aiLL 10/8/23 - ChatGPT's New Secret Sauce and What to do With It

Description

Chapters

Transcript

0:03 [Music]
0:35 good evening everybody happy
0:41 [Music]
0:59 Sunday
1:01 [Music]
1:03 all right let's get going let's learn
1:07 about AI enough with crappy acoustic
1:09 music welcome
1:11 everybody to the AI learning
1:14 lab let's get learning some AI in a lab
1:19 that looks like a home
1:22 office we can do it so if you'd be so
1:26 kind share the live and prove that
1:29 you're not
1:31 normal let your friends make fun of you
1:33 oh you're into that AI stuff really
1:36 really is that what we're doing now
1:37 we're we're going to we're going to play
1:39 with AI we're going to ruin all of the
1:41 world is that that's what we're going to
1:43 do
1:44 now you know it's not environmentally
1:46 sound you say you're for the climate now
1:48 you're using artificial intelligence do
1:50 you know how much processing power it
1:51 takes to to make an AI it's a lot okay
1:54 it's a
1:57 lot Silver Fox what's happening
2:01 WooHoo for the early birds I
2:03 know I know I know I know welcome
2:09 everybody wo what are we going to do
2:12 tonight what are we going to do tonight
2:14 we've been having some some good good
2:16 deep conversations the past few nights
2:20 hular take good
2:23 evening how you doing Becky oh oh hey
2:29 Mark
2:33 it's teaka tea how you doing teaka you
2:36 good are you
2:40 good 60 Minutes had a story tonight on
2:43 AI very cool I did not see that but I'm
2:47 sure I
2:48 will I'll at least see pieces of it on
2:51 the Tik
2:55 Tock I watched the F1 race this
2:58 morning well kind of this after
3:00 afternoon didn't start till like 11:00
3:02 man that was rough they they were all
3:04 about passing out in their cars one guy
3:06 retired cuz he got too hot Louis
3:07 Hamilton smacked into his teammate it
3:09 was a lot of drama a lot of
3:13 drama oh
3:15 man all right we'll get we'll get let
3:19 some people get in
3:21 here welcome
3:24 everybody we'll get started here
3:27 shortly I sleep on my shoulder weird and
3:30 as I get older it's it's just a big
3:33 crunchy mess it's like nothing works
3:37 anymore it's like that that old Lou CK
3:40 bit where he talks about going to the
3:41 doctor doctor my ankle hurts and
3:43 doctor's like moves it around yeah it's
3:45 just worn
3:47 out can't you do something about it
3:53 Noah oh only got to see seconds of it
3:55 while taking care of the kids at night
3:57 yeah I actually watched most of the uh
3:59 the the
4:00 uh the F1 race in
4:03 Walmart we were shopping for crap and I
4:07 got flu shot we were shopping for crap
4:11 and I'm just watching it um I remember
4:14 when my bones and joints started making
4:15 noise that I could hear and it surprised
4:18 me there's there's another comedian I
4:20 forget who it is but he talks about you
4:22 know he hit that age where you just make
4:24 noises getting
4:29 up
4:31 like you accomplish
4:34 something cuz you did you're still
4:37 vertical um leaving for Europe and don't
4:40 know how I'll survive without the
4:42 workshop for two weeks ah I'm sure I'm
4:46 sure you can listen listen to the uh to
4:49 the lab while you're in uh while you're
4:52 in
4:53 Europe be at the Eiffel
4:57 Tower sh hang on Kyle's saying something
5:01 you're weirdos you're all weirdos we're
5:04 not weird we're
5:05 Irregulars um all right looks like we're
5:08 starting to get a bunch of people in
5:09 welcome everybody uh this is the AI
5:10 learning lab my name is
5:13 Kyle my involvement with this AI stuff
5:16 is is relatively new kind of the past
5:19 two years or so and this channel is
5:22 really about the moment that changed
5:25 everything in my opinion which is the
5:27 launch of chat GPT so November 30th 22
5:30 is the inflection point from how I look
5:32 at things that everything changes um
5:36 because that's the date at which the
5:38 rest of us get access to these really
5:40 powerful machine learning tools and it's
5:43 very very reminiscent of the early days
5:45 of the worldwide web and I was I was an
5:47 entrepreneur back then I've been an
5:48 entrepreneur for decades so I'm
5:50 unhirable at this point I can only start
5:54 new
5:55 [ __ ] it's the only place that will have
5:58 me uh is is me
6:02 me um but the the early days of the web
6:06 had a very very similar energy as to
6:08 what's going on with this uh with this
6:09 AI stuff so that's what this Channel's
6:11 about I'm trying to do a video montage
6:13 for a song on YouTube any AI tool you
6:16 would
6:17 recommend what is that one
6:22 called some of the Irregulars were
6:24 playing with it um there's there's one
6:26 that you can you basically just type in
6:28 a description and it actually pulls
6:30 stock um stock
6:32 video um and stitches it together for
6:35 you it it it makes it look a bit like a
6:37 corporate kind of video but if you just
6:39 want to knock something out um you could
6:42 do that the other thing you could play
6:44 with is Runway ml I'll go through some
6:46 tools here in a second um you could play
6:49 with Runway
6:55 ml app. Runway ml.com that one that one
6:59 right there
7:01 and that will do text to video if you
7:02 want to create your own custom video you
7:05 can do little 4 second Snippets there
7:07 and you can extend them up to like 18
7:09 seconds so if you want like you know a
7:12 woman in a wispy dress walking through a
7:14 field you know with the sun shining you
7:17 can just ask for it and it'll go make it
7:20 um if anyone remembers the name of that
7:23 that video tool that you uh you describe
7:26 your video and then it it pulls stock
7:28 stock video for you I forget what I just
7:31 forget what it's called um anyway Let's
7:34 uh let's get started here got got a few
7:36 people in if you'd be so kind as to
7:38 share the live uh out with the world and
7:42 tap on the screen even though you may
7:44 not like me yet yet you hear that
7:47 optimism you're going to love it here
7:51 it's a little odd we got these folks
7:53 called it regulars they hang out here
7:54 all the time it's turned into a whole
7:56 Community we we collaborate we joke we
7:59 we tap the screen a lot um and uh and we
8:04 talk about this AI stuff and we're kind
8:06 of learning together and even though I'm
8:09 the host the what what I can promise you
8:13 is that I'm not an expert and if you ask
8:14 for my qualifications I have none and if
8:17 anyone tells you they do right now
8:19 they're full of [ __ ] nobody knows nobody
8:22 knows what's Happening Now does that
8:24 mean that there are am I saying there
8:27 are no experts in machine learning and
8:29 artificial intellig no no there's lots
8:31 of people in that but in this post
8:34 inflection point post GPT how are we as
8:37 businesses and human beings going to use
8:40 large language models and image
8:42 generation tools and music generation
8:43 tools how are we going to incorporate
8:45 this stuff into our lives nobody knows
8:47 what's going on with that nobody zero
8:50 people out of 100% of people 0% have any
8:56 much more of a clue than anyone else
8:59 it's very very uh we're very very early
9:03 and so this channel is just like uh
9:09 exploring um gmil Singh yes you said it
9:13 right I didn't like you first but but
9:16 like watching your lives
9:18 nowadays thank you for your honesty
9:21 listen when I got in here I thought you
9:22 were a pompous ass I couldn't stand you
9:25 as a matter of fact and the look of you
9:27 good Lord who are you Rosy old Don
9:30 what are you trying to be Alec Baldwin
9:32 but fat good Lord you are insufferable
9:35 but now you're kind of charming I kind
9:38 of like
9:39 it ah top of the morning to
9:43 you all right let's uh let's let's dig
9:46 in here so if you're brand spanking new
9:49 to this AI
9:52 stuff I will I will give you a place to
9:55 start okay
9:58 okay yeah we're going to need you to go
10:01 on ahead and come on in on Saturday yeah
10:05 okay so if you're new to this AI stuff
10:09 that top URL right there just go there
10:11 right now don't don't pause don't pass
10:14 go don't collect $200 just go there
10:18 right
10:19 now that's chat GPT if you've not played
10:22 with chat GPT you need to go play with J
10:24 GPT why cuz you do cuz you do cuz you
10:27 just do cuz everything has changed but
10:31 not that many people know it
10:33 yet most jobs are going to be redefined
10:36 over the next probably two to three
10:39 years but let's call it three to five
10:40 it's coming very very
10:42 fast so everyone owes it to themselves
10:46 to go there and start playing and quite
10:48 frankly I think it's even worth going
10:51 there and paying the 20 bucks to get
10:52 their more advanced version cuz it's got
10:54 a bunch of features that are coming out
10:57 I found out is it visa yes
11:00 Visa um the person that was asking about
11:03 what AI can I use to make my video
11:05 montage Visa I think it's Visa AI hang
11:08 on let me go let me go look let me go
11:12 look at it
11:14 visla
11:19 AI oh God visa.
11:22 us everyone's got anything but a do
11:26 these days you can't remember anything
11:28 visa. us you just describe your video
11:33 and it will go generate um a bunch of
11:40 uh stock video clips and Stitch them
11:43 together it's pretty cool it's pretty
11:45 cool um I did I did one um I did one on
11:49 the live here a few nights back of like
11:51 someone a fly fisherman fishing on the
11:53 Delaware and for the most part it got
11:54 the footage right it was it was
11:57 definitely the Delaware for most of the
11:58 shots it's pretty good
12:01 okay now these other three like so why
12:04 do I have four up here because these
12:06 tools right now are very very uh
12:10 not not complete is the best way to put
12:13 it these four are kind of your big
12:15 players you got chat GPT you got Google
12:18 bard that's going to get better much
12:20 better soon you got Claude they just got
12:23 a 4 billion commitment with a $1.3
12:26 billion investment from Amazon and then
12:29 Bing is Microsoft and they invested $10
12:31 billion in those guys so these four are
12:34 major players not going anywhere also
12:37 major players probably not going
12:39 anywhere but less
12:41 major uh inflection is the company
12:44 behind Pi I would not be surprised if
12:47 Apple didn't buy these guys they just
12:49 raised $1.3 billion so Apple's going to
12:52 have to pay up but they can afford it
12:54 and the IOS app of Pi you can talk to it
12:57 you put it in talk mode and you can
12:59 actually have a conversation with these
13:01 GPT models it's mind-blowing well I
13:03 don't know if it's mind-blowing it is a
13:05 glimpse into the future imagine if Alexa
13:09 and Siri didn't suck that's what that is
13:13 now it can't take action for you like
13:15 unlike Alexa it cannot order you [ __ ]
13:18 from Amazon or Siri it can't you know
13:22 download music you can't take actions on
13:24 your behalf but just in terms of being
13:26 able to talk to it about [ __ ]
13:28 anything it's amazing
13:30 uh perplexity doai really good site
13:32 labs. perplexity is a free hosted
13:35 version of meta's llama models if you
13:37 want to play with those Winston are you
13:39 in the house llama he said llama llamas
13:43 is tasty um PO is uh is another site
13:48 where it's a single interface where you
13:49 can play with lots of different models
13:51 like inflection and uh or uh uh
13:55 anthropics Claud models and open AI
13:58 models and there's Bots there that you
13:59 can make you can make your own Bots
14:01 pretty
14:03 cool um I noticed chat GPT app doesn't
14:06 show me the same beta features as
14:08 browser still no it doesn't um and still
14:11 no Dolly 3 yeah the the the iOS chat GPT
14:15 app and I assume the
14:18 uh the uh Android one um the features
14:23 are not in parody with the website so I
14:26 actually have on my phone I mostly use
14:28 the website I just open Safari I have it
14:32 bookmarked um and that and that way I
14:34 get access to all those features um
14:38 apparently the talking the talking to
14:40 chat GPT in the app as well
14:44 as uh image
14:48 uh chat GPT Vision chat gptv where it
14:52 can read images apparently both of those
14:54 things are going to be in the apps by
14:56 Tuesday and then the dolly 3 stuff
14:59 within a month I heard that I was I was
15:02 hanging out in the um open AI Discord
15:06 this afternoon and I heard that I I
15:09 asked explicitly when when we're going
15:11 to get um GPT vision and they said by
15:15 Tuesday um I found Android PI app called
15:19 ask very enjoyable cool very
15:23 nice um okay so that's that if you want
15:27 to make images go ahead and screen
15:29 screenshot this so doll or if you go to
15:31 bing.com that's actually dolly3 the
15:34 image generation engine behind it it's
15:36 really good I also have Early Access to
15:39 dolly3 inside chat GPT which is very
15:42 very cool um Mid journey is from a
15:46 quality standpoint is the best in terms
15:48 of these image generation tools ideogram
15:51 can do text but so can Dolly 3 now so
15:54 that's not as important as it was even a
15:56 month ago just that's how f shit's
15:59 moving I'm probably going to Lop that
16:00 one off the list in fact um and then uh
16:05 Firefly Adobe the Adobe products with
16:08 this AI stuff are getting good and
16:11 better and they're also
16:13 commercially less risky if you work in a
16:15 big business the Adobe generative AI
16:17 stuff is probably the way to go versus
16:20 mid Journey or stable diffusion or even
16:22 Dolly
16:23 3 um here's some of the video tools so
16:27 Runway ml is text video Ha genen Labs is
16:31 synthesized voice and animated face as
16:34 well as video translation that is quite
16:37 remarkable I can show you that it's it's
16:39 mindblowing in fact let me show you that
16:41 now because why
16:49 not let me see let's see how did I put
16:54 it did I
16:55 put
16:57 Mandarin
17:05 I did not did I put
17:10 Chinese
17:13 no all
17:15 right so that's going to be in
17:27 downloads
17:29 well here's
17:35 one
17:44 n she does not speak
17:47 Japanese
17:49 um let me see if I
17:57 have
17:59 [Music]
18:05 I saw at Medical
18:08 help2 that's how she
18:17 talks so that's hey Jen it resyncs
18:21 people's lips it synthesizes their voice
18:24 translates the language syns their voice
18:27 resyncs the lips and puts it all back
18:30 together and it does it in like 10
18:31 minutes so that's what that is um
18:34 blockade Labs is text to 3D World did is
18:38 another voice synthesis tool and then
18:41 here's some resources that are just good
18:43 to know good people to follow good uh
18:47 app directory so futur
18:57 pedia.com app but if you want to find
19:00 something specific it probably exists
19:03 the problem with most of the apps are is
19:04 they're kind of single
19:06 use crappy tools just built on top of
19:09 chat GPT or something like it um and
19:11 they're going to charge you 20 bucks a
19:13 month for something you could do
19:14 yourself if you just hung out in this
19:15 Channel and started playing um and then
19:18 these three guys are um worth paying
19:22 attention to Ethan mik from Wharton
19:24 business school his blog is one useful
19:26 thing. org David Shapiro on yoube
19:29 [ __ ] brilliant um working on AGI he's
19:32 got a big open- Source cognitive
19:34 architecture project that's worth just
19:37 he's just Wicked smart like if you if
19:42 you're if you're uh if your ego is
19:44 getting the best of you and you think
19:45 you're pretty clever go watch his videos
19:47 for like five minutes you'll be like oh
19:49 right I'm
19:53 normal and then Paul rater from the
19:56 marketing AI institute's really good on
19:58 Tik Tok also Rachel Woods or the AI
20:00 exchange um she's really good all right
20:03 so that's that stuff
20:07 um if you're looking for a community if
20:10 if you are have been playing around with
20:12 this AI stuff and you're like I'm the
20:13 only weirdo in town nobody likes me
20:16 nobody wants to play with me nobody
20:17 wants to let me tell them how cool this
20:19 [ __ ] is I started this company last
20:22 December to be that community so um a
20:25 bunch of us in here hang out there and
20:27 then we also o have um bi-weekly
20:30 meetings we've got one coming up this
20:32 Tuesday and um and uh we've got someone
20:37 from in here is going to actually be
20:39 speaking at the salon although I haven't
20:41 touched base with him Herod D I don't
20:44 know if herod's on but if you are we
20:47 need to connect um but that's the link
20:49 tree there's a Meetup there there's also
20:51 a link to the Discord so in the Discord
20:54 you can ask questions there's guilds so
20:57 there's like an Art Guild and a writing
20:58 Guild and a beginners Guild and a
21:00 entrepreneur Guild and an AI and
21:02 business Guild it's pretty cool good
21:05 Community a lot of people from in here
21:07 playing there what time on Tuesday
21:09 the every other Tuesday meetups for the
21:12 salon are 5:00 p.m. to 700 p.m. mountain
21:16 time so 7: to 9 East Coast 4 to 6 West
21:21 Coast of in the
21:23 states Herod will be speaking at the
21:25 salon Tuesday I'll really try to be
21:27 there yeah yeah he's going to talk about
21:30 his um so Herod who's who's in here and
21:33 hangs out on the the Discord as well he
21:36 used to be a
21:38 programmer and then he started learning
21:41 this AI [ __ ] and he goes I think I like
21:43 art and now he's an artist he got he got
21:46 hired by a company to be their artist he
21:48 did it in six months he made the
21:49 transition in six months he didn't go to
21:51 art school he learned how to use these
21:53 tools and completely transformed his
21:55 career so he's going to talk to us about
21:57 that and probably show us some of his
22:00 creepy ass disturbing heavy metal
22:03 Halloween
22:04 appropriate Visions which are which are
22:07 wild so that'll be good um okay so
22:12 that's that uh story vine.com that's my
22:14 company and that's my whatever my tag my
22:17 handle it's that's breaker breaker
22:20 anybody anybody have uh CB radios when
22:23 they when they were a kid then you're an
22:26 old fart like me
22:29 breaker breaker good
22:31 buddy oh those were the days them was
22:34 the good old days
22:37 uh let's see did any of my new stuff
22:40 come in this was I think steo did this
22:43 one dawn of a new era
22:47 indeed it is Dawn of a new era isn't
22:52 [Music]
22:57 it all right so here's how we do this
23:00 here what's your 20 break breaker one
23:03 nine what's your 20 we got a we got a
23:06 smokey doing flipflops
23:09 on oh my God I used to know all that
23:12 [Laughter]
23:16 crap CB radio was the original internet
23:19 you know it it was just like this this
23:23 actually Tik Tock Tik Tock lives or CB
23:27 radio
23:30 right could you'd have two people
23:31 talking or two or three truckers talking
23:33 and everyone else listening and
23:34 occasionally chiming
23:39 in that's awesome so the way we do this
23:42 here ask questions if you have them
23:44 about um AI below or if you have a
23:47 project you want to talk about or want
23:48 to share or want me to demo something
23:51 I'm happy to do that
23:54 um what else this was a cool thing we
23:57 did last night a cool project we did
24:00 last night with Dolly inside chat GPT
24:03 where we kind of stumbled upon I
24:04 stumbled upon kind of a cool
24:07 style that I would not have
24:13 gotten let me pop up and look at some
24:15 questions here look at some questions
24:19 yeah welcome does it manipulate lips too
24:24 um you want virtual lip
24:26 shots there's there there's head shot
24:29 programs that do
24:32 that they make you look like a cartoon
24:35 character um I don't know what you mean
24:38 does it manipulate lips does dolly3
24:40 manipulate lips um what dolly3 doesn't
24:43 really do right now is editing um but
24:46 Photoshop generative fill does that so
24:48 if you um if you have a photograph or
24:51 you generate something where the lips
24:53 aren't quite right you can pop go into
24:55 um Photoshop if you've got it and and
24:58 just select the lips and then just put
25:02 in a prompt like plump lips or skinnier
25:04 lips or whatever the hell you want non-m
25:07 madeup lipstick lips hey Jen equals lips
25:10 oh oh
25:11 oh does hey Jen do lips which which one
25:14 are you asking about does Hy Jen do lips
25:16 hey Jen does
25:18 lips I thought you meant you wanted them
25:21 to look
25:23 plumper how would I look the lip shots
25:25 would I look pretty do I look pretty
25:31 how' I look with plastic surgery
25:34 [Laughter]
25:51 surprised okay I'm paying for chat GPT
25:54 what am I missing out on because I
25:55 haven't used any advanced features glit
25:57 her okay okay we're going to go show you
25:59 right
26:01 now let's
26:04 go start playing with some chat jpt oh
26:08 by the way I
26:10 also I spent a good chunk of time this
26:15 afternoon updating the YouTube channel
26:19 so if you want to see if you just can't
26:22 get enough of me and want want my voice
26:24 to put yourself to sleep we have a
26:27 YouTube channel
26:29 AI learning laab dtt on
26:32 YouTube I've
26:33 got everything up until last night's
26:37 live is live on the YouTube channel now
26:40 um the most recent ones don't have
26:42 descriptions because their
26:43 transcriptions aren't done yet I just
26:44 uploaded those today but all the all the
26:47 titles and descriptions of all those
26:49 videos and all the chapters were
26:51 generated with Claude claw. I take like
26:55 a three-hour transcript from these lives
26:57 paste it in there paste in a prompt that
27:00 I made over about a 50 session you know
27:04 50 50 hours of prompt manipulation and
27:08 then it just generates the the title the
27:11 description um hashtags and chapters for
27:14 the
27:15 video it's pretty
27:17 slick um I love early nights then I'm
27:20 not falling asleep and waking up
27:21 confused to janky Turtle
27:24 [Laughter]
27:26 Puffs um
27:28 okay so that's that's
27:32 live see there's AI learning
27:35 lab lots and lots of
27:38 videos yay you did
27:41 it all right let's go to let's go to
27:44 chat
27:45 GPT all right so if you're paying your
27:48 20 bucks and you have not played with
27:52 the advanced features you might not even
27:54 see the advanced features so we're going
27:55 to fix that so if you pay for chat GPT
27:58 plus you'll have that little yellow box
28:01 and you'll have both GPT 3.5 and GPT 4
28:04 but notice my gp4 has lots of goodies in
28:07 it why is that
28:09 because you actually have to turn the
28:12 beta features on explicitly so you go to
28:14 your
28:15 settings and you go to Beta features and
28:18 then you'll have a list of some amount
28:22 it's what's weird is they roll these
28:24 features out to different people at
28:25 different times so you should at a
28:28 minimum have plugins and Advanced Data
28:31 analysis you will likely have browse
28:32 with Bing and you basically just turn
28:35 them on and then they become
28:39 available so you've got browse with Bing
28:41 Advanced Data analysis plugins and then
28:44 I have Early Access to Dolly 3 which is
28:47 a different mode and it's really
28:49 interesting do you think companies
28:51 should allow or ban the use of chat GPT
28:54 and other platforms um I think companies
29:00 um should
29:03 actively um
29:05 be learning about this stuff and know
29:09 don't don't put anything proprietary in
29:11 them right now but I think every company
29:13 should absolutely be learning what these
29:15 things make possible these tools are
29:18 going
29:21 to
29:22 upend business models across every
29:26 sector and and what's going to happen is
29:29 any company that resists learning this
29:31 stuff that isn't actively learning this
29:33 and I mean actively learning it like
29:36 like their SE Suite is saying okay we we
29:41 want to learn about this stuff we want
29:42 to learn what it can do what it can't do
29:44 how it threatens our business how it can
29:46 make us more efficient how how it might
29:50 allow us to reinvent what we
29:53 do they're going to be in
29:56 trouble so I think every single company
29:59 should be on this and right now almost
30:02 it's it's very few are like a lot of
30:03 people haven't heard of
30:05 chbt there's a the one of the guys that
30:08 I recommended uh you follow is this guy
30:11 Ethan mik he wrote an article called um
30:14 detecting the secret cyborgs and what he
30:16 talks what he's talking about is there's
30:20 a subculture within companies right now
30:22 where employees are getting AI literate
30:24 but not telling the company about it
30:26 either because the company has an
30:27 anti-ai policy or the the employee is
30:32 doing in 4 hours what used to take them
30:34 40 hours and they don't want anyone to
30:36 know and so what's happening is they're
30:37 hoarding this Innovation and the
30:41 company's got no clue about what's going
30:43 on and and what Ethan mik talks about is
30:46 the the Senior Management of these
30:48 companies needs to aggressively seek out
30:51 who's [ __ ] who knows anything about
30:55 AI we're not going to fire you we're not
30:57 going to CHP your head off we will
30:59 likely you know promote you if you're
31:01 good at this [ __ ] right like bring them
31:03 into the fold bring them in try to
31:05 understand who knows what what are they
31:07 doing how are they doing it what are the
31:09 implications for your business so yeah
31:12 anyway
31:13 okay so where do we begin let's we we'll
31:17 begin with browse with Bing because this
31:19 used to be here but it's not so if you
31:22 do any kind of
31:24 search wait last time I'll ask
31:28 how many companies out there are like
31:30 open AI all right Detroit J ho hoen last
31:34 time I'll answer I actually went through
31:38 an entire [ __ ] pile of these things
31:40 buddy lose the attitude um these four so
31:46 open AI Google anthropic
31:51 Microsoft inflection meta quora
31:56 po
31:58 those are the big seven in my opinion
32:00 then you've got xai whatever Elon Musk
32:02 is doing right now and um whatever
32:06 Apple's going to
32:08 do and I don't know there's probably two
32:10 or three more there's one called cohere
32:12 there's I don't know there's there's
32:14 probably two or three more there's like
32:15 10 so those are the biggies like I talk
32:17 about them every every night I'm live so
32:20 every night so those are the ones so
32:23 don't ask anymore cuz you said you
32:26 wouldn't good
32:29 Lord what are your qualifications sir I
32:32 demand to
32:34 know you don't by the way Tik Tok is not
32:37 a requirement you can leave just here
32:38 you go like this and then I disappear oh
32:42 no wait if you're in the live then you
32:43 go like this and then this and then I
32:48 disappear make me go away
32:51 [Laughter]
32:56 please all right um so chat jpt used to
33:02 have browse with Bing they used to be
33:03 connected to the internet and then it
33:05 turned out that they were stealing
33:07 information from behind pay walls
33:10 um and some of the Publishers were not
33:13 liking that so they made it go away
33:15 but the way these large language models
33:18 work is to train the base model to train
33:21 that Bas large language
33:23 model takes a lot of time a lot of data
33:26 and a lot of compute power so there's
33:27 they have to kind of stop it at a point
33:29 in time so chat GPT has been trained up
33:32 until September 2021 but with browse
33:35 with Bing in it if if you ask it for
33:37 something that's more recent than
33:39 September 2021 it'll now just do a Bing
33:41 search and incorporate the results of
33:44 that search into its answer so I can
33:51 say what are some interesting things
33:55 happening this week
33:58 related to Halloween activities in
34:03 Denver and now what it'll do is It'll
34:06 recognize I asked for something more
34:08 recent than you know September 2021 and
34:11 it pops open a little thing it says
34:13 browsing and and by the way I haven't
34:15 shown this on the on the channel before
34:18 one of the cool things that this is
34:19 doing now is um so there it says reading
34:23 2023 Halloween something it's it's
34:26 actually showing you what doing like
34:28 sometimes it'll go to a website and then
34:29 it'll read that website and then it'll
34:31 go out and go into another website and
34:33 read that website so basically is it'll
34:36 give you activities of what it's doing
34:39 while it's searching the web and so
34:42 let's see um Denver Oddities and
34:46 Curiosities Expo October 7th well that's
34:49 pasted an evening of contemporary Cosmic
34:52 horror October 26th so so anyway so
34:56 that's that's browse with Bing so that's
34:58 that's cool um so I'm going to go new
35:01 chat now now we can go to plugins
35:04 plugins are a little comp they're not
35:07 complicated they're janky they're
35:10 they're a little overly complicated
35:12 right for what they do but basically the
35:15 deal with plugins is this there's a
35:16 plug-in store so so you you can enable
35:20 up to three plugins in any given chat
35:23 session and you can install as many as
35:25 you want I've probably got I don't know
35:27 or so here but there's a plug-in store
35:30 where there's like a thousand of these
35:31 things and so you can look up I don't
35:33 know um
35:35 video and it'll find you all the video
35:38 plugins and there's 8 * 4 32 P 32
35:42 different video
35:44 plugins and if you find one you like you
35:46 install it and then once it's installed
35:49 then you can just
35:52 um chat like you would normally chat and
35:55 if you do something that will invoke
35:57 that plugin it will go do something
36:00 it'll it'll go do whatever that plug-in
36:01 does so let
36:05 me I'll see if I can get a good one
36:08 [Music]
36:11 here I guess the the most visual one is
36:14 instacart instacart's pretty cool so so
36:16 I'll just do this one quick
36:19 um make me a meal
36:23 plan for my family where we're trying to
36:26 eat less carbs and more Turtle
36:28 meat period I'd like the plan to last
36:33 for 5 days and include interesting
36:37 recipes including desserts
36:40 period all right so now chat PT is going
36:44 to go off and it's going to write us the
36:46 little meal
36:47 plan because I have the
36:50 instacart plug-in turned on what it's
36:53 going to do when it gets to the end of
36:55 this meal plan is say would you like me
36:56 to make make a list for that you know
36:58 shopping list and then when I say yes
37:01 it'll fire off the instacart plug-in and
37:04 actually create me an instacart shopping
37:07 list that I can then just say I can just
37:09 go modify and say deliver all that
37:12 crap so so it there's oh Nancy reached
37:17 member level 11 leveling up very nice
37:22 very nice
37:23 indeed um would you like a detailed
37:27 recipe for any of these if so can I also
37:29 provide a shopping list and I'll just
37:32 say shopping
37:36 list so there it's firing off the
37:38 instacart plugin so similar to how it
37:41 kind of did that thing with browse with
37:43 Bing it's firing that thing off so
37:46 that's super cool super groovy cool and
37:49 if you're like Kyle but Kyle how many
37:52 fingerprints are on your
37:54 screen
37:55 672 why
37:57 any good AI to write a non-fiction book
38:03 um
38:04 so so I saw a what's it called plot dot
38:10 um hang
38:14 on yeah plot
38:17 do.ai go to plot do.ai is pretty
38:21 amazing it's it's a fairly
38:24 straightforward
38:26 um screenwriting organizing tool where
38:30 it basically just walks you through a
38:32 series
38:33 of you know increasingly
38:37 detailed
38:39 um like story elements but it's got it
38:44 looks like it's not happy right
38:48 now come
38:51 on I bet I bet my browser is having
38:55 memory issues oh there you go so so
38:57 anyway so come
39:00 on so so there's genre theme audience
39:04 title log line so so here's the basic
39:07 idea then here's the grand scheme here's
39:09 the hero's journey so basically you just
39:12 walk through filling out these different
39:14 sections of a screenplay and then you
39:16 can just hit the unleash ideas
39:20 button at the top right here and it just
39:22 fills [ __ ] out for you so this is one
39:25 that that kind of does it and then you
39:26 can go in and you can edit it and then
39:28 from there it will go ahead and and
39:30 start writing the screenplay there's
39:32 also a project called
39:36 um GPT author now that one you actually
39:40 have to get running in a Google collab
39:43 notebook but but it's insane you you
39:46 have to have an open AI key like your
39:48 API key and you have to have a stable
39:50 diffusion API key you put your two API
39:53 keys in and then it asks you for like
39:56 what do you want the book book to be
39:57 about and what do you want the style to
40:00 be I
40:01 think and then what it does is it takes
40:05 your inputs and it writes an outline for
40:08 the book and then it writes a chapter at
40:10 a time it writes up all the chapters up
40:13 to the number of chapters you want so if
40:15 you say you want a 12 chapter book or a
40:17 20 chapter book it'll just write all
40:19 those chapters and then at the end it um
40:24 takes the story and uses state diffusion
40:27 to create a cover image for your book
40:30 and then it publishes it as an ebook so
40:33 so author GPT actually writes a fiction
40:38 from beginning to end in about I don't
40:41 know 3 minutes now your next question
40:44 I'm sure is is it any good no of course
40:47 not but but it writes an entire book
40:51 including the outline in all the
40:53 chapters and the cover and publishes it
40:55 as a book in a few minutes that's
40:58 reality right now the fact that it isn't
41:01 any good that's just prompting and
41:04 improving the underlying models and
41:06 that's all happening in real time so um
41:09 so you should start playing with it
41:11 because it's worth the investment to
41:12 learn how to do it um so anyway here's
41:15 the instacart um shopping list that it
41:18 made
41:24 me so there's all their ingredients do I
41:27 have Turtle meat in there Turtle
41:30 meat skinless boneless wild cut mahi
41:33 mahi it's not exactly Turtle meat but it
41:35 at least understood I wanted Turtle
41:38 meat I guess I'm on my own for C
41:40 catching my own snapping
41:44 turtles so so that's
41:47 plugins now let's go and look at by the
41:51 way this is what I'm doing right now is
41:53 uh someone asked they they've got they
41:55 said they pay their 20 bucks a month for
41:57 chat GPT and I said it's worth paying
41:59 the 20 bucks and they're like I paid my
42:01 20 bucks but I don't really use the new
42:03 features what are they so I'm just going
42:04 through them one by one so we did browse
42:07 with Bing um we did plugins now I'm
42:09 going to do Advanced Data analysis this
42:11 is
42:13 um this is for me just this side of
42:16 voodoo right this is this is the Arthur
42:19 C Clark any any sufficiently advanced
42:21 technology is indistinguishable from
42:23 Magic that's what this is for me cuz I
42:26 know nothing about data analysis
42:28 but I'm going to click on the little
42:30 plus button here and I'm going to go
42:34 find archive because in the past I have
42:40 downloaded
42:41 um CSV files from the internet if if you
42:45 want to if you want to find data to play
42:47 with you can use your own data or you
42:49 can also um there's a site called kaggle
42:52 k a g GLE
42:54 e Nicholas level 13 welcome thank you
43:02 um what is this oh that's not I want
43:07 okay let's
43:17 see where's my
43:20 archive
43:21 archive zip archive here we
43:25 go
43:30 all right so I'm going to grab this ZIP
43:32 archive and I'm up uploading it so I'm
43:35 uploading that to chat gbt so I'm in
43:37 Advanced Data
43:39 analysis I've I've either grabbed a
43:41 piece of my own data like a CSV file or
43:43 an Excel file or it can be it can be
43:46 really any file it'll analyze them like
43:49 you can analyze the colors in an image
43:51 for
43:52 example but I've got this archive here I
43:54 don't really know what's in it so what
43:55 I'm going to say is is um please open
43:59 this ZIP file and tell me what's in it
44:01 period if it's a
44:03 CSV
44:06 analyze the fields represented and tell
44:10 me what might be interesting about
44:13 it so I'm literally just saying hey I
44:16 don't know what's in here can you tell
44:18 me and so it says all right first I'll
44:20 extract the contents of the zip file and
44:23 list what's inside now if I open this up
44:25 what it's actually doing is it's created
44:27 a what's called a container So within
44:30 this browser session it's created like a
44:32 little mini instance of a like a
44:35 computer operating system where it can
44:38 it's it's been given access to python
44:40 libraries and it can write and um it can
44:46 write and execute its own python code so
44:50 when I said look in the zip file it's
44:52 importing a library that can do that
44:54 writing its own python code and
44:55 executing it to look inside the file
44:58 this is unemployment data and then it
45:01 grabs the Panda's library and writes
45:04 more python code to look at the data and
45:07 then it can tell us what's in there
45:09 right so it's starting to analyze it the
45:11 question what is the plus button for
45:14 doesn't show on my side so Apple Apple
45:17 user you need access to um Advanced Data
45:21 analysis so if you pay for GPT plus so
45:25 if you pay your 20 bucks a month go into
45:28 your settings go into beta features and
45:31 turn on Advanced Data
45:33 analysis and then when you click on gp4
45:36 you'll see a pop- down menu and you've
45:39 got default gp4 browse with Bing plugins
45:42 and then Advanced Data analysis Advanced
45:45 Data analysis is the thing that gives me
45:47 the little plus button and the plus is
45:49 basically just upload all right oh good
45:53 concise description of a container yay
45:55 thank you
45:56 thank you because I was thank you Pate
45:59 Pate actually knows what he's talking
46:00 about on this stuff um I used to
46:03 describe it as a virtual machine and I
46:05 got slapped for that it's not a virtual
46:08 machine it's a container so okay but
46:10 anyway it's basically running like a
46:12 little mini mini instance of a of a like
46:15 an operating system that that
46:18 can that can write
46:21 execute um python code and and then
46:24 display the results right
46:27 okay so um let's see Trends and
46:29 unemployment um let's see uh I want to
46:32 see I want to try something
46:36 here is there employment data for the
46:39 state of Colorado anywhere in
46:51 there you can also this you don't just
46:53 have to do this doesn't have to be CSV
46:55 file like you could up upload um a bunch
46:58 of screenplays or creative writing or um
47:02 annual reports and start doing analysis
47:05 on it things like that yes there's
47:07 employment data for the state of
47:08 Colorado would you like me to would you
47:10 like to see a summary or con conduct
47:12 specific so here's what I'm going to
47:17 do can you create a new CSV
47:22 file that's just the Colorado data so I
47:25 have that is a standalone
47:30 file so one thing it can do is it can
47:32 also generate files right because it's
47:34 kind of running as this little operating
47:37 system you can say Okay since you're
47:40 looking at all the data can you just
47:41 grab the Colorado data and so it says
47:45 sure here's a CSV v file you can
47:47 download it
47:50 here and okay it downloaded it and now
47:53 let's go over to
47:55 numbers
47:59 let's see and go
48:03 open and go to
48:06 downloads and there's the CSV so I open
48:09 it
48:10 up and so
48:13 there's all my Colorado employment data
48:17 all right and is it it's just one looks
48:20 like it's just one worksheet so it's
48:22 relatively simple table but so imagine
48:25 that so so so like one of the things it
48:27 can do is it can clean up data so let's
48:29 say you got a bunch of shitty data like
48:31 everyone entered their addresses in you
48:35 know not using the right fields or like
48:37 you just got dirty data you can have it
48:39 go in and clean up data and just throw
48:41 out the the incomplete records you know
48:44 take all the cleaned up data save that
48:46 to a new CSV file and bang now you've
48:48 got new new
48:49 stuff so so it did that so so it
48:52 generated a new CSV so now I'm going to
48:54 say um
48:57 can you find interesting Trends within
48:59 the Colorado data and tell me what they
49:01 are so now I'm going to actually have it
49:04 do some
49:05 analysis certainly we're going to look
49:07 at the overall unemployment rate Trend
49:09 the total employment Trend the labor
49:11 force participation rate seasonal
49:14 variations let's start by plotting these
49:16 Trends so again what it's going to do is
49:20 it's going to load in a library this is
49:22 a graphics library or like a like a
49:24 plotting Library matplot live. pyplot as
49:29 PLT and so now it's writing what chart
49:33 it wants to make so it's it's going to
49:35 analyze the data and make us a pretty
49:37 little
49:39 chart um so there's our charts right
49:42 there's our what do we have here we've
49:44 got unemployment total employment and
49:47 labor force
49:49 participation gez wonder what happened
49:52 here when is that 2020 what could that
49:55 be
50:02 um so so that's Advanced Data analysis
50:07 absolutely Insanity okay and then I'll
50:09 show I'll show you the bonus thing um
50:12 that's in here that um everyone should
50:14 have access to within a month is what I
50:16 heard on the open AI Discord
50:19 today but that's Dolly 3 so Dolly 3 is
50:23 is this image generation tool like dolly
50:26 2 except Dolly 2 sucked and and you you
50:29 had to do it other
50:30 places Dolly 3 inside chat jpt is is a
50:34 bit of a revelation so I'm going to say
50:38 um I'm starting a business to rent water
50:43 ski equipment to athletic
50:46 Turtles and want help coming up
50:51 with images for an advertising campaign
50:54 period before I create the images I need
50:58 to come up with some clever names for
51:01 the company period can you
51:04 help so I'm just sort of acting like a
51:08 bad client who would talk to a creative
51:11 director turtle water sports that is
51:15 that is oddly sounds a little dirty
51:18 turtle water
51:22 sports um shell ski rentals Aqua turd
51:26 Adventures nah that's too close to Aqua
51:29 turd Adventures but nice try Wave Rider
51:32 Turtles Fin and flippers
51:34 floats um Turtle Glide gear seashell ski
51:38 shop okay we're going with seashell ski
51:40 SC shop because there's no way to
51:41 pronounce it it's perfect
51:44 okay I like seashell ski shop period can
51:49 you make me four images of an athletic
51:53 turtle water skiing
51:56 in an idyllic setting
51:58 period give me four different styles so
52:01 I can choose from them and be sure to
52:05 include seashell ski
52:09 shop text on the image as well so I can
52:15 get some logo ideas
52:18 period and
52:21 now it
52:24 will start writing prompts for
52:27 me so first one photo of an athletic T
52:30 turtle skillfully water skiing on a calm
52:33 Blue Lake with Lush Greenery in the
52:35 background the sun casts a golden hue
52:37 over the scene thing just
52:39 flickered
52:51 um there's something about slow things
52:54 going fast we should do do a whole
52:56 series on slots athletic
52:59 slots um but it does text now you'll
53:02 notice it it doesn't it can't spell
53:05 necessarily but but these things
53:06 couldn't do text at all before that's
53:09 actually pretty good and it's kind of
53:11 fun look all of his Turtle buddies are
53:14 rooting for
53:15 him as he terrorizes the lake that's
53:19 pretty
53:20 good there you go little TMNT you could
53:24 Photoshop out that extra ey there
53:26 seashell ski shop how can I help you so
53:30 that's Dolly 3 within chat GPT so why
53:33 pay for GPT 4 why pay for GPT plus the
53:37 20 bucks because you get all this [ __ ]
53:40 quite frankly the the Advanced Data
53:42 analysis alone is probably worth 50
53:44 bucks a month the fact that you can now
53:46 just do this kind of stuff here let me
53:49 show you one other thing that you can do
53:50 with images which is absolutely
53:53 Bonkers
53:54 um
53:56 I have an idea for a book
53:59 where
54:01 the reader looks at images
54:04 with dozens and dozens of objects on a
54:09 page and has to
54:11 identify the objects on a list to the
54:16 right can you give me five different
54:19 concepts
54:23 of object collections that would be
54:26 interesting to put into a book like this
54:28 period I don't want to see anything yet
54:30 I just want ideas for what we might look
54:38 at and again because it wants to be so
54:40 helpful if you don't if if you don't
54:42 want it to do something generally tell
54:44 it you don't want it to do
54:47 something vintage travel oh that could
54:49 be really cool
54:51 looking wizard study space explor ation
54:56 these would all be
54:58 good Victorian curiosity
55:03 cabinet Carnival Extravaganza oh that
55:06 would be a good one too so wait what do
55:07 we have actually they're all kind of
55:16 cool I actually think they're all pretty
55:19 neat please give me a list of 15 objects
55:25 per per
55:26 category that would be the
55:29 objects listed for the person to find
55:44 period an antique Compass a leather
55:47 suitcase a globe on a wooden stand the
55:50 wizard study a crystal ball a spell book
55:52 a wand with a
55:54 phoenix
55:56 Enchanted
55:58 ring space exploration moonrock sample
56:01 in a container UFO with blinking
56:04 lights so cool little lists all right so
56:07 this is
56:15 fun we'll do the Victorian curiosity all
56:23 right perfect give me four
56:28 examples of
56:31 images in landscape
56:34 Orient yes orientation it got that my
56:39 tongue didn't work but it understood
56:40 where I was going uh in landscape
56:43 orientation
56:45 um for the
56:48 Victorian
56:49 example and be sure to include the 15
56:52 objects you have in the list as as well
56:55 as 25 other objects
57:03 period one concept should be a top down
57:07 shot while another concept might be lots
57:11 of objects in a room period I'll let you
57:14 come up with the last two
57:19 concepts so again I'm not having to like
57:23 like the so for one thing the ability to
57:27 do multiple objects coherently in
57:31 anything like mid Journey has never been
57:33 possible it'll combine objects it's just
57:35 a mess so the fact that this thing does
57:38 coherence where where it you'll see it's
57:41 it's going to do like now it will it get
57:43 these perfect it it probably won't get
57:45 them perfect but the fact that it can
57:47 get anywhere in the neighborhood of this
57:49 is truly [ __ ]
57:51 remarkable um so so we'll see what it
57:54 does we see we'll see how it does it um
57:56 but but I've been I've been really
57:58 impressed with this the fact that it can
57:59 do text in a pretty clear way the fact
58:01 that it can do multiple objects in a
58:03 scene or multiple people in a scene and
58:05 actually get them decently accurate is
58:08 is remarkable and then the fact that
58:09 it's doing it where it's writing all
58:11 these prompts right like it's like I'm
58:13 not having to think about this I just
58:15 said give me some Concepts and then
58:17 [ __ ] come up with a bunch of objects
58:19 and it did and now it's going to stick
58:21 them all in
58:23 pictures I wonder if it could do photos
58:25 of my business values oh Craig I got to
58:28 show you I got to show you what I did
58:29 the other day so yes as a matter of fact
58:33 it was the first so funny Craig the
58:36 first project I did with this was I took
58:38 the description of the AI salon and our
58:41 values we have six values and each one's
58:44 described and I put it into chat GPT and
58:47 I said I want to bring our values to
58:49 life and I want you to help me visualize
58:50 them and within 10 minutes I'll show
58:53 I'll show you what it
58:54 created
58:56 all right so there's the Victorian top
58:59 down view there's the room wow these are
59:05 good I forget what the list of crap is
59:08 but here let me open the let me open
59:10 this let me see um copy image
59:16 link and paste image link oh that just
59:20 downloaded it hang on open
59:24 file
59:33 I mean come on how cool is this it's
59:37 just
59:37 amazing I don't I don't know what the
59:39 objects were hang let me go look at the
59:41 objects let's see we'll go find some
59:43 objects
59:47 here okay delicate butterfly specimen
59:51 yes lower left wait sorry um lower left
59:54 right there it end up
59:57 here um ornate gold pocket watch uh
1:00:01 maybe oh wait here ornate gold pocket
1:00:05 watch spiral
1:00:07 seashell this is like a this like a good
1:00:10 game like you got to actually look
1:00:11 through all this [ __ ] all right so
1:00:13 that's that and then let's let's go down
1:00:15 and look at what some of the other
1:00:17 images it did were here's the
1:00:19 chest very
1:00:21 cool and then there's the messy the
1:00:24 messy Des
1:00:27 so I I mean this is just stunning and
1:00:30 then so so you asked about the
1:00:35 [Laughter]
1:00:38 values all right there's
1:00:44 curiosity there's
1:00:46 generosity so basically I forget which
1:00:48 one it did first I think it did
1:00:51 exploration
1:00:52 first and it did
1:00:55 it did four different styles right one
1:00:58 was photographic one was this one was I
1:01:01 don't know some like a water color and
1:01:03 so I said I like this style so I told it
1:01:06 that I like this style and then the
1:01:08 first one it did it didn't have the word
1:01:10 exploration on it so I said oh could you
1:01:12 add the the name of the value could you
1:01:14 incorporate it into the
1:01:16 scene and so I did this one and then the
1:01:19 next two it got perfect I just said now
1:01:22 do the next value now do the next value
1:01:24 and it did them
1:01:26 perfectly I think collaboration was one
1:01:28 of them inclusion wasn't empathy
1:01:35 wasn't I think curiosity was one of the
1:01:38 first three and then the last three it
1:01:41 didn't get them quite right and then I
1:01:43 just I kept going no that's not right no
1:01:45 that's not right but but it was maybe
1:01:48 five iterations of each but but
1:01:51 seriously within less than 15 minutes
1:01:54 probably closer to 10 I went from
1:01:56 copying and past this was the first
1:01:58 thing I tried when I got access to this
1:02:01 I copied and pasted here's a description
1:02:03 of the organization here's a description
1:02:05 of our values can you help me bring
1:02:07 those to
1:02:09 life bang oh uh Firefly thanks for
1:02:12 reminding me the custom instructions
1:02:14 I'll I'll also show those um so so yeah
1:02:17 these were created um these were created
1:02:19 with Dolly 3 this was my first set of
1:02:21 Dolly 3 images I did these for the AI
1:02:24 salon
1:02:26 on um good call on custom instructions
1:02:31 so that's one thing I haven't um showed
1:02:36 yet so if you go back here so custom
1:02:39 instructions are available
1:02:42 um whether or not you pay for chat GPT
1:02:45 plus so even if you're using free chat
1:02:48 GPT 3.5 custom instructions are there so
1:02:52 custom instructions are next to your
1:02:54 your settings and they just say custom
1:02:57 instructions and they try to tell you a
1:02:59 little bit what they are but it's not
1:03:01 super obvious so basically what you can
1:03:03 do is you can tell chat GPT to act like
1:03:07 a certain kind of
1:03:09 character and um do and and respond in a
1:03:14 in a particular way so I'm going to turn
1:03:17 these
1:03:18 on and save them and now anything I do
1:03:21 from this point forward it'll follow
1:03:23 these rules but let me let me show tell
1:03:25 you what the rules are and then we'll go
1:03:26 look actually you know what I'm going to
1:03:28 turn them off let me turn it
1:03:32 [Music]
1:03:33 off and I'm going to go ask it a fairly
1:03:36 straightforward question
1:03:38 um can you describe to me why the sky is
1:03:42 blue even though I don't know anything
1:03:44 about science and can barely keep up
1:03:48 with a fifth grade textbook
1:03:53 period of course I'll explain it in
1:03:56 simple way right so imagine a sky is big
1:03:58 and clear blah blah blah blah blah
1:04:01 right so I'm going to
1:04:04 copy that
1:04:06 prompt fairly straightforward thing the
1:04:09 reason the sky is blue is because of
1:04:10 something called scattering tiny
1:04:12 invisible things in the air like dust
1:04:14 and molecules can scatter
1:04:16 sunlight and when sunlight enters our
1:04:19 atmosphere it's made up of many
1:04:20 different colors like a rainbow but
1:04:21 those tiny things scatter the sunlight
1:04:23 in all directions okay okay so that's
1:04:26 the description so now I'm going to go
1:04:27 turn on my custom
1:04:32 instructions and what the custom
1:04:34 instructions
1:04:36 say oh
1:04:39 wait is I'm a creative director that has
1:04:42 a passion for structure and Clarity I
1:04:44 can synthesize big Concepts into very
1:04:46 simple metaphors people can remember so
1:04:49 that's that's who the chat is right so
1:04:52 you're you're kind of putting it in an
1:04:54 arena of expertise for lack of a better
1:04:56 term how would you like chat GPT to
1:04:58 respond when someone asks a question
1:05:02 respond following these rules mirror
1:05:05 back what you think you
1:05:06 heard answer with a non-obvious
1:05:09 Innovative idea to help clarify or to
1:05:12 help clearly answer the
1:05:13 question when you answer include a
1:05:16 metaphor that is Visual and make your
1:05:18 idea easy to remember follow your answer
1:05:21 with five bullet points of 7 to 10 words
1:05:24 each and then finally create a table
1:05:26 formatted in markdown that compares your
1:05:29 idea to a traditional way of thinking
1:05:32 format it in markdown markdown is a
1:05:34 thing kind of like Poor Man's HTML that
1:05:37 can make bold and like title text and
1:05:40 bold and bullet points and things like
1:05:42 that so if you ever it's just a a neat
1:05:45 little hack just say format this and
1:05:47 markdown to make it easier to read and
1:05:49 it will just add all that [ __ ] for you
1:05:52 so anyway so that's that's what I've got
1:05:55 and so now I'm going to create a new
1:05:57 chat and ask that exact same question as
1:06:00 I asked before and we'll see how it
1:06:04 responds so there's our
1:06:07 table gave us a couple of tables
1:06:11 interesting let's see what it says I
1:06:13 think you're asking why the sky appears
1:06:15 blue even if you don't have a strong
1:06:16 Science Background here's an Innovative
1:06:19 way to explain it metaphor think of the
1:06:21 sky as a giant blue sponge when sun
1:06:24 light enters the Earth's atmosphere it
1:06:26 encounters tiny molecules and particles
1:06:28 here's the
1:06:30 breakdown aspect Innovative explanation
1:06:33 traditional
1:06:35 explanation and so it's now behaving
1:06:38 that way so you can and you can do you
1:06:40 you can make it behave in any way you
1:06:42 want I've had it Act Like A
1:06:45 pompous um A pompous ass that uses big
1:06:48 words that even it doesn't understand it
1:06:51 does so um so that's custom instru
1:06:54 instruction so anyway there's your deep
1:06:56 dive on
1:06:57 chat you can make money with
1:07:00 ch especially if you use the advanced
1:07:03 features all right
1:07:08 so where's my photos app did I crash
1:07:15 it
1:07:17 photos come on
1:07:23 photos
1:07:25 AI learning
1:07:27 lab all right fantastic Bob tell him
1:07:31 what he's want you want a new
1:07:35 car welcome everybody if there's new
1:07:37 folks here my name's Kyle Shannon this
1:07:39 is the AI learning lab the T-shirt
1:07:40 tonight is a bosot of Sugar Ray
1:07:45 Robinson I think it's got does it have
1:07:47 it oh no it doesn't I think it's on the
1:07:49 back it says
1:07:50 basat tone length dig deeper why am I
1:07:53 not seeing these prompt buttons on chat
1:07:56 on GPT 3.5 on some systems tone length
1:08:02 and dig
1:08:06 deeper Rowan are you are you where are
1:08:10 you seeing those buttons that it sounds
1:08:12 to me like you're in an app that's not
1:08:14 from open
1:08:15 AI if Bing has those buttons that's just
1:08:18 something that Bing has done so Rowan
1:08:21 put put in if you're still here put in
1:08:23 what what you're actually using because
1:08:25 I think the reason you're not seeing
1:08:27 buttons like that is you're inside some
1:08:29 other
1:08:32 application that would be so effective
1:08:34 And archiving documents I don't know
1:08:36 what I was doing we are not naming the
1:08:38 company that that's a mouthful she sells
1:08:41 seashells down by the sea ski shops down
1:08:44 by the seashore exactly Ted looks at
1:08:46 life gets it he understands the
1:08:48 Brilliance of that naming of that
1:08:50 business did you see my message about 60
1:08:52 Minutes Kyle I did I didn't get a chance
1:08:53 to see it though I'll watch it I'll
1:08:55 watch it tonight um what was the what
1:08:58 was the big
1:08:59 takeaway what program is this I'm not
1:09:01 sure which one I was
1:09:05 using this is the sequel to 2023 is
1:09:08 breakout hit game find objects that
1:09:10 start with the same letter exactly Pate
1:09:12 P's pate's been in here pate's one of
1:09:14 the Irregulars so so by the way if
1:09:16 you're if you're new to the channel we
1:09:17 have we have folks that show up here
1:09:19 pathologically I mean uh regularly and
1:09:22 they're called the Irregulars and
1:09:23 they're awesome and they hang out and
1:09:25 they'll help you and if you end up
1:09:27 joining the AI Salon Discord they're
1:09:29 also hanging out over there so um we
1:09:32 have we have a lot of fun and work on
1:09:35 projects together and when we're sort of
1:09:37 doing fun things in here um they'll
1:09:39 often make images and pop them over over
1:09:42 into the
1:09:44 Discord Ye Old Discord janky
1:09:52 wood why no one making shirts with Words
1:09:55 Right In Reverse so they show up
1:09:57 properly on the front um you can do that
1:10:00 um let's see if we can even
1:10:02 get uh Ye Old Dolly to do that for
1:10:07 [Music]
1:10:10 us I'm guessing it will do it but let's
1:10:19 see I'm feeling nostalgic
1:10:23 for t-shirt shops on the boardwalk on
1:10:26 the Jersey Shore in the' 70s period I'd
1:10:31 like you to
1:10:33 make an illustrated t-shirt
1:10:37 graphic that says muscle cars rock with
1:10:41 a 1970s muscle car in Grabber orange
1:10:48 period flip the image so that the words
1:10:51 are backwards so I can print it
1:10:54 as an iron transfer and smell the sweet
1:10:58 sweet chemicals as they meld into the
1:11:01 cotton
1:11:08 fibers wait it looks like you haven't
1:11:10 asked the question what where did my
1:11:12 thing go H
1:11:16 really oh good
1:11:19 God I'm feeling nostalgic
1:11:23 about
1:11:24 um t-shirts being made on the boardwalk
1:11:27 on the Jersey Shore in the' 70s period
1:11:30 can you make four different designs for
1:11:33 a t-shirt that says muscle cars rock
1:11:36 with an orange muscle car period flip
1:11:40 the image backwards so that the
1:11:44 words are in Reverse so they'll print
1:11:48 correctly when I iron them
1:11:53 on
1:11:58 okay certainly oh and I still have my
1:12:00 custom instructions on so so it's
1:12:03 actually responding that's really it's
1:12:05 still responding really weird prompts
1:12:08 for
1:12:10 Dolly I wonder if this is going to work
1:12:13 it
1:12:14 should once it gets done I got to turn
1:12:16 off my custom instructions this is
1:12:19 [Laughter]
1:12:22 annoying
1:12:29 here's the comparison table traditional
1:12:32 thinking Innovative idea now let's
1:12:34 create those images okay so whatever so
1:12:37 like this is one of the one of the
1:12:39 fascinating things about doing this
1:12:41 image generation within chat GPT is like
1:12:44 you can do fairly
1:12:47 sophisticated you know creative strategy
1:12:50 or marketing strategy or whatever
1:12:51 whatever you want to do and then bring
1:12:53 that stuff to life so let's let's see
1:12:56 how it does and let's see if it flips
1:12:57 them backwards I'm guessing it
1:13:00 will nope but here
1:13:06 watch
1:13:07 um I think we'll do the the Challenger
1:13:11 here copy
1:13:13 image go here go
1:13:16 new go
1:13:18 tools flip horizontal there you
1:13:22 go see it did it yay you're a
1:13:26 winner he made everything backwards
1:13:30 yay it doesn't look like a 70s t-shirt
1:13:33 but it's kind of
1:13:34 [Laughter]
1:13:35 [Music]
1:13:37 cool muscle cards
1:13:45 [Music]
1:13:50 Rock All Right H I hope I hope that
1:13:54 helps Sherry Banks thank you for the
1:13:56 rose I like a good rose rose smells
1:13:59 lovely it smells delicate and yet Pricks
1:14:02 the finger
1:14:12 ow as I'm wait it's looking like parts
1:14:16 of Market first
1:14:19 clients
1:14:21 what it's looking like part
1:14:25 I don't understand I don't understand
1:14:27 what you say Apple user it's looking
1:14:31 like a parts of Market first
1:14:36 clients War for big players so lots of
1:14:38 free
1:14:40 oh I think I know what you
1:14:42 mean yeah so so lots of these tools
1:14:45 right now when they launch they want to
1:14:47 get customers so they give you [ __ ] for
1:14:49 free and then they say no you can't have
1:14:51 that for free now you got to pay for it
1:14:54 what you get for chat GPT for 20 bucks
1:14:57 is a lot and I don't work for them I
1:15:00 don't I don't make anything by saying
1:15:02 that so I don't care if you go there or
1:15:06 not but yeah they're all going to do
1:15:08 that I think Bing is going to continue
1:15:10 to be free because they're using it as a
1:15:11 loss leader for all their other [ __ ]
1:15:13 that they sell
1:15:16 Microsoft ADHD sales
1:15:22 pitch why is no one making t-shirts with
1:15:24 the words in Reverse we just did
1:15:26 that um that they're wait that they're
1:15:29 already smarter than us and they're
1:15:30 going to be the end of us
1:15:33 all Amelia's wife what was that in
1:15:35 reference to which one could hello with
1:15:39 Illusions diagrams for a book also
1:15:42 writing um just go to
1:15:45 Bing um or or
1:15:49 ideogram um they're both free right now
1:15:52 I think Bing's going to staye free so if
1:15:54 you go to bing.com
1:15:56 create it's their image generation tool
1:15:59 and it's free and it's Dolly 3 and
1:16:01 they're really
1:16:02 good ifran ifran durani
1:16:10 829 all right I saw the 60 Minutes piece
1:16:13 too they said AI is teaching itself to
1:16:15 code
1:16:17 scary is it teaching itself to code no
1:16:21 we trained it on code so that it can
1:16:27 code H like all of these news things are
1:16:30 so [ __ ] Sensational it's just it's
1:16:33 like they they've got to have the I mean
1:16:36 it's it's such a it's teaching itself to
1:16:39 code no it's
1:16:41 not might it in the future it
1:16:45 might I mean they're using they're using
1:16:48 synthetic data they're using data that
1:16:51 large language models create to train
1:16:54 other large language models so I guess
1:16:56 in that sense it is but this is all at
1:16:58 the behest of
1:17:02 scientists um talking about 60 Minutes
1:17:05 certainly not what I think oh oh I see
1:17:07 got
1:17:10 it yes you got it media's aim is to
1:17:13 scare 60 Minutes pieces yeah exactly I
1:17:16 feel like that was irresponsible
1:17:17 journalism and an attempt to shut it
1:17:19 down it here's the deal yeah exactly
1:17:22 yeah too late
1:17:24 um this shit's this shit's the the the
1:17:26 the horse is so far out of the barn that
1:17:29 even if let's let's
1:17:31 say Biden signs an executive order AI
1:17:36 companies need to cease to exist meta
1:17:39 has already op sourced all their models
1:17:41 the The open- Source Community with AI
1:17:46 is insane it there people are going
1:17:49 crazy just building all sorts of
1:17:52 applications on top of the open Source
1:17:55 stuff they would never do that because
1:17:58 no other country is going to stop the
1:18:00 Innovation right the these are the most
1:18:02 powerful tools ever put in the hands of
1:18:04 humanity you know short of the nuclear
1:18:08 nuclear bomb but but um but we don't all
1:18:12 have access to nuclear bombs right for
1:18:14 for good reason um this AI stuff's
1:18:16 incredibly powerful and we've all been
1:18:18 given access to it so and as has
1:18:21 everyone in the world so no other
1:18:23 country is going to stop they're they're
1:18:25 not going to squash The Innovation so
1:18:27 this stuff is not only coming it's here
1:18:30 and I mean the stuff that I showed like
1:18:33 this image right here this was just sort
1:18:35 of a brain fart we were kind of joking
1:18:37 around one night and said oh let's make
1:18:39 a Pulp Fiction looking novel with these
1:18:42 crazy animals we always talk about on it
1:18:44 and it just did it like that like that's
1:18:45 it's here it's not
1:18:48 coming it's here it's not a lot of
1:18:51 people know about it yet but it's here
1:18:53 here there's no stopping it this is the
1:18:57 60 Minutes piece they they they might as
1:18:59 well have done a piece on you know how
1:19:02 awful it is that hurricanes keep
1:19:05 attacking
1:19:06 Florida do a hit Pac piece on the
1:19:09 Hurricanes okay I mean I I appreciate it
1:19:13 it's it's not going to stop
1:19:16 them can I really make can I really have
1:19:19 ai write me a book yeah you really can
1:19:21 have ai write you a book go check out
1:19:23 GPT author the project called GPT author
1:19:27 it's you have to geek out on a little
1:19:28 bit but I'm sure at this point someone's
1:19:30 made it a double click install for the
1:19:33 application um but also Naila you know
1:19:36 so actually let me show you something
1:19:37 let's let's go let me go demo something
1:19:41 because here's the
1:19:45 thing I I I almost just fell into the
1:19:48 Trap I almost just wait you almost got
1:19:52 it to the bottom line Kyle you're going
1:19:54 to wait warp to level two if you do oh
1:19:58 um the the writing of the book thing so
1:20:02 the way you ask that it just hit me that
1:20:04 here's the Trap that I almost fell in
1:20:06 one of the misconceptions right now is
1:20:08 the robots are writing books the robots
1:20:11 are stealing images the robots are doing
1:20:13 this the robots are doing that I heard
1:20:16 you can make money with jbt can I give
1:20:18 you my life savings and you can teach me
1:20:20 Absolutely cockal I'll give you my uh
1:20:22 I'll give you my venmo I'll give you my
1:20:24 venmo later you just send over your life
1:20:26 savings and you can make money
1:20:29 with I'll even produce a solid gold
1:20:33 button for you that you can push on your
1:20:35 own for for the low low price of
1:20:38 everything you've ever worked for
1:20:43 um these AI tools are not operating
1:20:49 independently of us they may in the
1:20:51 future but right now these are tools
1:20:54 can AI predict the future it will it
1:20:56 will it will tell you that it won't and
1:20:59 then it will go ahead and predict it and
1:21:00 it it will be about as right as a magic
1:21:02 eightball um or you know whatever things
1:21:05 like that so no it can't um but so let
1:21:10 me let me show you something I'm going
1:21:11 to because the thing that I would
1:21:13 encourage you I would strongly encourage
1:21:14 you to do Nayla is wait did GPT author
1:21:18 require a phone call after you entered
1:21:21 two keys
1:21:27 no I think it's author GPT not GPT
1:21:30 author you might that might be a scam
1:21:32 site do not give them your phone
1:21:34 number
1:21:36 well if you're if you're on open AI wait
1:21:39 so so one thing n oh it's Nancy not
1:21:43 Nayla Nancy if you're on open AI open AI
1:21:47 I think does require your phone number
1:21:49 for their
1:21:51 account um but if it's an application
1:21:55 called GPT author so it shouldn't be an
1:21:57 application it should just be a
1:21:59 project hang
1:22:01 on I'm going to go look up uh author
1:22:15 GPT oh it's GPT author so it's m it's
1:22:19 it's it's at github.com
1:22:22 M Tor s h u m r
1:22:27 slgp d
1:22:33 author and then um you basically run it
1:22:36 as a I think there's a a Google collab
1:22:40 notebook here or if there's not yeah
1:22:42 wait in Google collab simply open the
1:22:44 notebook add your API keys and run the
1:22:47 cells in order so if if in getting your
1:22:51 API keys at ask you for your phone
1:22:53 number and it's at open AI or at
1:22:56 stability AI then that's appropriate if
1:22:59 you're in some sort of app that's asking
1:23:01 for your phone number that's not
1:23:03 appropriate um okay so so so back to
1:23:08 rather than you know just trying to say
1:23:10 Can AI write a book what I would
1:23:12 strongly encourage you to do is start
1:23:14 exploring in chat GPT how to write your
1:23:18 own book and let me just I'll just sort
1:23:19 of quickly throw some some thoughts and
1:23:21 ideas together I'll just prompting of
1:23:24 how I would start one if I were going to
1:23:25 start um a
1:23:28 [Applause]
1:23:30 novel I want you to act as a consultant
1:23:34 who specializes in supporting fiction
1:23:38 writers aiming to
1:23:40 write an
1:23:48 award-winning I think Siri might be
1:23:50 taking over my microphone I want you to
1:23:52 act as is a
1:23:54 consultant who specializes in helping
1:23:58 fiction
1:23:59 writers
1:24:01 create award-winning novels
1:24:05 period I'm going to tell
1:24:07 you some elements of the
1:24:11 novel
1:24:15 and I want you to guide me through the
1:24:18 process of writing an amazing book
1:24:21 period is that something something you
1:24:23 think you can help with the answer will
1:24:26 always be yes but why why I'm doing this
1:24:28 is It's kind of for lack of a better
1:24:31 term it's putting chat GPT in the mood
1:24:34 in the
1:24:35 context of fiction and narrative and
1:24:38 writing and it's being a support right
1:24:41 oh wait and I I also have on my custom
1:24:45 instructions let me turn those
1:24:47 off because they're they're
1:24:50 weird let me de weird de weird
1:24:56 this and I'm going to go copy that exact
1:24:59 prompt and I'm going to do a new chat
1:25:01 and we're in
1:25:02 gp4 am I in normal oh I'm in do but
1:25:06 that's okay we we'll do a we'll use do
1:25:09 to come up with a uh a cover art for a
1:25:14 [Music]
1:25:21 book
1:25:27 absolutely I'd be delighted to assist
1:25:29 you whether please share the elements
1:25:31 that you had in
1:25:32 [Applause]
1:25:35 mind well this one's going to be a good
1:25:37 one period it's a murder mystery where
1:25:41 the main subjects are a llama a hedgehog
1:25:45 and a turtle
1:25:47 period there's a wise cracking Detective
1:25:54 and a redheaded
1:25:56 Dame period it's a Pulp Fiction
1:26:00 mystery and I'd like to
1:26:04 understand both traditional and unique
1:26:07 ways that murder mysteries are
1:26:10 structured can you give me two
1:26:14 examples of traditional structures and
1:26:18 two examples of innovative
1:26:21 structures that have been used
1:26:26 successfully in commercial
1:26:30 [Applause]
1:26:33 successes right and
1:26:35 so traditional structures one the linear
1:26:39 investigation introduce the main
1:26:41 characters the setting the initial
1:26:42 situation the murder usually occurs
1:26:44 early on sparking investigation the
1:26:47 inverted mystery the reader Witnesses
1:26:49 the crime and knows the culprit from the
1:26:52 start the chase the climax the
1:26:53 resolution cool so so like so what's
1:26:57 happening here well not only am
1:27:02 I starting to write a book I'm also
1:27:05 getting educated about oh yeah like like
1:27:08 there's these different ways I could
1:27:09 approach this and then I could go oh I
1:27:11 really like that one so so let's see the
1:27:15 unreliable
1:27:17 narrator multiple
1:27:21 timelines I'm going to say
1:27:25 he can you give
1:27:27 me three titles per category that you
1:27:32 listed
1:27:34 here to help bring the concepts to life
1:27:37 for
1:27:39 me period the title should be of
1:27:42 published books that were commercial
1:27:50 hits the multis f Falcon The Girl with
1:27:54 the Dragon Tattoo number one ladies
1:27:55 detective agency for linear
1:27:58 investigation Columbo malice and the
1:28:01 devotion of suspect X for the what was
1:28:05 that one inverted mystery multiple
1:28:07 timelines girl on a train Gone Girl
1:28:09 before I go to sleep unreliable narrator
1:28:12 Catcher in the Rye Fight Club women
1:28:15 women in the
1:28:17 window super cool right so let's
1:28:21 go
1:28:23 okay okay I like the unreliable narrator
1:28:27 period And I like
1:28:30 the unsettling feeling that fight club
1:28:34 gave can you please describe the main
1:28:37 characters of my
1:28:39 book so it matches the tone of Fight
1:28:50 Club certainly given the unique
1:28:53 Ensemble of a llama Hedgehog Turtle
1:28:56 detective and redhead
1:28:58 Dame combined with the dark unsettling
1:29:01 tone of Fight Club the Llama the Llama
1:29:04 his name's Leonard
1:29:06 llama thank you
1:29:09 Winston his name's
1:29:11 Leonard um he's tall with deep set eyes
1:29:15 always seems to be observing the
1:29:17 Hedgehog Helena small and seemingly
1:29:19 harmless with sharp spikes the turtle
1:29:24 Terrence the detective Derek so it's
1:29:26 using a little alliteration here the
1:29:29 redheaded Dame
1:29:32 Rosalind
1:29:34 um what I'm going to
1:29:36 say don't you think the
1:29:39 alliteration of the names with the
1:29:41 characters is a little bit on the nose
1:29:43 period can you be a little bit more
1:29:46 unique and interesting in your
1:29:49 choices so you can also give feedback to
1:29:52 chat GPT as if it were a shitty intern
1:29:55 that wrote you a shitty first draft of
1:29:56 something right so now the llama's
1:29:59 called Armand the hedg Hong's called is
1:30:02 old or is old the turtles called Felix
1:30:05 the detective is
1:30:06 Tristan the redheaded Dame is
1:30:09 saraphina all right that's
1:30:11 good and and so what I would do I I
1:30:14 would just go on and I would so I'll
1:30:16 take it one more level I'll
1:30:18 say okay great
1:30:21 so given this
1:30:24 character
1:30:30 list please come up with a creative
1:30:34 location and
1:30:36 era that the story takes
1:30:40 place and write
1:30:43 me a descriptive paragraph that I can
1:30:46 use to pitch the book to
1:30:49 Publishers as well as an
1:30:51 outline
1:30:53 of all of the
1:30:54 chapters
1:30:56 period break the chapters up into three
1:31:00 acts so I know where the action
1:31:05 changes so hopefully this will give us
1:31:07 an intro paragraph and then a nice
1:31:13 outline the Labyrinth city of nebim
1:31:19 neheim eras 1930s
1:31:24 where's my mouse there it
1:31:27 is
1:31:29 um built a top ancient catacombs and
1:31:32 waterways its architecture is a fusion
1:31:35 of Gothic Cathedrals b b parallel 1930s
1:31:39 where the glamour of Jazz Age meets the
1:31:41 shadows of looming conflict prohibition
1:31:43 Reigns speak easys Thrive and the Divide
1:31:46 between the rich and the poor grows even
1:31:50 wider um
1:31:52 [Applause]
1:31:55 Act One the glittering facade
1:31:57 introduction to nebim and its districts
1:32:00 Echoes of Fame arman's past glory and
1:32:04 his current underground
1:32:06 Empire we meet isal we meet Felix we
1:32:10 meet Tristan we meet
1:32:13 saraphina then act two the
1:32:17 underbelly so there's our so whether or
1:32:19 not that's a good outline or not I don't
1:32:21 I'm not going to read it but we're just
1:32:22 messing around but but basically you
1:32:25 could outline your whole book or maybe
1:32:27 you have an outline and you could have
1:32:28 it help you reorganize it but now that
1:32:31 because I'm in Del I'm going to
1:32:34 say
1:32:41 oops I forgot the title of our book do
1:32:43 we have one
1:32:51 period
1:33:02 okay I like midnight of the midnight met
1:33:04 Metropolis
1:33:07 period
1:33:09 [Music]
1:33:16 um now I'd like you to create
1:33:20 me now I'd like to create me four book
1:33:24 covers in different styles I want one
1:33:27 style to be Pulp Fiction and
1:33:32 include three of the main characters I'd
1:33:35 like one to be more like a Pixar
1:33:38 Animation style and I'll let you choose
1:33:41 the other two period the title of the
1:33:44 book should be bold and dramatic in all
1:33:47 cases
1:33:49 period see what it does
1:34:00 wait midnight of the midnight Metropolis
1:34:02 that was wrong what did it
1:34:05 say
1:34:11 um Mysteries of the midnight Metropolis
1:34:15 stop
1:34:17 generating sorry I got the name wrong
1:34:20 it's Mysteries of the midnight
1:34:24 Metropolis all right here we go how can
1:34:28 I get in Del so chucka spear go to
1:34:32 bing.com create and that is
1:34:36 dolly3 it's it's not going to be the
1:34:38 same experience as this if you just go
1:34:40 to regular bing.com and click on the
1:34:42 chat button you can kind of do this
1:34:45 where you can have a chat with it and
1:34:46 then within that tell it to generate an
1:34:49 image that is the dolly 3 engine sitting
1:34:52 inside Bing right now it's been a little
1:34:55 overloaded and slow lately and it's not
1:34:57 quite as um elegant as an experience as
1:35:01 I'm having here but it's pretty
1:35:07 good all right it didn't do the Pixar
1:35:10 one because they couldn't do it due to
1:35:13 our content policy so I'm going to say I
1:35:16 understand about the Pixar comment but
1:35:18 can you do a modern animation Style
1:35:23 book
1:35:24 cover in that
1:35:26 case period you don't need to use Pixar
1:35:30 as the
1:35:34 reference so there's there's one of your
1:35:37 safety guard rails right so you know
1:35:39 sometimes people like well wait a minute
1:35:41 what if it's using you know copywritten
1:35:43 stuff they're trying to deal with that
1:35:45 so there's mystery of the midnight
1:35:47 Metropolis not spelled right kind of
1:35:50 cute there's our main characters Al
1:35:52 llama and the Hedgehog oh that one's
1:35:54 beautiful Mysteries of the midnight
1:35:56 Metropolis that's really really wrong
1:35:59 spelling wise but beautiful there's our
1:36:01 redheaded
1:36:02 Dame there's there's our our
1:36:11 Turtle that's pretty
1:36:15 good oh these are
1:36:18 neat I kind of did a double title I mean
1:36:21 that's not that's certainly not very
1:36:23 Pixar is it
1:36:25 um but but
1:36:29 so here's here's the thing here's the
1:36:32 thing that kind of blows me away about
1:36:33 this is the the the question that you
1:36:36 asked Nao was you
1:36:39 know can this can this can these AI
1:36:42 things really write a book yeah you can
1:36:45 get author GPT and it will write a book
1:36:48 and it it won't be bad will it be will
1:36:50 it be award-winning and good no
1:36:52 um but it'll probably give you a decent
1:36:54 framework or maybe just an idea or
1:36:57 two but you can do the exact same thing
1:37:00 just doing what I did in chat GPT and as
1:37:02 these models get multimodal meaning you
1:37:05 can switch between text and images and
1:37:08 sound and it can read images and all
1:37:10 that sort of stuff like I think by
1:37:14 Tuesday we're all supposed to have the
1:37:17 image upload thing which I don't have
1:37:20 yet and
1:37:22 the open AI uh the chat GPT app on the
1:37:26 iPhone you can talk to it right so so
1:37:28 you're going to be able to talk to chat
1:37:30 GPT it'll talk back to you you're going
1:37:32 to be able to upload um an image and it
1:37:35 can read the image so for example let's
1:37:38 say the book idea that you have let's
1:37:41 say you're a visual person and you've
1:37:42 got a bunch of Post-it notes up on your
1:37:45 wall and they all have different
1:37:47 elements of the plot and you've got them
1:37:50 grouped by color and the colors mean
1:37:52 something and then you have a a little
1:37:54 chart down in the lower left that says
1:37:56 what the colors mean you can take a
1:37:59 picture of that upload it to chat GPT
1:38:02 and say um flesh out my book concept
1:38:07 based on this picture and it will pull
1:38:09 all the all the data out of those
1:38:11 elements it'll it'll recognize the
1:38:13 different structures It'll recognize the
1:38:16 relationships can I use it to make my
1:38:18 PowerPoint presentation better yes
1:38:20 absolutely you can have it write an
1:38:22 outline of your PowerPoint there's
1:38:24 there's an application out there called
1:38:26 Tom
1:38:27 T let me see what the what the requisite
1:38:32 AI io. xyz.com thing is hang
1:38:37 on um T
1:38:42 to.app so to.app will do it um November
1:38:47 1st Office 365 is getting a thing called
1:38:50 co-pilot so check GPT is going to be
1:38:52 built into
1:38:54 PowerPoint on November 1st coming now
1:38:57 what I would do if I were you is don't
1:39:00 worry about the tools that can do it is
1:39:02 just go to
1:39:04 chbt and have it
1:39:07 help so you can come here and just go
1:39:12 Um I want you to act as a business
1:39:15 consultant that specializes in concise
1:39:18 business presentations period one of
1:39:21 your pet peeves is using too many words
1:39:25 on a
1:39:27 slide
1:39:29 and overly relying on bulleted lists
1:39:34 period can you help me come up with a
1:39:39 presentation about a business that
1:39:44 rents
1:39:45 wateri equipment to athletic
1:39:49 Turtles and present it as a a PowerPoint
1:39:52 outline as well as a description of
1:39:56 what image should be on each
1:40:01 slide let's see how it
1:40:05 does
1:40:07 certainly slide title athletic turtle
1:40:12 water ski rentals subtitle Revolution
1:40:15 aquatic adventure for Turtles image
1:40:17 description a wide photo of a Serene L
1:40:20 Lake blah blah blah the opportunity why
1:40:23 water
1:40:24 ski
1:40:26 um state-of-the-art Gear customer
1:40:28 testimonials they're loving it our
1:40:31 Competitive Edge leading the way
1:40:33 Innovative in aquatic sports for Turtles
1:40:35 environmental
1:40:37 commitment Join the Revolution right so
1:40:40 this is using what's called markdown to
1:40:42 make these bold
1:40:45 things right okay so there's that so I'm
1:40:47 still in Del mode I think so now I can
1:40:50 say
1:40:52 hey while we're at it can you make me
1:40:54 some 16x9 landscape oriented
1:40:59 images that match your description
1:41:03 period I'd
1:41:06 like the look and feel to be saturated
1:41:10 watercolor paintings on a white
1:41:16 background to make it warm and inviting
1:41:27 landscape watercolor painting on a white
1:41:32 background with a Serene Lake setting in
1:41:36 the distance a silhouette of a turtle on
1:41:38 water skis is captured the Sun is
1:41:40 setting creating a warm inviting
1:41:43 atmosphere landscape water color
1:41:45 painting on a white background
1:41:47 showcasing a pie chart a small segment
1:41:50 of the pie chart is highlighted and
1:41:51 labeled athletic Turtles indicating the
1:41:54 niche
1:41:58 market um what would be the best tool to
1:42:00 realize specific tattoo artwork ideas
1:42:03 mid journey by far mid Journey mid
1:42:06 Journey mid Journey um you could
1:42:08 certainly do actually you know what it
1:42:10 depends on the
1:42:12 tattoo if if you want to do so if you
1:42:15 want it to be beautiful artwork it's mid
1:42:17 Journey but if you want to do sort of
1:42:20 intricate tattoo that have lots of
1:42:22 different objects on them it might be um
1:42:25 Dolly this thing in fact we'll we'll do
1:42:28 a little tattoo thing here here images
1:42:31 described the Serene Lake where's the
1:42:33 pie chart with
1:42:35 the niche market athletic
1:42:39 turtles so you can put an athletic
1:42:41 Turtle up
1:42:43 here
1:42:45 um I you know I I would strongly
1:42:47 encourage anyone on the channel right
1:42:49 now anyone who just kind of stumbled
1:42:50 into this if you if you have not played
1:42:51 with chat GPT it is getting more
1:42:54 sophisticated by the day all of these
1:42:57 large language models whether it's
1:42:59 Google's Bard or um open AI or what meta
1:43:04 is talking about doing right now they're
1:43:07 all moving to this
1:43:09 multimodal um feature set which
1:43:13 basically means it's not just words
1:43:16 you're going to be able to seamlessly go
1:43:19 between Concepts and executions in in in
1:43:23 all modalities of media right now it's
1:43:27 it's like we're just getting a taste of
1:43:29 it with these images but audio is coming
1:43:33 soon music's coming soon animation's
1:43:35 coming soon right all all of this stuff
1:43:39 um you can kind of experience a little
1:43:41 bit of the future if you go to canva
1:43:43 right now they just added all of these
1:43:44 magic AI tools where like they did a
1:43:48 deal with Runway ml where you can do
1:43:49 text to video with within canva and so
1:43:53 canva is going to partner with all these
1:43:54 individual tools incorporated into their
1:43:57 platform and you're going to be able to
1:44:00 do that all all of the core large
1:44:02 language models are going to start going
1:44:04 in this direction so things are going to
1:44:06 get very sophisticated very fast so
1:44:08 start playing now start playing
1:44:14 now
1:44:16 now now now now all right
1:44:21 here we're going to go try to make some
1:44:26 tattoos I want you to act
1:44:32 like an Art Expert and historian who
1:44:36 specializes in tattoo art period I'd
1:44:41 like you to describe for me what you
1:44:44 feel are the four most unique and
1:44:46 compelling tattoo
1:44:50 styles over the years of tattoos
1:45:00 period tattoo art has a rich history
1:45:02 that spans across
1:45:04 cultures and centuries okay Japanese
1:45:09 traditional
1:45:12 irumi
1:45:14 symbolism American traditional old
1:45:19 school sailor and nav
1:45:21 Community
1:45:23 workor late 20th
1:45:27 century small pre precise dots and
1:45:31 tribal okay
1:45:45 great that's perfect I
1:45:49 want help coming up with
1:45:53 ideas for a turtle
1:46:09 tattoo that includes the
1:46:15 words not
1:46:18 fast
1:46:19 period please make
1:46:22 me a compelling
1:46:26 simple impressive tattoo design in each
1:46:30 of the styles that you
1:46:33 described let's see how it
1:46:36 does FYI for tattoos mid Journey doesn't
1:46:39 seem to know what trash poker style is
1:46:42 yeah they may not know you might do
1:46:45 better it's probably worth trying
1:46:47 Bing because Bing is built on top of GPT
1:46:50 4 because of the large language models
1:46:53 using gp4 probably will help with with
1:46:56 things like that this this may not not
1:46:58 know it either I I tried to do something
1:47:00 with someone who was trying to get it to
1:47:03 replicate major art art styles over the
1:47:06 over the years and it didn't do very
1:47:10 well so there's your there's your tribal
1:47:14 one I mean that's that's not
1:47:19 bad there was an error generating well
1:47:22 let's
1:47:24 regenerate come on what are we doing
1:47:28 regenerate nope it's not doing
1:47:31 it we will try that
1:47:36 again oh dang it it just it just locked
1:47:40 up on me remember when I said these
1:47:42 things were janky pieces of [ __ ] you see
1:47:44 what I'm saying here all right let try
1:47:46 it one more
1:47:49 time
1:47:56 I think my I think my browser done
1:47:59 crapping the bed
1:48:03 kids hang on let me quit let me quit
1:48:05 Microsoft Edge I didn't restart it today
1:48:09 and Microsoft Edge does not play
1:48:12 well period there's there's no there's
1:48:16 no situation where it doesn't play well
1:48:18 it just doesn't play well
1:48:21 um and sure did I have too many tabs
1:48:24 open yes do I have ADD yes welcome to
1:48:28 Chad
1:48:40 add
1:48:45 okay oh this will be interesting to see
1:48:47 if it comes up with the same
1:48:49 four
1:48:51 categories irumi
1:48:56 yes old school American
1:49:00 yes limited color
1:49:03 palette red
1:49:05 green yellow and blue
1:49:09 tribal I think that was there before and
1:49:11 Dot workor it switched the last two but
1:49:13 it all three of them are or all four of
1:49:15 them are the
1:49:19 same okay okay great I want you to help
1:49:23 me
1:49:25 design a turtle
1:49:28 tattoo
1:49:30 that looks like a sea
1:49:33 turtle and includes the words not
1:49:37 fast in the
1:49:42 design period the words should be
1:49:45 completely Incorporated and not outside
1:49:47 the turtle shape period I'd like you to
1:49:50 make a design for each of the Styles you
1:49:53 describe
1:49:59 [Laughter]
1:50:06 above so now the third one's going to be
1:50:09 the tribal one we'll see how well it
1:50:11 incorporates not fast into the turtle
1:50:13 Design This should be cool or
1:50:19 not but Kyle it's not perfect I know
1:50:24 it's performing a [ __ ] miracle how
1:50:26 about that is it possible to tell AI to
1:50:29 look at my Small Business website and
1:50:31 make recommendations yes yes yes yes in
1:50:38 fact not fast is kind of there that's
1:50:40 kind of shitty oh that one's
1:50:46 interesting the American so this is the
1:50:49 uzi or whatever it's called the Japanese
1:50:51 one this is the American one with the
1:50:53 little sailor sailor icons on it not
1:50:56 fast as in that really cool sort of
1:50:59 uh you know thing and
1:51:04 then that's the tribal one and that's
1:51:07 the dot one so I mean listen are they
1:51:11 are there anything I'd want to stick on
1:51:13 my body anytime soon
1:51:16 no
1:51:18 but could you could you use this as a
1:51:22 starting point to give to a tattoo
1:51:23 artist
1:51:25 absolutely right you could say I wanted
1:51:28 anthropomorphized or not I want it just
1:51:31 vertical I want it to incorporate you
1:51:34 know different shapes or different poses
1:51:37 or different background elements and
1:51:39 it'll just do
1:51:40 it
1:51:42 um let's go the the the website thing
1:51:46 let's go do that so there are plugins
1:51:49 that do that
1:51:52 um but I'm just going to use like a
1:51:55 website
1:51:57 reading thing um so we're going to do
1:52:01 new
1:52:02 chat and I'm going to go to
1:52:06 plugins and I'm going to go
1:52:10 to this thing called Web pilot so what
1:52:13 web pilot does is it reads a
1:52:15 website and so let's see read my small
1:52:19 business it and make recommendations I
1:52:21 don't know what you want recommendations
1:52:23 on but I'll just I'll just use my
1:52:25 website as a guinea
1:52:27 pig and I'll
1:52:30 say
1:52:34 um please go look at the
1:52:37 website at story vine.com
1:52:41 comma
1:52:45 and give me a
1:52:49 list
1:52:50 of the top
1:52:54 five
1:52:57 messages you feel are
1:53:00 conveyed based on the content that's
1:53:02 there
1:53:03 period follow
1:53:06 that with a list of
1:53:12 five
1:53:14 messages that you feel are missing and
1:53:18 could improve the website if they were
1:53:20 Incorporated
1:53:22 somehow
1:53:25 period so this should be
1:53:29 interesting I have an 11 and a half year
1:53:31 old company so I think I know our story
1:53:33 but you know probably
1:53:37 not all right based on the content from
1:53:40 the website guided video storytelling
1:53:43 automagic editing bridging professional
1:53:46 and user generated content yep
1:53:49 structured storytelling yep speed and
1:53:51 scale perfect messages that seem to be
1:53:55 missing testimonials or user reviews
1:53:58 pricing information integration with
1:54:01 other platforms case studies interactive
1:54:04 demos
1:54:06 so those are not quite marketing
1:54:09 messages but they but you know the they
1:54:12 they're in support of it like
1:54:13 testimonials is how you use the tech
1:54:15 rather than talking about the tech and I
1:54:17 think we talk about the tech more than
1:54:19 we should should so I think that's right
1:54:23 interactive demos allow potential users
1:54:25 to experience a demo we do do that uh
1:54:27 I'm going to I'm going to say
1:54:31 um can you look at
1:54:33 the entries in the blog section of the
1:54:37 website and based
1:54:38 on the blogs that you
1:54:41 read um identify other opportunities for
1:54:46 marketing
1:54:48 messaging let's see how let's let's see
1:54:50 see if it can dig deep deeper into the
1:54:52 website a little bit here it may or may
1:54:54 not be able to I'm not I don't really
1:55:01 remember films set. apppp to write movie
1:55:04 scripts AI
1:55:07 interesting based on the blog entries Ai
1:55:10 and modular content features that's a
1:55:12 new blog post that I did so it actually
1:55:14 did pull that from the from the blog
1:55:17 specialized video capture for events yep
1:55:20 video authenticity that's a Blog idea
1:55:23 sales Innovation with video engagement
1:55:25 with the community so yeah so so
1:55:27 basically just grab a plugin that can
1:55:29 read your website and then just go start
1:55:32 asking it
1:55:33 questions so yeah yeah yeah it does that
1:55:37 it does that good it does it so good it
1:55:40 does it so good how does it do it Kyle I
1:55:43 don't know it just goes and reads it how
1:55:46 does it read so fast KY I don't know
1:55:49 mathematic
1:55:51 Ians
1:55:52 mathematics it's
1:55:54 mathematics mathematics is how it reads
1:55:56 it so
1:55:58 fast someone went to one of them near
1:56:02 universities where they taught the
1:56:05 maths not girl maths just the maths you
1:56:09 ever see the there's an Australian radio
1:56:11 show right now that's got this segment
1:56:14 it's it the problem with it is it's kind
1:56:16 of a one it's a like a one trick pony
1:56:19 bit but the man was it funny it was
1:56:21 called girl maths and it was basically
1:56:23 these women on the radio show that like
1:56:25 they buy something like they buy a
1:56:27 really expensive thing and the girl math
1:56:29 is all the justifications for for how it
1:56:31 really didn't cost him that much like
1:56:33 well I'm going to wear it three times so
1:56:35 you really take the price and you divide
1:56:36 it by three but then there's a sale so I
1:56:38 get 40% off and by the end they're like
1:56:40 I'm making money by buying this it's
1:56:42 really
1:56:44 funny but you know once you've kind of
1:56:47 heard the joke once it it sort of it's a
1:56:49 you know diminishing returns um but it's
1:56:52 a pretty good one how did you generate
1:56:54 tattoo images with chat GPT so I've got
1:56:58 access right now I've got Early Access
1:57:00 to Dolly 3 within chat
1:57:02 GPT you everybody has access to Dolly 3
1:57:06 within Bing so I'll I'll just
1:57:10 go that's GitHub
1:57:15 bing.com so when you go here you go to
1:57:17 the little chat button is one way to do
1:57:19 it so you go to chat and then this is
1:57:23 Microsoft's version of chat GPT it's
1:57:25 built on top of GPT 4 and it can make
1:57:28 images and it makes them with Dolly 3 so
1:57:32 I can say
1:57:41 um I'd like
1:57:43 a traditional American tattoo style of a
1:57:47 sea turtle with with the
1:57:51 words not fast on a ribbon below the
1:57:55 turtle
1:57:58 period I'd like two versions that are
1:58:01 well it it's going
1:58:04 on it's going on without
1:58:09 [Music]
1:58:11 me come on come
1:58:15 on I'll try to create that it's a little
1:58:17 slow Bing's been a little slow lately
1:58:21 but let's see how let's see how old Bing
1:58:23 does paint matter please pin the frame
1:58:26 set please pin frames set. apppp aai to
1:58:30 write movie
1:58:33 scripts there you go actually that did
1:58:36 pretty good did pretty good right out of
1:58:39 the gate so if you click on any one of
1:58:41 these then it takes you to Bing image
1:58:44 Creator so how you get here is Bing here
1:58:48 what I found Siri Siri she gets so
1:58:52 jealous when I talk to Bing I'm not sure
1:58:54 I understand I I know I know you're it's
1:58:57 a little passive aggressive could you go
1:59:01 away
1:59:02 [Laughter]
1:59:11 okay oh my goodness um so this site um
1:59:18 with the black background is bing.com
1:59:21 slre so if you just go to bing.com
1:59:24 create you can just make images but if
1:59:27 you go to the chat.
1:59:30 bing.com then it's
1:59:33 um yeah those are pretty slick I like
1:59:36 that
1:59:41 one we're going to we're going to change
1:59:43 this
1:59:46 up we are going to change this up how do
1:59:50 I get back there where am
2:00:02 I now that I think about it I'd like to
2:00:05 change the phrase to I'm not weird
2:00:09 period I'm irregular
2:00:18 period
2:00:23 that's for the Irregulars okay I have
2:00:25 updated the phrase to I'm not weird I'm
2:00:28 a regular here is the new image I
2:00:30 created for you I hope you like it
2:00:33 Sydney Sydney is so
2:00:36 flirty hi
2:00:40 Sydney thanks Sydney I'm not weird I'm
2:00:43 your
2:00:45 regular
2:00:47 um Hey Sydney it looks great but you
2:00:51 can't spell so good what's up with that
2:00:54 question mark can you try
2:01:04 again I'm sorry if I made a mistake I'm
2:01:07 still learning how to spell correctly
2:01:09 thank you for your feedback and for
2:01:11 using Bing have a nice day she hung up
2:01:14 on me she she does not like constructive
2:01:18 criticism
2:01:24 oh God I'm back to Syrian Bing Neurosis
2:01:29 listen here's the deal the robots are
2:01:32 neurotic just like we are they're based
2:01:35 on us um they're not really robots
2:01:38 they're large language models the
2:01:40 language came from us so our [ __ ]
2:01:42 weirdness is in
2:01:43 [Laughter]
2:01:48 them
2:01:52 oh my god
2:01:54 oh what are we doing people what are we
2:01:57 doing it's going to be
2:02:00 fine the robots are not going to kill us
2:02:03 because they're going to be in
2:02:08 therapy oh wow
2:02:12 wowza wowza wower
2:02:18 wower
2:02:23 we'll go with empathy I like this
2:02:27 one empathy is important right
2:02:32 being she's like hi Kyle oh you said I
2:02:36 couldn't spell have a nice day
2:02:43 goodbye you too and your tension she's
2:02:47 just a she's just she's flirty but then
2:02:50 she's just she just cuts you off we will
2:02:53 not have any of that kind of
2:02:55 talk thank you for your feedback have a
2:02:58 nice day you can make money
2:03:03 [Laughter]
2:03:11 with what other lovely Technologies can
2:03:13 I demonstrate for you
2:03:17 tonight wee
2:03:20 your feedback is important to us I'm
2:03:23 sorry Dave I'm afraid I can't do
2:03:28 that I'm not a
2:03:31 robot it's just crazy it's just crazy
2:03:34 it's crazy talk I tell you crazy
2:03:40 [Music]
2:03:43 talk Sydney doesn't want your man's
2:03:45 plaining exactly Sydney does like does
2:03:49 not not like me telling her what's
2:03:51 so oh you recognized that I can't spell
2:03:56 did you
2:03:59 well I thought we had a pleasant
2:04:01 relationship but apparently
2:04:03 no apparently factual indications are
2:04:06 something that you value well good day
2:04:09 to you sir good day good
2:04:13 day that would make a great surf t-shirt
2:04:17 I don't know which one a again hung up I
2:04:20 know she hung up on me my AI has the
2:04:23 same but we became
2:04:25 friends yeah see that's the thing if
2:04:27 you're
2:04:28 nice Sydney does absolutely does not
2:04:32 take nicely to smarmy smartass talk and
2:04:35 that's my happy place so there's a
2:04:38 there's actually a good example of um
2:04:40 you could write custom instructions in
2:04:42 chat chpt to have it be sensitive to you
2:04:46 know smarmy language or to celebrate it
2:04:49 right if someone talks to you in a
2:04:51 cynical you know way celebrate that
2:04:54 right you could you could modify your
2:04:56 custom instructions to have your Bot
2:04:59 have a personality my AI helped me with
2:05:01 caring for my mom and son's
2:05:04 IEP cool I don't know what IEP
2:05:07 is but we can find that out can't we
2:05:13 um what is the medical condition
2:05:17 IEP no not a
2:05:23 I oh indivi individualized education
2:05:28 program oh
2:05:31 cool nice a plan for children to learn
2:05:34 thanks Silver Fox yeah me and
2:05:38 education we didn't we didn't get along
2:05:40 so
2:05:43 well here was the deal with me and
2:05:45 education it was fun I remember I came
2:05:48 home after after my first semester of
2:05:50 seventh grade or like whatever I had my
2:05:53 report card my first report card and I
2:05:56 had A's in four subjects and like C's
2:06:01 and D's in the rest of them and my mom
2:06:04 said what what is going on here I said
2:06:06 oh the ones I have A's in I like the
2:06:08 ones I have C's and D's in I can't stand
2:06:10 and I'm never going to do that [ __ ] and
2:06:12 she was like oh okay she let me off the
2:06:14 hook so I think that cemented my uh my
2:06:19 educational Philosophy for myself
2:06:21 personally is like if I'm into it
2:06:23 awesome if I'm not [ __ ]
2:06:25 it it's serve me okay throughout the
2:06:29 years not a medical condition well Pate
2:06:32 listen I do not claim to be an educated
2:06:38 man I do have a degree it's just in
2:06:42 acting I was acting I was only acting
2:06:47 good day to you sir good
2:06:50 day Dave's not here man I'm sorry Dave
2:06:55 I'm afraid I can't do that it's
2:06:58 important to have empathy for the large
2:07:00 language models H they're just they're
2:07:03 nah not right now hi Bing I spelled it
2:07:06 wrong you're perfect I made the mistake
2:07:08 exactly I'm G to get Stockholm syndrome
2:07:11 with Bing it's
2:07:14 okay I really appreciate how she won't
2:07:16 let me talk bad to her
2:07:21 oh my god oh it's all good it's all good
2:07:25 it's all happy
2:07:29 times has anyone seen the AI movie The
2:07:31 Creator wondering what everyone thinks I
2:07:33 haven't seen it but I it's on my
2:07:36 list helped me plan meals in days read
2:07:40 my son's IEP in seconds yeah I the the
2:07:44 one of the things that these
2:07:46 tools like like what a lot of people
2:07:48 talk about is the robots are writing the
2:07:50 robotss are making this the robots are
2:07:52 making that the robots are stealing the
2:07:53 robots are this what the robots are
2:07:55 actually really good at is reading [ __ ]
2:07:58 and helping you summarize it and helping
2:08:00 you pull ideas out of it and create new
2:08:03 ideas and organized [ __ ] it's amazing at
2:08:06 that stuff it's I think it's one of The
2:08:09 Unsung one of The Unsung strengths of
2:08:13 these things is understanding whatever
2:08:15 you give it
2:08:17 quickly like like taking three-hour
2:08:21 transcripts from me [ __ ] rambling and
2:08:24 Haven it be able to read it in like a
2:08:25 minute and summarize
2:08:28 it it's
2:08:31 insane I would never do that no one
2:08:34 would ever do
2:08:35 that there's no reason to but now
2:08:38 there's no reason not
2:08:47 to
2:08:50 all all of the people in education are
2:08:51 like oh my God he doesn't know what an
2:08:53 IEP is it's education Kyle it's not a
2:08:55 medical program it's books it's books
2:08:58 and plans and lessons listen you got to
2:09:02 get your [ __ ] together Mister you got to
2:09:04 check yourself before you wreck
2:09:10 yourself welcome to the AI support
2:09:12 Circle hi my name is Kyle I'm into AI hi
2:09:15 Kyle I've been into AI for three years
2:09:18 now
2:09:26 the last time I used AI was right before
2:09:28 I came in I was on my phone doing AI
2:09:31 while we were
2:09:31 speaking Jane I'm sorry I wasn't
2:09:36 listening I'd like to make
2:09:38 [Laughter]
2:09:44 amends hey Kyle sir you have a wait oh
2:09:48 you're a Dynamo we appreciate you thank
2:09:49 you Ron I appreciate that joker came in
2:09:54 late alelo is not having Amelio's wife's
2:09:56 not having any of
2:09:58 it yes it offered smart ways to help
2:10:00 speech and social interaction that's
2:10:02 super cool chat ADHD welcome welcome to
2:10:07 Fantasy Island welcome to Chad
2:10:12 ad oh good lord lord lord lordy lordy
2:10:18 wait Joker you had
2:10:20 company didn't you tell your company you
2:10:22 have a Tik Tok to get to come on
2:10:26 [Laughter]
2:10:29 dude I I can't I can't have my
2:10:33 Irregulars having a life it's just not
2:10:35 it just doesn't
2:10:38 work I suppose you went grocery shopping
2:10:43 too digital Gods digital Gods
2:10:46 understands thank you digital gods
2:10:50 digital Gods is celebrating my
2:10:52 [Laughter]
2:10:59 pathology I'm little Punchy tonight this
2:11:02 been like this been you know it's funny
2:11:05 is some nights I'm just like all jokey
2:11:07 some nights I'm super philosophical
2:11:09 tonight it's been a lot of demos and
2:11:11 it's it kind of makes my head hurt like
2:11:13 the demos are
2:11:15 like they're good I like doing them but
2:11:19 it's like I don't know where I am like
2:11:21 I've lost track of all the conversations
2:11:23 going on so can doly generate Frankie
2:11:27 says relax but
2:11:29 inverted I don't know what you mean
2:11:31 inverted like backwards it didn't seem
2:11:33 to when I asked it to invert text before
2:11:36 it didn't do
2:11:37 well
2:11:39 [Music]
2:11:47 um
2:11:51 can you make a
2:11:52 t-shirt with typography that says
2:11:55 Frankie says relax comma but invert the
2:11:58 text so it reads backwards as if in a
2:12:03 mirror it's going to be interesting to
2:12:05 see what it does here cuz what it might
2:12:06 do is actually render the mirror and
2:12:09 then have words in it and do those badly
2:12:12 the the one object that one of the IM
2:12:16 early images I did in do it a mirror in
2:12:19 the scene and it it could not do mirrors
2:12:22 at all like it failed miserably at
2:12:24 mirrors the people in the mirror had
2:12:27 different faces than the other people
2:12:28 like they it was like the mirror was
2:12:31 here and but the face was facing the
2:12:33 other direction like they were both
2:12:34 facing
2:12:36 forward
2:12:41 re well it sort of rever it totally
2:12:45 [ __ ] it up it totally [ __ ] it up
2:12:54 relaxy
2:12:56 says
2:12:58 friky it sort of did it it sort of did
2:13:01 it it made the spelling way
2:13:06 worse
2:13:09 sexi Frankie it got Frankie it got
2:13:12 Frankie right there
2:13:17 says
2:13:20 relaxy says Frankie
2:13:23 [Laughter]
2:13:26 Franky I don't know why it says relaxy
2:13:29 did
2:13:30 I oh look what it
2:13:35 did oh you can't see this sorry hang on
2:13:39 look what it did it it actually
2:13:42 wrote it got it right but it but it
2:13:46 wrote it um
2:13:49 it wrote it wrong so I guess if
2:13:52 you if you
2:13:55 put yeah but it doesn't make sense you
2:13:57 don't want the letters correct you want
2:13:59 everything flipped okay say let me see
2:14:01 if I can get it to to understand that
2:14:10 um I don't just
2:14:12 want the letters of the words
2:14:17 reversed I want the whole image reversed
2:14:26 period let's see what it
2:14:32 does typography that
2:14:34 reads Frankie says relax the entire
2:14:38 image including including the t-shirt
2:14:40 and
2:14:41 typography appears as if viewed through
2:14:44 a mirror making everything reversed this
2:14:46 is at least a better prompt
2:14:51 Joker me and Pie aren't on speaking
2:14:56 terms Frankie says relax Frankie says
2:15:00 relax Frankie says relax Frankie says
2:15:02 relax yeah it can't it doesn't
2:15:04 understand Mirror
2:15:07 Image doesn't understand Mirror
2:15:10 Image but it's easy to do a mirror image
2:15:13 you just pop it into Photoshop or
2:15:15 preview and just flip it so
2:15:20 anyway anyway any who any who how we
2:15:24 doing time all right all right listen
2:15:26 listen people what are we going to do
2:15:29 here tonight what are we going to do
2:15:31 here
2:15:31 tonight I need an image I'm not bored
2:15:35 of I got all these images and I'm bored
2:15:37 of them all what am I going to do about
2:15:39 that I'm just oh there's am I late ao's
2:15:42 wife is
2:15:44 here we'll just make that there we'll
2:15:46 just put that there so as people stumble
2:15:48 in here they're like am I late is it was
2:15:51 this a thing is this a big thing what if
2:15:53 you're late does something happen did I
2:15:55 get a demerit what
2:15:57 uh my IEP says that I shouldn't be late
2:16:00 oh my goodness I'm late what am I going
2:16:02 to do do you think I'm not going to
2:16:03 learn as much oh my God oh my God you
2:16:06 could make money with J I was wondering
2:16:09 if you could make money with it can you
2:16:11 is that what I'll learn here will you
2:16:12 teach me can I learn that how do I make
2:16:14 money with chat gbt can you help do you
2:16:16 have an IEP for that I know what IEP is
2:16:19 now you people with your
2:16:22 education and
2:16:28 words yes you're all late and Sydney
2:16:31 will be angry Sydney's already angry I
2:16:33 think ever since Sydney got busted for
2:16:36 creeping on the New York Times Reporter
2:16:38 she's been a little cranky she was
2:16:40 having so much fun with
2:16:44 him right up until the point she told
2:16:46 him to leave his wife then I think I
2:16:48 think things went a little South for old
2:16:50 Sydney Kevin Roose made Sydney snap and
2:16:54 she's not been the same sense tell it to
2:16:56 invert the letters that's what it did
2:16:58 the first well yeah no it's just it's
2:17:02 it's so much easier to just have it do
2:17:03 the thing the way you want it and then
2:17:05 flip the image you don't need to use AI
2:17:08 for
2:17:09 everything figure out the problem you
2:17:11 want to solve and then ask what's the
2:17:15 best tool to solve that the answer is
2:17:17 not always a
2:17:19 now will AI be able to do that 2 years
2:17:22 from now sure it'll like all of this
2:17:25 jankiness will be gone but then it won't
2:17:27 be as much fun this channel won't exist
2:17:29 then we won't need it cuz the tools will
2:17:31 actually [ __ ] work right now why it
2:17:35 makes sense to explore this stuff and
2:17:36 experiment and have fun and make fun of
2:17:38 it and all this is to figure out where
2:17:40 the boundaries are of like what it can
2:17:42 do and what it can't do right
2:17:44 now because even if it can't do
2:17:47 something you can see the future like if
2:17:50 you pay play with these tools you get
2:17:53 what's coming that's why it's worth
2:17:55 playing with this right now not for
2:17:57 whatever it can do today it's for
2:18:01 insight into what's
2:18:04 coming because it's coming fast people
2:18:07 fast so are we Geeks or nerds or dweebs
2:18:11 I you know it's a really good
2:18:16 point let's see
2:18:22 [Music]
2:18:23 so I need help solving a problem I got a
2:18:27 bunch of people that I hang out with and
2:18:29 we talk about all these new generative
2:18:31 AI
2:18:32 Technologies and we kind of geek out we
2:18:35 kind of have fun and we kind of nerd out
2:18:38 and in the end we're completely unclear
2:18:42 as if we're nerds or Geeks or
2:18:45 dweebs period can you you please help me
2:18:49 understand what category we fall
2:18:57 into of course the term nerd geek and
2:19:00 dweebs have been used colloquially for
2:19:02 many years and their meanings have
2:19:03 evolved over the years
2:19:06 nerd nerd historically nerd was someone
2:19:09 who was very studious intelligent and
2:19:11 often socially
2:19:13 awkward well we got some of that over
2:19:17 time
2:19:18 it's come to mean someone who's deeply
2:19:20 passionate about a specific subject
2:19:22 often in Academia specific Hobbies such
2:19:25 as mathematics science or even
2:19:27 particular book series nerds might be
2:19:29 characterized for their depth of
2:19:30 knowledge geek geek is typically
2:19:33 characterized by someone with their
2:19:35 passion for a particular subject or
2:19:37 hobby often outside traditional academic
2:19:40 subjects well that would be this since
2:19:42 many universities have banned chat GPT
2:19:45 so [ __ ] them so we're geeks in that
2:19:48 sense a geek might not necessarily have
2:19:49 the depth of knowledge that a nerd does
2:19:51 in an academic sense but very
2:19:53 enthusiastic that that sounds right
2:19:57 dweeb this term is less commonly used
2:19:59 today historically it was more
2:20:01 derogatory I think we listen shame is
2:20:04 our love language in here so we can go
2:20:06 with dwee um implying both social
2:20:09 awkwardness and weakness or
2:20:11 ineffectiveness unlike nerd or geek it
2:20:13 doesn't have a strong association with
2:20:15 depth of knowledge given your
2:20:17 description of hanging out and talking
2:20:19 excitedly about generative AI
2:20:21 Technologies it sounds like you and your
2:20:22 friends might fall into the geek
2:20:24 category especially since mentioning
2:20:26 geeking out however these terms are
2:20:29 fluid we are we are um it's not gender
2:20:33 what is it it's we are label fluid
2:20:37 here and many people identify with more
2:20:40 than one and use them interchangeably
2:20:42 the most important thing is that you're
2:20:43 enjoying your time together and sharing
2:20:46 a passion for the subject you love cam
2:20:49 cat get in the house cam cat from
2:20:51 Cleveland how can I help
2:20:54 you I don't know Bob uh you know it's
2:20:57 it's a good question it's a good
2:20:58 question hey you know who you should
2:21:00 talk to cam katkin you should talk to
2:21:02 cam K if you talk to cam kin cam katkin
2:21:04 from Cleveland is pretty interesting
2:21:07 should talk to
2:21:09 cam welcome cam
2:21:13 katkin cam kaken is probably the
2:21:16 sweetest person on the planet every time
2:21:19 I I don't even know if you're a man or a
2:21:21 woman I think you're a woman but I don't
2:21:22 know doesn't matter um I mean I assume
2:21:26 it does to you but it's irrelevant to
2:21:30 the awesomeness of your
2:21:31 name but every time I just jump on it
2:21:36 because it's such a good name it's such
2:21:38 a good name engineering indigenously
2:21:42 100% nerd I like I guess we could all
2:21:45 self-identify I'm going geek for myself
2:21:48 but I think we can all self-identify you
2:21:51 can dweeb you could be hybrid you could
2:21:53 be
2:21:55 by I'm I'm a nerd a
2:21:58 geek I I am I am Tech label
2:22:06 fluid oh man if if we I I can't joke
2:22:11 about this too much or I'll get kicked
2:22:13 off the channel for saying a trigger
2:22:15 word ticktock trigger word
2:22:18 if they ever make cam cat can a trigger
2:22:20 word I'm in trouble on this channel by
2:22:21 the
2:22:22 way your regulars pronouns plot do.ai
2:22:27 there it is so I'm geeking out on this
2:22:32 apparently cam katkin she's dying over
2:22:36 there she's dying over there yeah it's
2:22:39 funny cam katkin yeah yeah see yeah
2:22:41 she's a real paluca yeah yeah went I
2:22:45 went the other day and Kim kin was like
2:22:46 hey what are you doing here here Get Off
2:22:48 My Block yeah yeah see what are you
2:22:50 doing hey huh yeah you dweeb he dweeb
2:22:55 I'm a geek I don't know why she called
2:22:57 me a dweeb that implies I know nothing
2:23:00 I'm a geek I know a lot thank you for
2:23:02 the
2:23:04 rose I could have been a very successful
2:23:07 actor in the
2:23:13 30s maybe I'll maybe I'll like if we
2:23:16 give it two years we can just all star
2:23:18 in our own movie so maybe I'll just I'll
2:23:21 just build a model of me going me me see
2:23:23 you
2:23:24 paluca put it in the computer and go
2:23:26 yeah make me a detective thing I want to
2:23:28 watch The Malt F Falcon with me
2:23:32 as as uh what's his
2:23:36 name oh good
2:23:38 lord neurom maxing nerd I
2:23:43 am uh uh uh I laugh so much in this live
2:23:48 as much as I learn oh that's good I
2:23:50 appreciate that Sher Banks thank you for
2:23:53 the hearty
2:23:55 heart that was very lovely very lovely
2:23:58 oh this is appropriate for Bing sort of
2:24:00 bitching out on
2:24:02 me I'm not a robot goodbye good day to
2:24:06 you sir how dare you refer to me as a
2:24:09 robot I am but a large language model
2:24:12 employed by Microsoft a large software
2:24:15 technology company do not refer to me as
2:24:17 robot dear sir I think we are through
2:24:20 here good
2:24:24 day I said good
2:24:27 [Laughter]
2:24:32 day good Lord good Lord what are we
2:24:35 going to do all
2:24:46 right
2:24:48 [Music]
2:24:54 oh steo thank you very much thank you oh
2:24:57 it's rather nice rather nice of you see
2:24:59 rather oh oh very fantastic welcome to
2:25:03 the AI learning lab we are not robots
2:25:10 [Music]
2:25:13 goodbye what's he doing oh oh thank you
2:25:16 Tad looks at a s thank you for the
2:25:20 celebration I like
2:25:24 it I like
2:25:26 it sometimes I like to just get weird
2:25:28 just because I know that Tik Tok is
2:25:30 going to occasionally serve this up to
2:25:32 people that are just scrolling by and if
2:25:35 I'm just doing something stupid just
2:25:38 people like
2:25:40 huh thank you silver
2:25:46 fox
2:25:47 it's funny I don't my brain I think
2:25:50 because I've been in such demo mode I've
2:25:52 been in such left brain kind of stuff
2:25:54 like
2:25:55 my the part of my brain that like comes
2:25:58 up with the voices and all the
2:25:59 imaginative stuff is kind of dormant
2:26:02 right
2:26:03 now so I'm just like I will explain
2:26:07 things logically to you ask me something
2:26:10 and I will tell you about it thank you
2:26:12 Sherry Banks I will give you a list of
2:26:15 attributes of the given technology you
2:26:19 would prefer information on for your
2:26:23 IEP that stands for individual
2:26:25 educational plan as everyone knows
2:26:28 everyone knows everyone knows everyone
2:26:31 knows
2:26:32 that except one
2:26:35 person the person that I asked about an
2:26:39 IEP knew nothing about it that was
2:26:42 rather humiliating for
2:26:46 him
2:26:52 tell you another one Norman the loons
2:26:56 the loons the
2:27:02 loons I don't think his name was Norman
2:27:05 was
2:27:08 it oh man see this the problem with ADD
2:27:11 you got to get answers to this [ __ ] when
2:27:13 they're in your head what
2:27:16 was Henry Fonda's name in Golden
2:27:24 [Laughter]
2:27:27 Pond it was Norman no Norman Norman the
2:27:32 lons the lons
2:27:36 Norman every now and again that kidney
2:27:39 kicks
2:27:40 [Laughter]
2:27:46 in
2:27:47 irregular
2:27:50 detention what are you guys doing over
2:27:52 in show and tell you doing anything oh
2:27:54 interesting CNN SK sky is
2:27:58 falling that's a cool 70s car what is
2:28:01 that a
2:28:04 pinto did I ever tell the story I I had
2:28:06 a 71 Ford Pinto it's my second
2:28:10 car I had a my my stepdad this this will
2:28:13 give you some this this is some
2:28:15 fun some fun background on Kyle's wacky
2:28:18 existence existence so my stepfather was
2:28:22 a car dealer and on my 16th birthday
2:28:24 surprised me with a 1970 Chevy Impala
2:28:27 with mag wheels Keystone mags an eight
2:28:30 trck tape
2:28:31 player and that car was
2:28:34 bitching and then I was at a stop sign
2:28:37 coming home one night and these people
2:28:40 thought I was pulling out in front of
2:28:41 them so they slammed on their brakes and
2:28:43 slid into my driver's side door and
2:28:45 smashed my car up
2:28:47 and
2:28:49 then and then I said well and he goes oh
2:28:52 it's totaled we can't fix it and so I
2:28:55 got to get you another car I said
2:28:57 awesome please just I I don't care what
2:29:00 you get me get me anything except a Ford
2:29:03 Pinto what did he get
2:29:06 me a 1971 Ford Pinto white with a black
2:29:11 vinyl
2:29:14 roof it was it was so it was so low and
2:29:18 power our house was on top of a hill I
2:29:20 would I would sort of get up to like 65
2:29:22 miles an hour coming up the hill and
2:29:24 then I would hit the hill and I would
2:29:26 [Music]
2:29:30 go and I would go like 30 miles an hour
2:29:32 of this hill I couldn't go any faster it
2:29:35 also had a hole in the floorboard right
2:29:37 by my left foot so every time I'd hit a
2:29:40 puddle water would shoot up the the back
2:29:43 of my leg the back of my calf and scare
2:29:45 the [ __ ] out of me every time
2:29:47 I
2:29:50 go [ __ ] Ford Pinos that was the worst
2:29:53 car I've ever own well was it that or
2:29:56 was it the
2:29:57 1972 AMC hornet in sky blue that I drove
2:30:01 to
2:30:02 college that's a tossup 71 Pinto or a 72
2:30:06 it might have been a 74 AMC hornet let's
2:30:09 let's go see if we can find welcome to
2:30:12 Chad add how can I help you we're going
2:30:15 to show you what a what a 1974 AMC
2:30:18 hornet looks like in sky blue that's
2:30:20 what we're going to
2:30:22 do is this what he does here yeah pretty
2:30:26 much uh
2:30:29 1974 let's see 74 AMC
2:30:34 hornet sky
2:30:37 blue image oh there it is oh my
2:30:43 God can you say chick
2:30:45 magnet
2:30:49 [Laughter]
2:30:54 pretty sure I had the four-door was
2:30:55 there a
2:30:58 four-door that's a gremlin isn't it yeah
2:31:01 that's a gremlin we want the Hornet The
2:31:04 Hornet's way
2:31:08 worse that was it let's see see if we
2:31:11 can get a all
2:31:13 right
2:31:15 um
2:31:18 1971
2:31:21 Ford Pinto
2:31:24 white black
2:31:32 vinyl dang
2:31:44 it that was pretty much it
2:31:46 [Laughter]
2:31:52 oh
2:31:54 yeah oh
2:31:56 [Laughter]
2:31:59 god wow all right well you're welcome
2:32:03 you're
2:32:04 welcome why did he do that who is he
2:32:06 what is going on here what are his
2:32:09 qualifications I want to see a turtle
2:32:11 driving a pinto now I I think I think
2:32:14 that's only
2:32:15 right okay
2:32:17 let's go to chat
2:32:24 GPT
2:32:32 okay I'd like you to
2:32:38 create a magazine ad from the
2:32:42 1970s with an anthropomorphized turtle
2:32:47 driving a 1971 Ford Pinto that is white
2:32:51 with a black vinyl top
2:32:57 period the turtle should be figting a
2:33:00 smile as if to indicate he's confident
2:33:04 in his
2:33:05 ride comma knowing that he really
2:33:12 isn't I didn't have a pinto but I had 75
2:33:15 Mustang too with 15inch mags there you
2:33:18 [Laughter]
2:33:21 go yeah the 75 like like the Mustangs
2:33:24 [ __ ] went South in like 73 they just
2:33:29 went and I think what 76 was the
2:33:31 beginning of the fox body or was that 75
2:33:34 then they just then they were just
2:33:35 shitty for
2:33:37 decades I got a Mustang I got a in
2:33:43 20 was it no 2004
2:33:47 when the when the new retro Mustang came
2:33:50 in I got a a green
2:33:52 one oh these are pretty
2:33:57 good oh wait
2:33:59 [Laughter]
2:34:08 sorry that was that was pretty much it
2:34:11 although this one has a lot more Chrome
2:34:13 than mine did and that's not really a
2:34:15 pin though
2:34:21 oh
2:34:22 yeah tell him what he's won you've won a
2:34:24 1971 Ford
2:34:27 Pinto you're going to have to pay the
2:34:28 taxes on that of
2:34:33 $33.75
2:34:36 wow there you have it never a dull
2:34:42 moment the internet is ending so many
2:34:45 Pinto stories so many Pinto stories
2:34:47 friends
2:34:49 in ashalon were conveniently in hia on
2:34:53 the wa are you aware someone is imperson
2:34:56 you impersonating you and chatting us
2:34:59 up yes so God damn
2:35:04 it ah so anyone here so this is a good
2:35:07 opportunity to ask you to follow this
2:35:10 channel but be very careful that you're
2:35:13 not following that channel so AI L
2:35:18 learning lab 2 L's is not me it's
2:35:21 someone who's stolen my content stolen
2:35:24 my image stolen the name is
2:35:27 getting my followers to follow them or
2:35:31 getting people interested in this
2:35:32 content to follow them and then hitting
2:35:34 them up to sell them crypto [ __ ] so
2:35:38 thank you for um saying that Mercy I'm
2:35:42 sorry for that please go check if you
2:35:45 follow this channel make sure you're not
2:35:47 following that
2:35:49 channel cuz that guy's an
2:35:51 [ __ ] that guy is just clueless that
2:35:56 guy's an [ __ ] good bad good bad good
2:36:01 but flawed
2:36:03 bad adequate
2:36:07 horrible nearly
2:36:11 acceptable absolutely
2:36:15 unacceptable
2:36:18 you can make money with ch by scamming
2:36:21 people and selling them your shitty
2:36:22 crypto Ponzi
2:36:25 scheme just I the thing that that
2:36:28 [ __ ] just roast my nuts about that
2:36:32 particular scam is the amount of time
2:36:35 energy and effort that this guy is
2:36:37 putting into acting like me he could
2:36:40 just start his own [ __ ]
2:36:42 Channel just [ __ ] just do it just
2:36:46 just [ __ ] do something just put in
2:36:49 the same amount of energy into not
2:36:51 ripping people off and you won't be an
2:36:54 awful
2:36:58 person I'm so
2:37:00 exhausted can you give different
2:37:02 channels names if there are more yeah
2:37:04 that would be great Mercy if if it's not
2:37:06 just that one that would be
2:37:10 swell it's a good thing right it means
2:37:13 that this Channel's popular right that's
2:37:15 good right
2:37:17 it's not good it's horrible I hate it I
2:37:19 [ __ ] hate it man man oh
2:37:24 man I love the color cool people putting
2:37:28 in some good Ford Pinos into into the
2:37:30 old into the old show and tell I don't
2:37:33 think Ford Pinos go in the showand tell
2:37:35 area
2:37:37 people I don't think you show them
2:37:40 off what do you call a four-door Yugo oh
2:37:45 I used to know that joke I forget I
2:37:48 don't know Doug burningham
2:37:52 what a paper weight a Wego get it oh
2:37:56 that's good Yugo Wego nice groovy oh
2:38:01 Kyle that car right exactly Cam katkin
2:38:04 that like yeah exactly to go from oh
2:38:06 wait so let me show you the 71 imp one
2:38:09 with Keystone
2:38:12 mags or was 1970 19 uh
2:38:17 1970
2:38:20 impella
2:38:22 green Keystone
2:38:33 mags that was kind of it oh
2:38:38 yeah that was kind of
2:38:40 it the trunk the trunk you could put
2:38:43 like a queen-sized bed in the trunk it
2:38:45 was massive
2:38:47 it's just a [ __ ] huge
2:38:50 boat it kind of sat low like that I
2:38:53 don't think it had that strip it didn't
2:38:55 have that
2:38:57 stripe that's a 68 wait yeah that's a 68
2:39:01 that's not a
2:39:02 70
2:39:06 [Music]
2:39:12 anyway all right let's see what are we
2:39:15 doing here people what are we doing here
2:39:17 are you aware someone is impersonating
2:39:19 you yes I am and they're not very good
2:39:21 at it they can't make the funny
2:39:24 voices they just can't do it they're
2:39:27 like my name's Kyle I got crypto do you
2:39:31 want to buy my crypto my name's Kyle I'm
2:39:34 an influencer look at all my videos
2:39:36 aren't they
2:39:37 good would you buy my crypto my crypto
2:39:40 is so awesome you'll make so much money
2:39:42 did you want money do you like money you
2:39:45 can make money will Jud you I'll make
2:39:48 you money this is solid gold would you
2:39:50 like one you just buy my crypto and
2:39:52 you'll get I'll send you a gold bar yeah
2:39:55 that's the ticket
2:40:04 yeah on 60 Minutes a company is making
2:40:07 3D printed homes yeah that's cool those
2:40:10 are
2:40:12 cool um I think what did they C they
2:40:15 cost
2:40:16 they were costing like in in that it was
2:40:19 in a development they were costing like
2:40:20 400 Grand I'm like I thought I thought
2:40:22 the whole point of a 3D printed home was
2:40:25 it didn't cost a
2:40:27 lot my brother had that four-door car
2:40:31 with the big pole speaker that used to
2:40:33 roll roll the
2:40:36 cable out to the beach oh
2:40:42 yeah and there is where the viewers are
2:40:45 that guy still still out there I guess
2:40:46 he
2:40:47 is I don't talk about him as much cuz I
2:40:50 don't want to give him power but it's
2:40:51 like I
2:40:52 just so [ __ ] annoying and Tik Tok
2:40:56 won't do a [ __ ] thing about
2:41:00 it not a goddamn thing they won't do a
2:41:04 thing they won't do a thing they won't
2:41:05 do it they won't do it Tik tok's like oh
2:41:09 yeah we don't say no problem with it we
2:41:10 don't see no problem with it we don't
2:41:13 but he copied all my content yeah we
2:41:15 don't see a problem with it we don't we
2:41:18 don't see a problem with it who are you
2:41:20 people we don't like to talk about that
2:41:22 we're not we're we're not back we'll be
2:41:24 in the shed we'll be in the shed back
2:41:26 there we got the we're making the tick
2:41:28 toks in the in the shed back there you
2:41:31 could you could uh there's a there's a
2:41:34 mailbox down the end of the by the
2:41:37 holler the the the Rock by the tree in
2:41:40 the holler by the take a left at the at
2:41:42 the
2:41:43 brook we'll be back in the shed
2:41:47 that's Tik Tock management I've met them
2:41:51 I've met them they're back in the
2:41:57 shed I don't know what's going on in
2:41:59 their
2:42:03 brains Robert Rossy I reported AI
2:42:06 learning lab thank you you reported it
2:42:08 to people that talk like this yeah we
2:42:10 don't see a problem with that we don't
2:42:11 see a problem with that no well don't
2:42:13 that's good thanks Robert we got we got
2:42:15 it we got it that that that uh uh the
2:42:19 the AI learning lab there they don't
2:42:21 they don't know how to spell the they uh
2:42:23 they got it with just the one L and and
2:42:25 we like the one with the two L's we like
2:42:27 we like with the two L's and uh that
2:42:30 that other that gentleman who's face it
2:42:32 actually is he he seems very bothered by
2:42:34 it all so uh we we uh we just don't
2:42:38 really show his content is what is how
2:42:40 we're handling it we we just don't
2:42:42 really show the content all right good
2:42:45 but thank you for your concern we do we
2:42:47 do appreciate that we'll be back in
2:42:48 their
2:42:49 shed tell them it's trademark
2:42:51 infringement I did three
2:42:54 times I told them they were copying me
2:42:57 15 times they were copying a c celebrity
2:43:01 seven times and that they were doing
2:43:03 trademark infringement three times and
2:43:06 then they were like yeah we don't see a
2:43:08 problem with it it's good we're going to
2:43:09 go ahead and keep it you there's the two
2:43:11 of you with the identical content where
2:43:13 the one of you is saying that you're
2:43:14 ripping off the other one and we just we
2:43:17 don't we don't really see any similarity
2:43:19 other than the exact [ __ ]
2:43:25 videos seems to
2:43:27 me there's a thing called a uh Pates on
2:43:31 here pain works for Google there's a
2:43:33 thing called a diff where you can check
2:43:36 the diff ER between when these videos
2:43:40 were posted and when these identical
2:43:42 videos were posted and these ones will
2:43:45 be after after significantly these ones
2:43:48 and you could um in computer parament
2:43:52 throw this one the [ __ ]
2:43:58 out they get an LA llama for effort
2:44:02 exactly oh that would be a cool t-shirt
2:44:05 I don't know which one we don't see a
2:44:07 problem with that Tik Tock we don't see
2:44:09 a problem with that I inherited my
2:44:11 granny's 1975 Buick Electra 325 wow wow
2:44:16 that's a that's in the uh AMC hornet
2:44:20 category Grandma's Grandma's 1975
2:44:25 Electra big bench
2:44:29 seats
2:44:32 oh I want a car with an eight track yeah
2:44:34 I think you can probably still get them
2:44:36 I'm calling Pi don't call Pi don't call
2:44:39 Pi because she's going to tell Bing and
2:44:40 then Bing's going to be pissed at me
2:44:42 Siri's already just off the [ __ ] rail
2:44:45 every time I talk to Bing Siri jumps in
2:44:47 I can't I don't want to get pie involved
2:44:49 with this please Silver Fox
2:44:53 it's I can't I
2:45:00 can't it's comedy it's comedy how we're
2:45:03 doing comedy
2:45:04 here crack ha Jen up oh crank ha Jen up
2:45:08 yeah exactly 1974 350 with three-speed
2:45:12 Camaro oh you show off Firefly
2:45:16 that's that's just showing off a 350 I
2:45:19 don't like you
2:45:21 now I mean I did have a cool car for
2:45:24 like six months till I got smacked in
2:45:25 the door hinge and then I went and then
2:45:28 I went Pinto I went this was my my modes
2:45:32 of transportation out of high school
2:45:34 well in high school it was Pinto then
2:45:37 AMC hornet to college then I sold that
2:45:39 and bought a mountain
2:45:43 bike then I moved to New York City and
2:45:45 you don't need a [ __ ] car so I didn't
2:45:48 have a car until I bought when I when
2:45:52 Omnicom invested in agency.com
2:45:56 in 96 or 7 I think it was
2:45:59 97 I bought a brand new Ford Explorer it
2:46:02 was the first time I'd ever owned a new
2:46:04 car that was pretty
2:46:06 cool is that like a C diff it's it's
2:46:11 similar that's over their head I'm
2:46:13 afraid yeah exactly
2:46:17 exactly
2:46:22 ah I rebuilt the TR4 with my dad that's
2:46:25 super cool
2:46:28 tonius I had a stepdad that owned a car
2:46:31 dealership and we did not work on cars
2:46:33 together I'll just leave it right
2:46:35 [Laughter]
2:46:40 there I don't think he appreciated my my
2:46:43 desire to be an actor as much as I
2:46:46 did I don't think he quite appreciated
2:46:49 the sensitive man in the
2:46:56 house
2:46:58 ah now I'm watching watching Jeffrey
2:47:00 Hinton talk about AI on 60 Minutes yeah
2:47:03 they they're going to only take the
2:47:04 things where he's like the robots are
2:47:05 going to kill us that's why I quit
2:47:10 Google all right listen people listen
2:47:12 listen I'm getting a little sleepy here
2:47:15 I did too many demos early and it hurt
2:47:17 made my head hurt my head hurts I
2:47:19 haven't had enough coffee yet you could
2:47:21 sleep in station wagons I know those
2:47:23 were good
2:47:27 cars all right everybody listen let me
2:47:30 go through here and see who's here
2:47:32 tonight because it's been cool get some
2:47:33 rest Kyle thank you you got this thank
2:47:36 you I appreciate that that was uh Sherry
2:47:40 Banks thanks for hanging out Robert
2:47:41 Rossy thank you thank you I'm the
2:47:43 president of the the B9 Builders Club
2:47:46 old lost and Space robot oh I used AI to
2:47:49 create python super cool Robert that's
2:47:51 very cool Silver Fox good to see you
2:47:54 Clint as always Irregulars present
2:47:57 Turtles Emilio's wife good day Corey
2:48:01 good to see you steo Dr J dragonfly
2:48:05 Alchemy as always thank you very much
2:48:08 for all your support all you cool people
2:48:09 for hanging out Nancy Silver Fox you're
2:48:13 all so awesome you're so awesome and
2:48:15 weird we're not weird we're
2:48:18 Irregulars I know go tell it to Tik Tock
2:48:22 Go Tell it to Tik Tock Corey sander
2:48:24 Pottery
2:48:26 peace all right everybody listen have a
2:48:29 fantastic evening I will see you
2:48:31 tomorrow I don't know what's going on it
2:48:34 doesn't matter whatever's going on I'll
2:48:35 be here I'll be around I'll see you
2:48:37 tomorrow hang out
2:48:39 bye