1
00:00:00,000 --> 00:00:07,200
Hey, welcome back to the Po-Party, it's Jules. Today we are talking about key art, making

2
00:00:07,200 --> 00:00:12,160
key art. And this is going to be a bit of a workshop going on the vein of branding and

3
00:00:12,160 --> 00:00:17,080
the components of a brand package. We've discussed a lot about brand this past month. That's

4
00:00:17,080 --> 00:00:23,640
your typography, your logo, colors, mood boards, etc. There are other workshops on this podcast

5
00:00:23,640 --> 00:00:27,920
on how to develop the basics of a brand. And I give you a few of my opinions about how

6
00:00:27,920 --> 00:00:33,760
to get started easy, fast, and also find inspiration. So check out those first few episodes of the

7
00:00:33,760 --> 00:00:39,120
Po-Party. They're in the feed. They're in this season 1, 2, and 3. Today we're going

8
00:00:39,120 --> 00:00:45,120
to talk about key art, what it is, how it contributes to a brand package, when you need

9
00:00:45,120 --> 00:00:51,480
it. And I'll go into some examples of what key art can be and how you would make it for

10
00:00:51,480 --> 00:00:56,200
a music festival, for example. Then we'll also touch a little bit on AI because since

11
00:00:56,200 --> 00:01:00,400
we're talking about images and how to make them, it's almost impossible not to talk about

12
00:01:00,400 --> 00:01:06,520
this juicy topic today. I've generated a few AI images myself. I've also photoshopped the

13
00:01:06,520 --> 00:01:13,080
crap out of a ton of them. So I have a few ideas on how to use still generative AI art.

14
00:01:13,080 --> 00:01:17,360
And I'll give you some of my thoughts on what AI is good for, what it's not good for. But

15
00:01:17,360 --> 00:01:21,200
that in itself can be another episode. So we'll touch on it briefly. But today, let's

16
00:01:21,200 --> 00:01:26,720
go ahead and dive into this topic of making key art, what it's for, and we'll get to

17
00:01:26,720 --> 00:01:31,320
know a little bit more about how it fits into your overall brand. Alright, let's go.

18
00:01:31,320 --> 00:01:36,460
So what is key art? Key art is essentially all of the graphics or illustrations, maybe

19
00:01:36,460 --> 00:01:41,800
even photography that you would typically see for a campaign. Now, what is a campaign?

20
00:01:41,800 --> 00:01:48,360
A campaign can be any sort of movement that you make for marketing, whether that's advertising

21
00:01:48,360 --> 00:01:54,800
the Olympics or a new flavor of Doritos or a new album you have coming out. Generally,

22
00:01:54,800 --> 00:02:01,160
we call these branding fleets campaigns because they serve a specific purpose, whether it's

23
00:02:01,160 --> 00:02:05,840
to announce the drop of a new line, you want to tell people that you have something new

24
00:02:05,840 --> 00:02:11,240
coming out. But the idea is that it serves the purpose of getting the information out

25
00:02:11,240 --> 00:02:17,260
to the public and then that campaign is over. You might refer to it in the future, but generally,

26
00:02:17,260 --> 00:02:23,120
the art really matters a ton because while your brand is the key components of what you

27
00:02:23,120 --> 00:02:28,240
look like through your fonts, your logos, your colors, the key art can be seasonal,

28
00:02:28,240 --> 00:02:34,420
it can define the moment. And I think a lot of the times we see key art used during special

29
00:02:34,420 --> 00:02:40,920
seasonal events like the Superbowl, Valentine's Day, name any holiday that you would celebrate.

30
00:02:40,920 --> 00:02:46,280
Lunar New Year, for example. I personally love seeing New Year's campaigns. I think

31
00:02:46,280 --> 00:02:51,360
they're fun and then also many people in the field of astrology, I think could also

32
00:02:51,360 --> 00:02:57,800
benefit from utilizing key art for the season. So think of key art as foundational work that

33
00:02:57,800 --> 00:03:03,360
you would show in a seasonal campaign. And I know a lot of examples, but even if you're

34
00:03:03,360 --> 00:03:08,760
an artist or designer trying to launch your brand and you are offering a new service,

35
00:03:08,760 --> 00:03:14,200
that service might have specific forms of key art that you're using in order to advertise

36
00:03:14,200 --> 00:03:20,320
your product. So another question people might have is, do you need key art? And honestly,

37
00:03:20,320 --> 00:03:25,660
you do not always need key art. I think key art belongs to the illustrators and visual

38
00:03:25,660 --> 00:03:31,160
designers of the world, maybe even the photographers and the presentation designers. I think that

39
00:03:31,160 --> 00:03:36,320
using key art is important when you're trying to create a campaign, but what's even more

40
00:03:36,320 --> 00:03:41,560
important is getting the look and feel of the key art down with the messaging of the

41
00:03:41,560 --> 00:03:45,000
campaign. Let's say there's an illustrator who has a particular style you like and you

42
00:03:45,000 --> 00:03:49,880
think it matches the campaign, you found them on Behance, that's perfect. You can go ahead

43
00:03:49,880 --> 00:03:55,400
and contact them and ask them to work on your campaign and maybe get like three to four

44
00:03:55,400 --> 00:04:01,280
visuals that you could then use on your email marketing, your posters, your mailers, et

45
00:04:01,280 --> 00:04:06,160
cetera. Now key art is important for this specific reason because generally in a campaign,

46
00:04:06,160 --> 00:04:10,860
you might have very many different ways of communicating with people. I mentioned a few,

47
00:04:10,860 --> 00:04:14,700
maybe you're reaching out to them with flyers or you're creating a new landing page. If

48
00:04:14,700 --> 00:04:19,880
you have specific graphics for that campaign, it's going to be easily tied to the moment,

49
00:04:19,880 --> 00:04:24,240
the season, the announcement you're trying to make for that specific campaign. And you

50
00:04:24,240 --> 00:04:28,000
really can just pop those graphics in there and that just ties everything together. That's

51
00:04:28,000 --> 00:04:33,160
one of the best things about key art. Now I touch upon this already, but one of the

52
00:04:33,160 --> 00:04:38,940
key things to know about key art is that it is so important still to get the look and

53
00:04:38,940 --> 00:04:44,520
feel of your brand down with key art. Let's talk about an example. Let's say you're

54
00:04:44,520 --> 00:04:50,480
launching a Superbowl campaign for Doritos during the halftime show and you're trying

55
00:04:50,480 --> 00:04:55,760
to make key art. It might not really make a lot of sense to make key art for Doritos

56
00:04:55,760 --> 00:05:01,960
that feels really pastel and bubbly and lovely because that doesn't really fit with the

57
00:05:01,960 --> 00:05:07,400
brand of Doritos. When I think of Doritos, I think angular, spicy, flavorful, crunchy,

58
00:05:07,400 --> 00:05:13,000
like loud and red. So it would be really important to find someone who's really good at matching

59
00:05:13,000 --> 00:05:18,280
that style. Also, it really depends on what the flavor is and if there's something particularly

60
00:05:18,280 --> 00:05:23,480
new about this Dorito. Let's say for example, that this Dorito was a puff chip. Maybe you

61
00:05:23,480 --> 00:05:28,340
would get someone who is really good at a 3D art style that is puffy. There's lots

62
00:05:28,340 --> 00:05:33,120
of different types of key art you can make, but generally speaking, since key art can

63
00:05:33,120 --> 00:05:39,440
be so many different things like 3D art, doodles, photography, AI art, I think it's a good idea

64
00:05:39,440 --> 00:05:45,840
to keep it simple and maybe just choose one to two categories of key art for a campaign

65
00:05:45,840 --> 00:05:50,220
just so it doesn't start getting too busy. Because one of the things I mentioned about

66
00:05:50,220 --> 00:05:54,680
key art is that it's easy to pop in and then just tie back to a campaign. If you start

67
00:05:54,680 --> 00:05:59,960
getting into the territory of way too many different styles, then it becomes harder to

68
00:05:59,960 --> 00:06:04,560
tie it all together. And this is what art directors are for, I believe, just really

69
00:06:04,560 --> 00:06:10,640
making sure that messaging is the same across all campaigns visually and then also working

70
00:06:10,640 --> 00:06:15,400
with the other creators who are on a project. And going on a little bit of a tangent, but

71
00:06:15,400 --> 00:06:20,040
in the same vein, these days, many brands work with content creators. I think one of

72
00:06:20,040 --> 00:06:25,960
the hardest things to nail is actually dignified, but put together look, style, and feel among

73
00:06:25,960 --> 00:06:31,040
all different types of content that creators can contribute to a campaign. One of the other

74
00:06:31,040 --> 00:06:36,720
great things about key art is that sometimes people develop whole experiences around key

75
00:06:36,720 --> 00:06:43,160
art. So I think a really good example of this is with character designs and games. So let's

76
00:06:43,160 --> 00:06:48,560
say for example, Zeltnett. Sometimes this is just an example, but sometimes character

77
00:06:48,560 --> 00:06:54,280
designers might show the designs for a new character and then a game is created entirely

78
00:06:54,280 --> 00:06:59,440
around that look and feel of a character. This is because it's much easier to get one

79
00:06:59,440 --> 00:07:04,480
component, one small piece of a world down rather than creating the world from scratch

80
00:07:04,480 --> 00:07:09,080
first and then creating the small components. So if you're working on a team with lots

81
00:07:09,080 --> 00:07:14,480
of different designers and artists, then maybe someone that an art director creates a piece

82
00:07:14,480 --> 00:07:18,800
of key art, just one of them, and then everyone else nails down that style, that look and

83
00:07:18,800 --> 00:07:24,160
feel and then bases the work off of this initial design. This is great because it can become

84
00:07:24,160 --> 00:07:29,840
a beacon for a team of artists on large sets. Games are a great example because they take

85
00:07:29,840 --> 00:07:35,080
so long to develop and there's a lot of different moving pieces. So we've talked a little bit

86
00:07:35,080 --> 00:07:40,200
about what key art is, what it can do for you and your campaign and how it might help

87
00:07:40,200 --> 00:07:44,960
working on a large team of designers. Now we're going to go into an example of making

88
00:07:44,960 --> 00:07:49,520
key art for a music festival that takes place in the state of Florida. And I'm using this

89
00:07:49,520 --> 00:07:55,360
example from a very particular music festival that I once went to, which was based in Okeechobee.

90
00:07:55,360 --> 00:08:00,560
So I'm back and we're in the backyard, but let's go ahead and work out this exercise

91
00:08:00,560 --> 00:08:07,720
and making key art for a music festival based at a place in Florida called Okeechobee. So

92
00:08:07,720 --> 00:08:12,480
Okeechobee is a place full of swamps and I don't know if you've ever been to the Okeechobee

93
00:08:12,480 --> 00:08:17,440
music festival, doesn't really matter. But when I was there, they had all these little

94
00:08:17,440 --> 00:08:22,360
islands. There were some different stages. I felt like an alligator was going to crawl

95
00:08:22,360 --> 00:08:27,840
out at any moment. But if we're going back to this idea of making key art for a poster,

96
00:08:27,840 --> 00:08:32,820
one of the ways that I would brainstorm essentially approaching the task of illustrating for this

97
00:08:32,820 --> 00:08:40,380
festival would be to brainstorm different ideas of visuals that represent the place.

98
00:08:40,380 --> 00:08:46,160
So that could be people who are there, the setting itself, where the event takes place,

99
00:08:46,160 --> 00:08:52,120
and then the types of things you might see there like food, music, activities. And I

100
00:08:52,120 --> 00:08:57,120
love music festivals, especially for illustration exercises, because if you've ever been to

101
00:08:57,120 --> 00:09:02,200
a festival, you know that there are some key things that stay the same every single year,

102
00:09:02,200 --> 00:09:09,040
but then there are usually seasonal elements that change, maybe new vendors or like a really

103
00:09:09,040 --> 00:09:16,000
dramatic piece of art that is just put up somewhere in the middle of the campground.

104
00:09:16,000 --> 00:09:19,460
And so as an artist and an illustrator, I think one of the ways that I would approach

105
00:09:19,460 --> 00:09:25,000
making key art for this type of event is to basically write down the types of things that

106
00:09:25,000 --> 00:09:30,920
I would see at this event, maybe even discuss with the festival organizers what would go

107
00:09:30,920 --> 00:09:37,040
on. And then from basically brainstorming and creating a word cloud of ideas, what we

108
00:09:37,040 --> 00:09:43,660
could then do is do some really quick sketches, set a timer for about 15 minutes and not really

109
00:09:43,660 --> 00:09:49,760
doing complete fleshed out sketches, but just doing little thumbnail designs of what these

110
00:09:49,760 --> 00:09:53,840
images would look like. Now, if you've never heard of a thumbnail before, this is a great

111
00:09:53,840 --> 00:10:00,160
time for me to explain what it is. A thumbnail is essentially a one inch by one inch drawing

112
00:10:00,160 --> 00:10:04,720
typically done on scratch paper in pencil or with pen. I like pencil because you can

113
00:10:04,720 --> 00:10:10,680
erase and you get the idea that these are just previews. And then you try to essentially

114
00:10:10,680 --> 00:10:16,040
get down the drawing. So let's say I'm trying to draw a frog. I'm going to spend two minutes

115
00:10:16,040 --> 00:10:21,280
drawing a frog. And then if I like it, great. If not, then I'm going to spend another two

116
00:10:21,280 --> 00:10:26,520
minutes drawing a different frog. The idea here is not to get these drawings perfect.

117
00:10:26,520 --> 00:10:30,520
We don't want to get these drawings perfect, but what we do want to do is get a bunch of

118
00:10:30,520 --> 00:10:37,520
ideas down onto paper, warm up our hand and get the brain flowing. If you are an artist,

119
00:10:37,520 --> 00:10:42,040
designer, illustrator, someone who usually works with drawing images, this might be a

120
00:10:42,040 --> 00:10:46,760
little bit easier for you, but I would say still engage with this exercise even if you're

121
00:10:46,760 --> 00:10:52,080
not used to it, because it'll start to get you thinking about images, not being so hard

122
00:10:52,080 --> 00:10:57,160
on yourself. And then just playing with shapes and being a little bit more abstract is important

123
00:10:57,160 --> 00:11:01,320
here. Just try to be super quick with these drawings because then you'll get a sense for

124
00:11:01,320 --> 00:11:06,700
what's actually worth spending your time on once you get to the computer and start making

125
00:11:06,700 --> 00:11:12,840
more fleshed out. Something I sometimes see with newer designers is that they really love

126
00:11:12,840 --> 00:11:17,040
playing with color. Maybe they'll choose like some pastels or crayons and they'll get really

127
00:11:17,040 --> 00:11:23,280
sucked into the weeds and the details and composition of one single drawing. And I think

128
00:11:23,280 --> 00:11:28,880
that's fine for a single piece of artwork, but even before you get to the stage of deciding

129
00:11:28,880 --> 00:11:33,680
you want to flesh out your artwork more, I think it's a great idea to just get all the

130
00:11:33,680 --> 00:11:39,520
ideas out there. And so with this thumb nailing exercise, maybe you even spend an hour, but

131
00:11:39,520 --> 00:11:44,640
you just spend five minutes on every single little idea that you have and you could draw

132
00:11:44,640 --> 00:11:50,440
the squares beforehand. So if we're dividing an hour by five minutes, what does that give

133
00:11:50,440 --> 00:11:54,880
us? That gives us drawings in the end and then you can choose the best ones that you

134
00:11:54,880 --> 00:12:00,880
like and then spend more time on just the best ones. The reason why we want to brainstorm

135
00:12:00,880 --> 00:12:05,400
and get all the ideas out on paper first is that we know that we've given ourselves the

136
00:12:05,400 --> 00:12:12,920
time to explore the possibilities and then we can worry about things like color, style,

137
00:12:12,920 --> 00:12:17,600
how we're going to execute fleshed out art. We're just doing these in pencil and so let's

138
00:12:17,600 --> 00:12:24,400
brainstorm out loud what kind of images we could make for this festival in Florida. I'm

139
00:12:24,400 --> 00:12:31,280
thinking alligators, islands, swamps, stages. What kind of food do you like to eat at festivals?

140
00:12:31,280 --> 00:12:37,520
I like hot dogs, I don't know, smoothies, guitars, dancing. Dancing might be a hard

141
00:12:37,520 --> 00:12:41,840
thing to draw just abstractly so if we narrow it down, maybe we do people, maybe there's

142
00:12:41,840 --> 00:12:46,540
a type of style of person. You can do stick figures if you like, but don't get so caught

143
00:12:46,540 --> 00:12:51,960
up on it. Sun, maybe it's water. We're just trying to make these kind of quick and then

144
00:12:51,960 --> 00:12:57,280
as we explore drawing out these designs, you can see what's really going to be feasible,

145
00:12:57,280 --> 00:13:00,720
but it's important to just let it flow and if it starts taking a little bit too much

146
00:13:00,720 --> 00:13:07,360
time that's a good indicator of moving on to the next idea. Alright, great. So at this

147
00:13:07,360 --> 00:13:11,600
point you should have all your thumbnails laid out. There's probably going to be about

148
00:13:11,600 --> 00:13:17,880
12 of them. What you can do is choose really five to six of the images that speak to you

149
00:13:17,880 --> 00:13:24,600
the most. What we would do with these images is try to make higher fidelity art with just

150
00:13:24,600 --> 00:13:30,560
the sketches. There are a few ways to do this. If you like to draw and use a very physical

151
00:13:30,560 --> 00:13:36,200
pen to paper process with markers, you can do that. You can just outline them in black.

152
00:13:36,200 --> 00:13:39,960
If you are a digital artist, you can take this to your computer. I would recommend from

153
00:13:39,960 --> 00:13:45,280
here actually starting another mood board process just for the illustrations and the

154
00:13:45,280 --> 00:13:49,980
reason why I would do this is because there's so many different ways to approach illustrations

155
00:13:49,980 --> 00:13:56,000
and we want the branding, the look, the feel and the vibe to really match the music festival.

156
00:13:56,000 --> 00:14:02,040
And so for example, while it might make sense to do kind of slings for something that's

157
00:14:02,040 --> 00:14:08,680
a little bit more handmade for a music festival, maybe you want to explore 3D art or you want

158
00:14:08,680 --> 00:14:14,800
to try some AI generated art. It's really up to you to decide as an artist and a designer,

159
00:14:14,800 --> 00:14:20,120
but the possibilities are endless. Now is the time to narrow down the possibilities

160
00:14:20,120 --> 00:14:25,180
and when you create mood boards, you can choose three directions. You can put together three

161
00:14:25,180 --> 00:14:30,160
different mood boards of what you might like to do. And if you're unfamiliar with mood

162
00:14:30,160 --> 00:14:34,800
boards, I highly recommend that you visit the episode on mood boards or finding your

163
00:14:34,800 --> 00:14:40,440
tastes and the episode on where to find design inspiration. And that's on this podcast.

164
00:14:40,440 --> 00:14:44,160
Go ahead and give those a listen and then you're just going to create some mood boards

165
00:14:44,160 --> 00:14:50,360
from websites like Designspiration, Beheath and find some illustrations, at least six

166
00:14:50,360 --> 00:14:55,560
images that have one specific style and then that's going to be one direction. You can

167
00:14:55,560 --> 00:14:59,640
create as many mood boards as you like, but at the end of the day, you need to choose

168
00:14:59,640 --> 00:15:04,560
a one direction that's really strong, really dignified and that's going to be applied to

169
00:15:04,560 --> 00:15:09,080
your illustrations. Don't spend too much time on this because the next thing you want to

170
00:15:09,080 --> 00:15:14,280
do is actually try to execute making your illustrations in the style that you like.

171
00:15:14,280 --> 00:15:18,600
And the thing is, if you try to make it too perfect, picking out mood boards and styles

172
00:15:18,600 --> 00:15:23,360
before you actually even try to execute it, you might find that, well, you wanted your

173
00:15:23,360 --> 00:15:28,920
illustrations in this style, but maybe you lack the execution ability or it's just not

174
00:15:28,920 --> 00:15:32,960
really working in the way you thought it would. It's good to try things really quickly and

175
00:15:32,960 --> 00:15:37,880
then find solutions. As I mentioned before, you can always hire another designer to make

176
00:15:37,880 --> 00:15:43,120
your illustrations in a specific style if you like the style that they have. A lot of

177
00:15:43,120 --> 00:15:47,840
times illustrators and designers can even work out this subnailing process for you,

178
00:15:47,840 --> 00:15:53,160
but I think it's fun to be engaged and even if this is not your main job, you want to

179
00:15:53,160 --> 00:15:58,600
take it to pen and paper first. That's going to be the easiest ways to get your creativity

180
00:15:58,600 --> 00:16:02,720
going and then feel like you had a say in the process. Now, when it comes to developing

181
00:16:02,720 --> 00:16:07,720
key art on a timeline, I feel that depending on the event you have, you want to give at

182
00:16:07,720 --> 00:16:12,960
least one to three months to your artist or designer to get this project done. If you

183
00:16:12,960 --> 00:16:19,360
have more graphics, I would say add on an additional week for every single image above

184
00:16:19,360 --> 00:16:24,640
the number three that you include. And I think that's fair just because genuinely, if you

185
00:16:24,640 --> 00:16:29,640
want the execution to be done correctly, the more you add on, the more time it takes, it's

186
00:16:29,640 --> 00:16:35,080
only fair. And if you ever wanted illustrations and graphics more quickly within a one to

187
00:16:35,080 --> 00:16:40,000
three month timeline, I would say that's when you start looking at images and graphics that

188
00:16:40,000 --> 00:16:44,400
are already made, that are already pre-done, that you can just buy the license to use.

189
00:16:44,400 --> 00:16:48,520
They exist out there and they're on websites like Creative Market. You could even venture

190
00:16:48,520 --> 00:16:53,920
to use icons on the noun project. There are lots of different solutions, but try not to

191
00:16:53,920 --> 00:16:58,760
rush an artist or a designer. Give them time. It takes time. If you're in a rush and you

192
00:16:58,760 --> 00:17:03,800
have this bigger vision, lower your expectations.

193
00:17:03,800 --> 00:17:08,240
So I have to be honest with you. I feel like I've said pretty much everything I can tell

194
00:17:08,240 --> 00:17:13,920
you about Key Art just by doing a voice audio workshop, but I did promise you that I was

195
00:17:13,920 --> 00:17:20,120
going to talk a little bit about AI art and generative AI today. And why do I have any

196
00:17:20,120 --> 00:17:25,240
authority to talk about this? I did work at an AI startup for about six months until it

197
00:17:25,240 --> 00:17:30,240
went defunct. At the time we were building a generative AI product, it was essentially

198
00:17:30,240 --> 00:17:35,600
what Adobe Firefly looks like right now. If you have no idea what I'm talking about, generative

199
00:17:35,600 --> 00:17:43,720
AI is essentially a text to visual AI product. It uses large language models to match words

200
00:17:43,720 --> 00:17:51,520
like dripping pink gooey rug on the beach to visual images from a visual library. And

201
00:17:51,520 --> 00:17:56,120
just from a string of words like that, you can generate images that take inspiration

202
00:17:56,120 --> 00:18:04,080
from the library AI has. You can add in your own image references and style references.

203
00:18:04,080 --> 00:18:09,800
There's actually a really good product available on the market called Adobe Firefly. It's currently

204
00:18:09,800 --> 00:18:15,520
free to use. And another reason why I have authority to talk on this topic is because

205
00:18:15,520 --> 00:18:21,240
I've used it quite a few times, especially working with my current agency, the Creative

206
00:18:21,240 --> 00:18:26,520
Exchange. Now I'm going to tell you what I think generative AI is good for, what it's

207
00:18:26,520 --> 00:18:32,340
not good for. And I truly believe this could be a longer episode. So I'm going to give

208
00:18:32,340 --> 00:18:38,400
you the TRDR, the too long don't read, or basically the cliff notes. Genit, I feel,

209
00:18:38,400 --> 00:18:43,440
especially for talking about Adobe Firefly, this is basically like a website that you

210
00:18:43,440 --> 00:18:50,120
can use. It will generate images for you. I think that the website itself still takes

211
00:18:50,120 --> 00:18:55,440
a lot of fine tuning to use. There are pretty good tutorials online. I'd say on TikTok,

212
00:18:55,440 --> 00:19:01,560
I found one by Pink Pony Creative that someone at my agency sent over to me. And I think

213
00:19:01,560 --> 00:19:06,320
what it's really good for is if you want to put in like a word, let's say that word is

214
00:19:06,320 --> 00:19:11,760
giveaway, and that is the structure that you want to generate, like a really cool dripping

215
00:19:11,760 --> 00:19:17,400
pink gooey style. You can put in your own vector arts and then change the style pretty easily,

216
00:19:17,400 --> 00:19:22,320
pretty quickly. You do have to tweak with the settings. And I've had a few experiences

217
00:19:22,320 --> 00:19:26,200
where maybe I put in a word that I wanted to change into a different style. Maybe it's

218
00:19:26,200 --> 00:19:31,600
like chromatic, fuzzy, whatever. But if you don't get the structure style down correctly,

219
00:19:31,600 --> 00:19:38,000
it can end up changing just entirely the word that you wrote out. So this work is not good

220
00:19:38,000 --> 00:19:43,020
on its own. I think it still needs a lot of massaging, as one of my design friends said.

221
00:19:43,020 --> 00:19:48,400
And the other thing too is I just think that you still need to have good design sensibilities

222
00:19:48,400 --> 00:19:55,400
to use these products. You still need to know different image types, like PNGs versus JPEGs,

223
00:19:55,400 --> 00:20:00,600
and what ratio of images are good to use. I think sometimes AI can actually generate

224
00:20:00,600 --> 00:20:05,960
some cool textures that, or like elements of like stylized words, for example, that

225
00:20:05,960 --> 00:20:11,240
are cool to cut out in Photoshop and then add to other designs because I'm a big Photoshop

226
00:20:11,240 --> 00:20:17,380
user. I love it. I love collage style. I think it's good for that, but I think on its own,

227
00:20:17,380 --> 00:20:22,960
it still looks AI generated to me. And I think that can be its own use. It's like going into

228
00:20:22,960 --> 00:20:28,760
a clip art folder or in like just, you know, it gives me more variety and ways of playing

229
00:20:28,760 --> 00:20:33,680
around with ideas, especially if you're not like a Photoshop user, you don't have the

230
00:20:33,680 --> 00:20:38,800
patience for Canva. But frankly, I'm not even going to tell you that this is a tool that's

231
00:20:38,800 --> 00:20:43,840
super easy to use. It takes some time. It takes some learning, but I do like it when

232
00:20:43,840 --> 00:20:48,480
visual tools are a little bit more accessible for people. But yeah, I think standalone as

233
00:20:48,480 --> 00:20:53,840
a product in itself, it's almost like a new category of visuals that I just don't think

234
00:20:53,840 --> 00:21:00,480
anyone can confidently say can stand on its own legs without some sort of experience,

235
00:21:00,480 --> 00:21:04,560
not even design experience, but like time with the product, I would say it takes 10

236
00:21:04,560 --> 00:21:08,880
hours to really get used to it, which isn't a super high investment. I would say Photoshop

237
00:21:08,880 --> 00:21:15,480
takes anywhere from 50 to a hundred hours just to get some usable range of actually

238
00:21:15,480 --> 00:21:21,560
creating some good outputs. Now, if you are a Photoshop or Adobe Suite user, you know

239
00:21:21,560 --> 00:21:26,840
that AI is really integrated into a lot of these products as well. So Illustrator, Photoshop,

240
00:21:26,840 --> 00:21:32,280
their generative fill backgrounds on Photoshop. I don't think these are as good as using

241
00:21:32,280 --> 00:21:39,360
Adobe Firefly. And the reason that is has to do a lot with basically how AI works. And

242
00:21:39,360 --> 00:21:46,320
I find like the best AI products actually has to do with some technical like CPU, computer,

243
00:21:46,320 --> 00:21:51,720
graphics card synergy that I fully even can't explain. Again, even at the AI startup, I

244
00:21:51,720 --> 00:21:57,320
was a product designer and I was designing the interface and I have like limited knowledge

245
00:21:57,320 --> 00:22:03,520
of the backend. So what I can say on a separate but related note is that a lot of companies

246
00:22:03,520 --> 00:22:09,600
are integrating AI into their features these days. Even Figma recently at config demonstrated

247
00:22:09,600 --> 00:22:16,840
some really cool uses of AI generating user flows of user interfaces. And they did say

248
00:22:16,840 --> 00:22:21,320
though, I will say this like very clearly that what they demonstrated on stage is like

249
00:22:21,320 --> 00:22:27,240
a private beta that they tested with some permission using other people's interfaces

250
00:22:27,240 --> 00:22:33,440
and whatnot. And the thing is what I know about AI and generative AI is it's actually

251
00:22:33,440 --> 00:22:40,600
extremely expensive. The computing cost of it is not cheap. So the reason why I feel

252
00:22:40,600 --> 00:22:45,160
like a lot of these big companies are coming out with like some like good products is because

253
00:22:45,160 --> 00:22:51,120
they can actually afford the cost of generating. But that's almost like barely. I think one

254
00:22:51,120 --> 00:22:55,960
thing that I learned from the AI startup fail is that the big players in the tech industry

255
00:22:55,960 --> 00:23:01,280
now are the ones that have the money and the resources to really do anything with generative

256
00:23:01,280 --> 00:23:06,960
AI. It's very resource expensive and yet no one's really figured out the monetization

257
00:23:06,960 --> 00:23:13,240
or revenue aspect. And I think even there's been jokes that they're just hoping that AI

258
00:23:13,240 --> 00:23:19,160
can figure out the cost revenue model. For this reason, I am highly skeptical and even

259
00:23:19,160 --> 00:23:27,040
wary of highly relying on generative AI and a design and visual process because what companies

260
00:23:27,040 --> 00:23:31,440
are essentially doing right now is giving us these products to use for free. They're

261
00:23:31,440 --> 00:23:37,000
doing the old excuse my metaphor drug dealer technique where like the first few times are

262
00:23:37,000 --> 00:23:41,440
free and then once you're hooked on it, they're going to start charging you like a $50 a month

263
00:23:41,440 --> 00:23:47,860
subscription. And you can see this even with a lot of really good tools that exist out

264
00:23:47,860 --> 00:23:53,640
there for like architects that may be leveraging some sort of AI like they're really charging

265
00:23:53,640 --> 00:23:59,360
$60 a month. And I think if you are in an industry where you need this, I wouldn't mind

266
00:23:59,360 --> 00:24:04,280
paying for that. You just include it in your cost. But I also think right now like I see

267
00:24:04,280 --> 00:24:09,880
designers using these free products and not really charging enough. Like generative AI

268
00:24:09,880 --> 00:24:15,360
is not supposed to be a cheap solution. Essentially what companies are doing right now is they're

269
00:24:15,360 --> 00:24:19,640
making it free, they're eating the cost. And once they find the market for it, they're

270
00:24:19,640 --> 00:24:24,200
going to start charging. So I think that's something to consider. I think it's funny

271
00:24:24,200 --> 00:24:30,000
because I see sometimes people like freelancing out generative AI services for work saying,

272
00:24:30,000 --> 00:24:34,600
oh, this is like a cheap and easy fast way of producing work. But genuinely, I think

273
00:24:34,600 --> 00:24:39,920
it actually is more work for designers sometimes to figure out how to make AI look good, especially

274
00:24:39,920 --> 00:24:43,960
with like things like Adobe Firefly. And then the second thing is no, it's absolutely not

275
00:24:43,960 --> 00:24:48,800
going to be free and cheap forever. This is just a way of getting people using the product

276
00:24:48,800 --> 00:24:53,920
and then maybe finding a way to rely on it before they start charging a lot for it. So

277
00:24:53,920 --> 00:24:59,360
that's a little bit of my cynical but like business perspective of generative AI. The

278
00:24:59,360 --> 00:25:04,240
other thing too is yeah, absolutely. I think it's a horrible idea to use other people's

279
00:25:04,240 --> 00:25:10,080
artwork in generative AI and rip on their style without any permission. I still think

280
00:25:10,080 --> 00:25:17,720
that is like copying and a lazy way of creating visuals that requires no sort of like talents

281
00:25:17,720 --> 00:25:22,840
or trust relationship, which as you've probably heard on this podcast, like I care about quite

282
00:25:22,840 --> 00:25:28,200
a lot. So I would say be careful how you use generative AI. And it's all fun in games to

283
00:25:28,200 --> 00:25:33,760
add in a vector word group of something like, oh, giveaway, because I want to do like a

284
00:25:33,760 --> 00:25:39,640
fun giveaway for a brand and just make the words giveaway look fun and match their style.

285
00:25:39,640 --> 00:25:45,800
But it's totally different. I think totally not cool in my opinion to take someone's illustrations,

286
00:25:45,800 --> 00:25:51,200
like copy their style, make a bunch of work that tries to replicate like basically knock

287
00:25:51,200 --> 00:25:56,880
off their work and like the years of expertise. I think that's foul play. And I also think

288
00:25:56,880 --> 00:26:03,820
people need to do research on how generative AI takes information, your information to

289
00:26:03,820 --> 00:26:10,120
learn human expectations based upon the info you give it. So if you give generative AI

290
00:26:10,120 --> 00:26:15,040
information that needs to be private, or you don't really want to share, that's too bad.

291
00:26:15,040 --> 00:26:21,120
You gave up the goods. So I think that's the gist of what I can offer on generative AI

292
00:26:21,120 --> 00:26:25,320
at the moment. But if you have any questions, if you think I said something like slightly

293
00:26:25,320 --> 00:26:29,880
inaccurate, you can go ahead and correct me. I don't mind. But these are my thoughts, my

294
00:26:29,880 --> 00:26:36,360
take. And I think after six months of being at a crypto startup that turned AI, I realized

295
00:26:36,360 --> 00:26:41,200
that the team I was working with just had very many issues and no product market fit

296
00:26:41,200 --> 00:26:45,760
to begin with. And that's one of the largest reasons that I was inspired to create this

297
00:26:45,760 --> 00:26:51,000
podcast because I truly feel that brands and founders need to have their product mission

298
00:26:51,000 --> 00:26:57,540
and their values down before they start hiring people. I think it's lazy and cheap moral

299
00:26:57,540 --> 00:27:03,600
character to just hire people to make things for you on the fly and chase trends. I think

300
00:27:03,600 --> 00:27:09,100
those types of ways of operating create products that have no backbone. And why should I want

301
00:27:09,100 --> 00:27:14,960
to stay and continue to work with you if you don't have any sort of strong moral character

302
00:27:14,960 --> 00:27:19,840
or work ethic or background just because you have money to pay me for a service that I

303
00:27:19,840 --> 00:27:25,080
could charge to literally anyone else and market on my own? It is my skills and expertise

304
00:27:25,080 --> 00:27:29,400
after all, is it not? So designers, I think you should think about using AI in a very

305
00:27:29,400 --> 00:27:35,360
similar sense. And that is not what I expected to say at the end of this podcast, but those

306
00:27:35,360 --> 00:27:40,840
are truly my thoughts and feelings and I will stand on that rock. All right. So this has

307
00:27:40,840 --> 00:27:47,800
been my episode on key art. My name is Jules. I am building out a lot of my websites this

308
00:27:47,800 --> 00:27:52,800
month. We're going to keep on talking about generating AI tools and resources for designers,

309
00:27:52,800 --> 00:27:57,400
brand and our education. I have some strong opinions, but like I think those are good

310
00:27:57,400 --> 00:28:02,160
to have, especially in the world where you could essentially do anything you want, do

311
00:28:02,160 --> 00:28:07,800
something you love, decide what you want to do and don't let other people choose for you.

312
00:28:07,800 --> 00:28:11,960
So this has been the pool party. Go ahead and visit my website, take a look at my services.

313
00:28:11,960 --> 00:28:16,280
I have a new illustration package out, send it for my newsletter on my website. That's

314
00:28:16,280 --> 00:28:23,320
J U L I A E S P E R O dot me and Julia Sparrow dot me. Maybe even add me on LinkedIn. And

315
00:28:23,320 --> 00:28:28,120
if you listen to this podcast episode, go ahead and give this episode a five stars,

316
00:28:28,120 --> 00:28:34,000
follow and listen wherever you listen to podcasts, maybe even share it with a friend. All right,

317
00:28:34,000 --> 00:28:46,800
folks, have a nice day and I'll talk to you next week.