WEBVTT
1
00:00:11.359 --> 00:00:15.480
So we're in I. I was just showing you, guys,
2
00:00:15.560 --> 00:00:22.399
Sea Bass and Mike this app from Adobe called Project
3
00:00:22.519 --> 00:00:27.160
Indigo that lets you use your iPhone to take pictures
4
00:00:27.160 --> 00:00:29.679
in just a really different way than has happened before,
5
00:00:29.800 --> 00:00:33.759
where you know, like live photos on the the on
6
00:00:33.840 --> 00:00:37.119
the iPhone, there's just it's basically just a little snippet
7
00:00:37.159 --> 00:00:39.119
of video and then you can just find your spot
8
00:00:39.200 --> 00:00:41.880
in it and make that the actual picture, and it
9
00:00:41.960 --> 00:00:43.759
kind of makes a decision for you, but you can
10
00:00:43.799 --> 00:00:46.960
override that. This thing takes a bunch of pictures and
11
00:00:47.039 --> 00:00:50.359
knits them together with AI to make the perfect version
12
00:00:50.359 --> 00:00:54.039
of the picture, which is crazy AI. You know, it
13
00:00:54.079 --> 00:00:56.759
does a lot of good things, but I read a
14
00:00:56.960 --> 00:01:03.840
bit yesterday about the problems with AI, and of course
15
00:01:03.880 --> 00:01:05.519
it's not going to open up for me.
16
00:01:05.680 --> 00:01:07.480
Do we want the perfect picture though?
17
00:01:08.040 --> 00:01:09.159
Oh ah?
18
00:01:09.200 --> 00:01:15.120
Sometimes yeah, real estate, real estate, but like humans, you know, yeah, true,
19
00:01:16.319 --> 00:01:19.439
Like we're just getting further away from like capturing a moment.
20
00:01:19.519 --> 00:01:21.959
I think, yeah, well, which is the point, right?
21
00:01:22.040 --> 00:01:25.200
Yeah? I think that. I mean some of these pictures,
22
00:01:25.239 --> 00:01:26.680
as great as they look, have a little bit of
23
00:01:26.680 --> 00:01:29.760
an uncanny valley quality to them. Yeah, there's like forty
24
00:01:29.799 --> 00:01:31.400
eight of them on the website and you can go
25
00:01:31.439 --> 00:01:32.840
through them, and some of them look like, oh, that
26
00:01:32.840 --> 00:01:34.680
looks really cool and normal and whatever, and the other
27
00:01:34.719 --> 00:01:36.560
one looks this looks cooked.
28
00:01:36.480 --> 00:01:41.359
Yeah, like a landscape, real estate, a car, like objects obviously,
29
00:01:41.400 --> 00:01:43.359
but like, yeah, if you're taking it like a sports
30
00:01:43.400 --> 00:01:44.519
game or something, you.
31
00:01:44.439 --> 00:01:45.480
Want a little grain in there.
32
00:01:45.560 --> 00:01:47.040
Yeah, and you want to like feel it a little bit.
33
00:01:47.079 --> 00:01:49.000
And I feel like the only way you feel things
34
00:01:49.079 --> 00:01:52.000
is with flaws, Like you don't feel perfection. Yeah, this
35
00:01:52.079 --> 00:01:52.359
is true.
36
00:01:52.599 --> 00:01:56.359
That's why. That's actually one of the keys to making
37
00:01:56.400 --> 00:02:00.719
something esthetic is having the in a flaw.
38
00:02:00.799 --> 00:02:03.079
Yeah, something right or even yeah, like in tiar design.
39
00:02:03.400 --> 00:02:05.599
I'm never taking courses, but I imagine you don't want
40
00:02:05.599 --> 00:02:08.360
everything perfectly symmetrical. It's just like it's just that your
41
00:02:08.360 --> 00:02:10.080
eyes don't like that. It feels weird.
42
00:02:10.319 --> 00:02:15.240
The Catholic church edict yeah that churches and cathedrals cannot
43
00:02:15.319 --> 00:02:15.919
be perfect.
44
00:02:17.479 --> 00:02:18.719
So yeah, it just makes more sense.
45
00:02:18.759 --> 00:02:20.680
And there's like more windows on one side than the other.
46
00:02:20.919 --> 00:02:22.560
And we're already sucking that up right, like because there's
47
00:02:22.599 --> 00:02:26.439
already photoshop and like face tune. Like the more I
48
00:02:26.479 --> 00:02:29.280
see it on Instagram, it's insane, Like I try not
49
00:02:29.360 --> 00:02:33.919
to pay attention, and luckily it's not really my career anymore,
50
00:02:33.919 --> 00:02:36.080
although it's adjacent to everything we do and a part
51
00:02:36.120 --> 00:02:41.639
of what we do. But like scrolling around and so
52
00:02:41.719 --> 00:02:44.520
my shit's more cars and you know, finance shit I
53
00:02:44.560 --> 00:02:46.560
like to read, which is great rather than what I
54
00:02:46.599 --> 00:02:48.680
have to create. But scrolling around, I still see a
55
00:02:48.719 --> 00:02:51.319
bunch of fitness stuff and it's just like so apparent
56
00:02:51.400 --> 00:02:54.960
from so far away that every part of every human
57
00:02:55.039 --> 00:02:57.319
has been smoothed out. I'm like, you guys don't have
58
00:02:57.400 --> 00:03:01.560
like hair, you don't have a single wrinkle, you know,
59
00:03:01.840 --> 00:03:04.199
like literally some of these people, I think if I
60
00:03:04.199 --> 00:03:06.000
saw him in person, I wouldn't recognize them, Like it's
61
00:03:06.680 --> 00:03:09.400
that extreme. And so we get the perfect picture and
62
00:03:09.439 --> 00:03:11.280
then we're gonna get the perfect face tune on top
63
00:03:11.319 --> 00:03:14.800
of that, like we're literally cook. Well, there'll there just
64
00:03:14.840 --> 00:03:18.000
be no room for a creator. Yeah right, it'll just
65
00:03:18.080 --> 00:03:20.400
be it'll just make it AI. It already is right,
66
00:03:20.439 --> 00:03:22.759
And then we were talking all fair about the stormtroopers
67
00:03:22.759 --> 00:03:26.800
and the Bigfoot and the monkeys like it's so good yeah,
68
00:03:26.919 --> 00:03:30.479
and it's funny yeah, and like the prompter, the writers
69
00:03:30.800 --> 00:03:33.479
will only get better and who knows, twenty years from now,
70
00:03:33.520 --> 00:03:35.680
they might be obsolete, right, AI might be able to
71
00:03:35.840 --> 00:03:42.159
algorithmically write something funny that, right, But yeah, I think
72
00:03:42.159 --> 00:03:44.000
it's gonna change shit so fast, and we've already I
73
00:03:44.000 --> 00:03:46.439
think talked about in the past where part of the
74
00:03:46.560 --> 00:03:50.240
huge strike in acting was like many things, I think
75
00:03:50.479 --> 00:03:54.159
we're on the list, but like using people's face just
76
00:03:54.240 --> 00:03:56.120
once and then being able to use it forever, right,
77
00:03:57.639 --> 00:04:01.280
and so like it's cool and I love tech, but
78
00:04:02.159 --> 00:04:04.919
part of what makes content cool. And another reason I
79
00:04:05.759 --> 00:04:09.360
almost purely watch live streams now is because you can't
80
00:04:09.400 --> 00:04:12.479
hide the authentic. Yeah, you know, podcasts are similar in
81
00:04:12.479 --> 00:04:15.199
some senses, although you know, maybe thirty percent or edited
82
00:04:15.280 --> 00:04:17.399
or who knows, maybe more now because we don't know
83
00:04:17.399 --> 00:04:19.519
all these people we don't, you know, because in the
84
00:04:19.519 --> 00:04:21.959
beginning I thought YouTube was fully authentic, and then I
85
00:04:21.959 --> 00:04:24.120
started doing YouTube collapse and I see how many takes
86
00:04:24.120 --> 00:04:27.639
and edits people do. Yeah, and obviously YouTube's elevated now
87
00:04:27.720 --> 00:04:30.040
mister Beast is making TV shows.
88
00:04:29.920 --> 00:04:34.040
You know, Mister Beast is rumored to be buying the Angels.
89
00:04:34.079 --> 00:04:36.319
No, I don't doubt it, the Angels. I don't doubt it.
90
00:04:36.319 --> 00:04:37.399
He's got so much money.
91
00:04:37.639 --> 00:04:39.000
Yeah, I don't know if it's really gonna hap. I
92
00:04:39.040 --> 00:04:43.000
don't know that the ownership. You know that the owner's
93
00:04:43.000 --> 00:04:44.199
committee would actually approve.
94
00:04:44.360 --> 00:04:46.199
I think they should. I actually think it would be
95
00:04:46.240 --> 00:04:49.319
their smartest play ever. Mister pieces go crazy with that.
96
00:04:49.319 --> 00:04:52.120
Probably he would probably implement like game. Imagine the content
97
00:04:52.199 --> 00:04:53.920
he would create that would promote that. Like he's the
98
00:04:54.000 --> 00:04:56.120
number one marketer on the planet. You know, you can
99
00:04:56.240 --> 00:04:58.519
listen to all the hormoses and the Gary V's, but
100
00:04:58.560 --> 00:05:01.199
he's the best at getting it tension, and he would
101
00:05:01.199 --> 00:05:03.079
make authentic. That's what he still does. All though his
102
00:05:03.120 --> 00:05:06.759
shit's edited, it's still authentic from the heart, Like he's
103
00:05:06.800 --> 00:05:09.720
a creator and he's the creator's creator, and so like,
104
00:05:09.800 --> 00:05:11.480
I bet you he would make their shit pop.
105
00:05:11.519 --> 00:05:13.439
If you're the MLB, you probably like step in like
106
00:05:13.519 --> 00:05:13.800
a bro.
107
00:05:13.959 --> 00:05:16.199
The MLB might stop it before the Angels. I mean
108
00:05:16.199 --> 00:05:18.000
the Angels might too because they're old school. But if
109
00:05:18.160 --> 00:05:19.639
I think if they had a brain and want to
110
00:05:19.720 --> 00:05:22.199
make cool shit, mister Beasts is the first guy goes
111
00:05:22.199 --> 00:05:22.720
to for sure.
112
00:05:23.000 --> 00:05:27.120
I think the problem is that the on field performance
113
00:05:27.160 --> 00:05:28.879
has to be up.
114
00:05:28.759 --> 00:05:31.439
To of course before well, I don't know. The bananas
115
00:05:31.439 --> 00:05:34.160
say different but well that's because they're fun.
116
00:05:34.759 --> 00:05:35.839
The bananas are fun.
117
00:05:36.000 --> 00:05:37.279
Yeah, you know what, I'm just saying.
118
00:05:37.279 --> 00:05:39.879
There's anything fun about watching the Angels most of the time.
119
00:05:39.800 --> 00:05:41.560
For sure. But you could do a combo both, right,
120
00:05:41.600 --> 00:05:43.639
Like you up revenue somehow, then maybe you can get
121
00:05:43.639 --> 00:05:47.480
better players or maybe the next generation. Shout on my
122
00:05:47.519 --> 00:05:50.800
boy agent. We went to the NBA playoff game one
123
00:05:50.920 --> 00:05:52.879
every person and he got to go to the media. Yeah,
124
00:05:52.920 --> 00:05:55.000
and he's next to ESPN, all these guys, all the
125
00:05:55.000 --> 00:05:57.879
players knew who he was. So you don't think if
126
00:05:57.879 --> 00:05:59.920
you're an up and coming twenty five year old pitcher,
127
00:06:00.240 --> 00:06:02.720
not fuck that I'm trying to play for mister Beast. Yeah,
128
00:06:02.800 --> 00:06:04.600
you know, I think I just don't see a negative
129
00:06:05.319 --> 00:06:08.000
Like the memes, bro the memes and they probably watch
130
00:06:08.040 --> 00:06:10.040
this stuff. They probably legitimately watch this stuff.
131
00:06:10.160 --> 00:06:12.560
Well you know that that MLB just bought a piece
132
00:06:12.600 --> 00:06:18.839
of John Boy media, like because nobody was making better
133
00:06:19.839 --> 00:06:25.680
like true baseball content in a social media format.
134
00:06:25.720 --> 00:06:29.240
Smart. I think a lot of the big leagues are
135
00:06:29.240 --> 00:06:31.120
still behind. I think we've talked about this too in
136
00:06:31.160 --> 00:06:33.959
media versus. Like UFC's doing a really good right now
137
00:06:33.959 --> 00:06:36.079
where they hired this chick Nina who does all like
138
00:06:36.319 --> 00:06:40.040
very authentic TikTok style interviews, and it just feels so real.
139
00:06:40.720 --> 00:06:42.839
And obviously it's more complicated because there's many teams in
140
00:06:42.879 --> 00:06:45.639
the league versus a fight night once a month and
141
00:06:45.800 --> 00:06:49.079
you only have ten fighters interview. But still that's what
142
00:06:49.079 --> 00:06:51.560
people are longing for. I think, you know, as TV
143
00:06:51.639 --> 00:06:54.360
and movies have gotten so out there, has everything on,
144
00:06:54.720 --> 00:06:58.000
it just feels so fake. Being able to like just
145
00:06:58.040 --> 00:07:00.920
talk to the creators normal was so it makes a
146
00:07:00.959 --> 00:07:01.519
lot of sense.
147
00:07:01.720 --> 00:07:05.319
Yeah, so back to AI. This is the thing that
148
00:07:05.360 --> 00:07:11.000
I encountered yesterday. It is MIT completed a study on
149
00:07:11.279 --> 00:07:17.319
a brain brain scan study on Chat GBT users, So
150
00:07:17.519 --> 00:07:20.600
four months of data and here's what they said. Eighty
151
00:07:20.639 --> 00:07:23.199
three percent of people who used chat GPT to write
152
00:07:23.600 --> 00:07:27.839
couldn't remember anything they wrote just minutes later. Only eleven
153
00:07:27.920 --> 00:07:32.360
percent forgot what they used Google or their brain to
154
00:07:32.399 --> 00:07:39.959
come up with to write. Neural connections had a forty
155
00:07:40.000 --> 00:07:45.480
seven percent collapse. Like that's like if your laptop lost
156
00:07:45.600 --> 00:07:46.360
half its power.
157
00:07:47.399 --> 00:07:47.720
Damn.
158
00:07:49.720 --> 00:07:52.360
Human brain, when left to think for itself, fires on
159
00:07:52.399 --> 00:07:57.480
all cylinders. In contrast, large language model assisted writing risks
160
00:07:58.839 --> 00:08:04.399
risks leaving that engine idle teachers called them soulless, perfect
161
00:08:04.399 --> 00:08:07.759
grammar and zero personality, no depth, nuanced and recycled structure.
162
00:08:08.240 --> 00:08:12.120
But the worst of it is that when chat GBT
163
00:08:12.439 --> 00:08:15.360
users were forced to write alone, their brains stayed flat,
164
00:08:15.879 --> 00:08:19.600
their essays got worse, their mental muscles didn't recover. Meanwhile,
165
00:08:19.639 --> 00:08:24.279
those who wrote without AI finally tried chat GBT, their
166
00:08:24.319 --> 00:08:27.079
brain sparked at first and then settled into a steady,
167
00:08:27.199 --> 00:08:31.680
active rhythm. That's a cognitive debt. When the bot does
168
00:08:31.720 --> 00:08:34.679
the thinking, your brain stops training and you need it
169
00:08:35.000 --> 00:08:39.759
when you need it most, and it might not respond.
170
00:08:40.159 --> 00:08:44.159
At Oregon State, students use chat gbt just to brainstorm,
171
00:08:44.200 --> 00:08:46.799
but wrote their own words and produce their own creative work,
172
00:08:49.159 --> 00:08:52.279
and they produced more creative work than people who didn't
173
00:08:52.360 --> 00:08:56.480
use AI. But in Canada, two large studies showed people
174
00:08:57.240 --> 00:09:00.679
who were more creative with AI, but a weekly without
175
00:09:00.720 --> 00:09:03.279
it performed worse than before.
176
00:09:03.759 --> 00:09:06.559
I wonder how they're judging creativity. I don't know, because
177
00:09:06.600 --> 00:09:10.559
like you know, like like I guess at its roots,
178
00:09:10.559 --> 00:09:14.440
creativity's original thought, our original creation. But yeah, how the
179
00:09:14.480 --> 00:09:17.000
fuck do you judge that? Because even the IQ test
180
00:09:17.039 --> 00:09:20.279
you can argue, right, like, isn't the best way to know,
181
00:09:21.240 --> 00:09:22.559
so it's all so difficult.
182
00:09:22.879 --> 00:09:26.000
Yeah, well, I mean they did this with brain scans, so.
183
00:09:26.440 --> 00:09:29.039
Yeah, that makes it even harder for me to comprehendively.
184
00:09:29.120 --> 00:09:32.879
Yeah, I mean, well obviously it's it's more quantifiable, yeah,
185
00:09:32.919 --> 00:09:34.200
for sure than what they're doing.
186
00:09:34.440 --> 00:09:36.559
But yeah, yeah, I guess the creative part of your
187
00:09:36.559 --> 00:09:40.399
brain's more activated. But it doesn't mean you're more creative
188
00:09:40.440 --> 00:09:43.840
per se, because it's probably versus yourself, I would imagine, right,
189
00:09:43.840 --> 00:09:46.279
because it's probably verse your baseline. And I think that's
190
00:09:46.480 --> 00:09:49.720
two things. One did you see the semi viral some
191
00:09:49.799 --> 00:09:52.600
kids graduating because all these graduations are going on, Yeah,
192
00:09:52.960 --> 00:09:54.840
and he showed is he had an iPad in the
193
00:09:54.840 --> 00:09:56.919
crowd with his little fucking tassels and all that he's
194
00:09:56.960 --> 00:09:59.519
about to graduate. And he showed on his iPad like
195
00:09:59.559 --> 00:10:03.240
his thesis that was wrote by Chap gbt's basically proving
196
00:10:03.279 --> 00:10:05.960
that he cheated to get his degree with that, which
197
00:10:06.000 --> 00:10:10.480
is fucking wild to me. AI's it so interesting because
198
00:10:11.120 --> 00:10:14.039
even this study, if they ran it like six months apart,
199
00:10:14.080 --> 00:10:17.080
the one in Canada versus here, AI is already so different, right,
200
00:10:17.200 --> 00:10:19.399
Chat GBT two versus three and four is coming out
201
00:10:19.480 --> 00:10:22.960
or whatever, So it's like hard to get a baseline there. Two,
202
00:10:23.039 --> 00:10:27.200
I see it all just like the Internet, Like, oh,
203
00:10:27.240 --> 00:10:29.639
the Internet makes you dumber. Well yeah, man, if you're
204
00:10:29.960 --> 00:10:32.879
like scrolling tiktoking on porna hub all day, you're dumber.
205
00:10:33.320 --> 00:10:36.320
But like I'm researching and now I can not go
206
00:10:36.399 --> 00:10:38.480
to the library and read a thousand books, and I
207
00:10:38.519 --> 00:10:40.440
can get my answer quick and move on to the
208
00:10:40.440 --> 00:10:42.320
next answer, move on to the next answer and kind
209
00:10:42.320 --> 00:10:45.120
of erase the fluff and not again, not like trying
210
00:10:45.120 --> 00:10:46.519
to say I'm some genius, but that's how I use
211
00:10:46.600 --> 00:10:49.600
chat GPT. Also, yeah, I mean saybot's use it for
212
00:10:49.639 --> 00:10:51.159
a lot of our captions. Hey, give me a couple
213
00:10:51.200 --> 00:10:53.240
of captions and then we rip pull it out kind of
214
00:10:53.240 --> 00:10:55.000
like they said, rip it apart to make it more
215
00:10:55.360 --> 00:10:58.240
us and then put it and post it. Like why
216
00:10:58.240 --> 00:11:02.399
wouldn't you use it in those manners versus Yeah, like
217
00:11:02.799 --> 00:11:07.120
cheating or like footnoting probably will be obsolete, Like there
218
00:11:07.159 --> 00:11:09.919
probably won't be a thing right because how do you
219
00:11:09.960 --> 00:11:11.840
regulate it, which you're already having issues with.
220
00:11:14.519 --> 00:11:16.559
Well, I a couple of weeks ago, when I had
221
00:11:16.639 --> 00:11:22.519
josh On, I used a AI program to rip a
222
00:11:22.559 --> 00:11:25.600
bunch of clips and I posted one of them. It
223
00:11:25.759 --> 00:11:29.879
was so much less work. I mean, the amount of
224
00:11:29.879 --> 00:11:33.399
work was really just like going through the clips that
225
00:11:33.399 --> 00:11:36.240
looked like they were most likely and it's supposed to
226
00:11:37.360 --> 00:11:40.320
This was the free version and so there's no like
227
00:11:40.440 --> 00:11:44.519
virality score, but now like I actually subscribe to it,
228
00:11:44.559 --> 00:11:46.840
and now there's like a virality score.
229
00:11:46.639 --> 00:11:48.679
When it just makes sense because like, first you would
230
00:11:48.720 --> 00:11:50.679
have to do it yourself in twenty ten and that's
231
00:11:50.679 --> 00:11:52.399
going to take forever. Yeah, right, if you're going to
232
00:11:52.440 --> 00:11:54.399
look through an hour long podcast and come up with
233
00:11:54.519 --> 00:11:57.080
three clips that you think are good, then edit.
234
00:11:56.960 --> 00:11:59.600
Those clips into vertical format.
235
00:11:59.320 --> 00:12:04.320
Rightly that, yeah, and then you know, twenty sixteen, twenty seventeen,
236
00:12:04.360 --> 00:12:06.519
you're going to we work and you're outsourcing it around
237
00:12:06.559 --> 00:12:08.279
the world, but that's still gonna cost you fifty to
238
00:12:08.279 --> 00:12:11.279
one hundred and then like do they know the virality
239
00:12:11.320 --> 00:12:13.320
better than you? And the truth is, like some people
240
00:12:13.320 --> 00:12:15.559
are really good at it. Obviously again, the mister Beasts
241
00:12:15.600 --> 00:12:18.080
and shot Out Ludwig and some of these creators are
242
00:12:18.120 --> 00:12:21.679
just so tapped into you know, human entertainment. But like
243
00:12:22.559 --> 00:12:25.399
to me, that again isn't like forging or cheating that's