give it examples.
970
00:52:12,870 --> 00:52:14,250
Now we're gonna feed an example.
971
00:52:14,339 --> 00:52:23,100
So here's a prompt y'all in,
um, in square brackets, right?
972
00:52:23,430 --> 00:52:25,380
Is what I want you to accomplish or write.
973
00:52:25,560 --> 00:52:26,880
I want add copy.
974
00:52:27,180 --> 00:52:29,069
What's the perspective, right?
975
00:52:29,100 --> 00:52:35,670
Is expert inside of, um, parentheses
and then it's going to be a Facebook ad.
976
00:52:36,575 --> 00:52:38,195
And here is after the colon.
977
00:52:38,495 --> 00:52:39,875
Here's what I want you to write about.
978
00:52:44,815 --> 00:52:45,595
That's the prompt.
979
00:53:01,125 --> 00:53:03,945
And then if we give
how it would translate.
980
00:53:09,794 --> 00:53:13,334
The question mark symbol
means this example.
981
00:53:13,575 --> 00:53:16,544
So this now says previous output.
982
00:53:17,865 --> 00:53:19,004
I got a question about it.
983
00:53:19,785 --> 00:53:25,575
Will trans will translate into, explain
your reasoning for your last output.
984
00:53:26,955 --> 00:53:32,714
All you typed in was that now
you wanna talk about saving.
985
00:53:33,839 --> 00:53:35,160
Um, characters.
986
00:53:35,250 --> 00:53:39,240
We're saving characters and we're
getting the computer closer to
987
00:53:39,240 --> 00:53:40,920
what it already understands.
988
00:53:45,029 --> 00:53:45,990
We have the minus symbol.
989
00:53:45,990 --> 00:53:47,339
I'm not gonna go go through all of these.
990
00:53:47,345 --> 00:53:49,560
There's, there's a ton, right?
991
00:53:49,950 --> 00:53:54,210
Um, and then we have examples
of how this all works out.
992
00:53:56,835 --> 00:53:57,165
Okay.
993
00:53:57,495 --> 00:54:00,375
Uh, one of their favorites here is
you could take the previous output.
994
00:54:00,375 --> 00:54:03,645
So let's say it wrote something and
you want it to be funnier, or you
995
00:54:03,645 --> 00:54:06,675
want it to be sillier, or it was too
silly and you want it to be serious.
996
00:54:06,885 --> 00:54:08,955
Well, you can give it a
perspective and tell it that.
997
00:54:16,165 --> 00:54:16,455
Okay?
998
00:54:16,715 --> 00:54:19,275
You can even go to the point
of giving it a percentage.
999
00:54:19,275 --> 00:54:21,345
So I wanted 20% more serious.
1000
00:54:23,640 --> 00:54:25,590
Here's my goal here, okay?
1001
00:54:25,590 --> 00:54:28,380
Because I know this is a lot and
this prompt is I'm not even done.
1002
00:54:28,380 --> 00:54:29,280
The prompt keeps going.
1003
00:54:29,550 --> 00:54:37,320
My goal, as I told you guys a little while
ago, there is stuff that you can do with
1004
00:54:37,320 --> 00:54:40,860
AI that I know that's different than what.
1005
00:54:42,015 --> 00:54:46,545
Um, other people know, and we're all
working collaboratively together.
1006
00:54:46,755 --> 00:54:50,295
And the purpose of all of us working
collaboratively, collaboratively
1007
00:54:50,300 --> 00:54:57,255
together is so that we can
create greater, there is greater.
1008
00:54:57,255 --> 00:55:03,285
And this is a, this is a basic advanced
prompt, if that makes any sense to you.
1009
00:55:04,965 --> 00:55:05,205
Okay.
1010
00:55:05,205 --> 00:55:07,185
When we get into advanced
prompting, this is the, the,
1011
00:55:08,175 --> 00:55:09,885
this just scratches the surface.
1012
00:55:10,800 --> 00:55:12,870
There's so much more that you can do.
1013
00:55:13,200 --> 00:55:15,090
There's so much more that you can learn.
1014
00:55:15,180 --> 00:55:19,860
So I'm also going to challenge
you that as you're learning, there
1015
00:55:19,860 --> 00:55:22,410
are a lot of people who are giving
you the same generic answers.
1016
00:55:22,560 --> 00:55:23,190
Go deeper.
1017
00:55:25,110 --> 00:55:25,770
Go deeper.
1018
00:55:25,800 --> 00:55:27,660
There's a lot more here to learn.
1019
00:55:28,620 --> 00:55:28,950
Okay.
1020
00:55:29,190 --> 00:55:30,210
What does that all look like?
1021
00:55:30,720 --> 00:55:34,320
All right, so we're gonna write a
video script from an expert opinion.
1022
00:55:34,320 --> 00:55:38,100
It's gonna be YouTube short about why
digital market asserts are the best.
1023
00:55:39,645 --> 00:55:40,005
Okay.
1024
00:55:40,245 --> 00:55:43,605
I like it to include a hook,
scene, selection and voiceover.
1025
00:55:43,605 --> 00:55:47,985
So I took the, the basic prompt
and I then I said I want specific,
1026
00:55:49,515 --> 00:55:51,645
um, output format as well.
1027
00:55:51,915 --> 00:55:54,255
Um, the please write in English
is optional, whether you want
1028
00:55:54,255 --> 00:55:55,845
that or not, and it did that.
1029
00:56:00,835 --> 00:56:01,125
Okay.
1030
00:56:02,175 --> 00:56:06,915
Then I wanted it to be funnier
because that's pretty dry.
1031
00:56:07,950 --> 00:56:12,630
Are you tired of sifting through endless
online courses, wondering which ones
1032
00:56:12,630 --> 00:56:16,620
are actually worthy or worth your time?
1033
00:56:16,800 --> 00:56:18,210
Eh, that's pretty dry.
1034
00:56:19,050 --> 00:56:21,240
Mark wouldn't write that, right?
1035
00:56:21,240 --> 00:56:21,570
Mark?
1036
00:56:23,640 --> 00:56:24,270
Or maybe he would.
1037
00:56:24,420 --> 00:56:24,960
No, he wouldn't.
1038
00:56:24,990 --> 00:56:25,470
Okay, good.
1039
00:56:25,860 --> 00:56:30,570
All right, so stick of scrolling through
endless online marketing courses.
1040
00:56:30,690 --> 00:56:34,320
Get certified with digital marketer.com
and watch your marketing game.
1041
00:56:34,320 --> 00:56:35,790
Go from zero to hero.
1042
00:56:39,330 --> 00:56:39,660
Right.
1043
00:56:39,720 --> 00:56:40,350
It goes on.
1044
00:56:40,350 --> 00:56:46,319
And then the, the, um, funnier actually
came in, in the voiceover, right?
1045
00:56:46,319 --> 00:56:47,670
Don't be like Dave.
1046
00:56:47,730 --> 00:56:53,430
They thought he could handle marketing
with just a few flyers and a prayer,
1047
00:56:53,879 --> 00:56:57,149
but now Dave's business is struggling.
1048
00:56:57,629 --> 00:57:00,240
Don't be like Dave.
1049
00:57:01,125 --> 00:57:05,205
Get certified with digital
marketer.com and take your
1050
00:57:05,205 --> 00:57:07,185
marketing skills to the next level.
1051
00:57:07,365 --> 00:57:10,665
It's time to become a marketing superhero.
1052
00:57:11,475 --> 00:57:12,225
Absolutely love that.
1053
00:57:12,230 --> 00:57:13,515
It uses superhero motif.
1054
00:57:13,520 --> 00:57:15,315
I didn't ask it to, but it it, it did.
1055
00:57:15,465 --> 00:57:16,005
So it was cool.
1056
00:57:17,025 --> 00:57:17,355
Okay.
1057
00:57:17,715 --> 00:57:19,335
Um, all right.
1058
00:57:19,605 --> 00:57:27,105
I know I just gave you a lot, so I'm gonna
pause right now and take some questions.
1059
00:57:29,445 --> 00:57:29,685
Okay.
1060
00:57:32,970 --> 00:57:40,950
Vicky with the limit on the data that
chat g b t has, let's say, ingested.
1061
00:57:41,400 --> 00:57:49,260
Um, our topic is something, is something
now that has been growing since 22 to 20.
1062
00:57:49,290 --> 00:57:49,470
Okay?
1063
00:57:50,130 --> 00:57:51,660
Uh, so there isn't much content in chat.
1064
00:57:51,660 --> 00:57:56,610
G b T go to Bing, not, not
bing, go to, um, go to Bard.
1065
00:57:57,090 --> 00:57:58,590
That's my answer for you, Vicky.
1066
00:57:58,860 --> 00:57:59,350
Go to Bard.
1067
00:58:01,890 --> 00:58:06,180
Hopefully when chat g p t comes out with
Ablo browser plugin, that would help.
1068
00:58:06,720 --> 00:58:10,589
But, um, right now I'd say go to Bard.
1069
00:58:16,229 --> 00:58:16,839
Have you gotten back
1070
00:58:20,700 --> 00:58:21,370
you saying something, mark?
1071
00:58:21,790 --> 00:58:23,939
No, I was gonna just go
to the list of questions.
1072
00:58:24,720 --> 00:58:24,930
Oh.
1073
00:58:26,790 --> 00:58:31,185
Uh, What is the benefit of
saving the number of characters?
1074
00:58:31,245 --> 00:58:32,145
Um, Harry.
1075
00:58:32,235 --> 00:58:33,705
Um, thanks for your question, Harry.
1076
00:58:34,005 --> 00:58:36,945
Uh, the benefit of saving the number of
characters depends on what your output is.
1077
00:58:36,945 --> 00:58:42,135
If you prioritize your output and need
a lot of output and you're out, you're
1078
00:58:42,135 --> 00:58:44,685
starting to reach the output limit.
1079
00:58:44,745 --> 00:58:47,805
Then the num, you wanna save as
many characters as you can so that
1080
00:58:47,805 --> 00:58:50,145
you can get better and more output.
1081
00:58:51,225 --> 00:58:51,555
Okay.
1082
00:58:52,035 --> 00:58:56,055
Um, I guess one of the things to also
understand with me, I, I get that you
1083
00:58:56,055 --> 00:58:58,035
can write more pleas and all of this.
1084
00:58:58,335 --> 00:59:01,485
I'm very much into writing one
prompt, getting my answer and leaving.
1085
00:59:02,685 --> 00:59:05,535
I get that, that chat g p t
was created so that it could
1086
00:59:05,535 --> 00:59:07,515
be chain of thought prompting.
1087
00:59:07,965 --> 00:59:08,595
I get that.
1088
00:59:09,165 --> 00:59:14,595
However, chain of thought prompting
is based on single thought prompting
1089
00:59:15,645 --> 00:59:18,825
and they just extended the model,
so they're trying to get you to that
1090
00:59:18,830 --> 00:59:21,345
anyway, so, I just tried to go there.
1091
00:59:32,325 --> 00:59:32,775
Yes.
1092
00:59:32,775 --> 00:59:33,944
So you can use ic.
1093
00:59:33,975 --> 00:59:34,305
Sorry.
1094
00:59:34,365 --> 00:59:37,214
Uh, mark asked, um, I can't
remember to read the questions.
1095
00:59:37,665 --> 00:59:43,725
Um, so, uh, you can use the I C P
throughout the rest of the chat.
1096
00:59:43,815 --> 00:59:45,315
In the variable example.
1097
00:59:45,315 --> 00:59:45,805
Yes, you can.
1098
00:59:49,200 --> 00:59:52,650
Harry, you just said by using
variables, so I'm not sure.
1099
00:59:53,280 --> 00:59:59,522
Oh, was that the, was that, oh, that was
the extension of your earlier question.
85830
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.