Apr 19, 2024 - From Poetry to Music in Real-Time!

Description

Chapters

Transcript

0:00 ampy
0:03 [Music]
0:09 [Music]
0:49 is this place I can rest my poor
0:52 [Music]
0:55 head G my thoughts in sweet silence
1:01 [Music]
1:04 is place where the feelings are
1:08 [Music]
1:10 near an over exposure to
1:14 violence this this place I can slowly
1:18 face the only one I truly can
1:23 know these are tears from a long time
1:27 ago got these tears from a long time ago
1:32 I need to cry 30 years or
1:35 so these are tears from a long
1:39 [Music]
1:50 time good evening happy Friday night
1:53 date
1:54 night so Brandon there welcome back I
1:58 know you were traveling I saw a picture
2:00 you uh sporting AI Salon gear in public
2:04 that is highly highly
2:06 [Music]
2:14 encouraged Champ's on tonight
2:22 [Music]
2:40 good evening good evening good evening
2:43 good evening want to see something cool
2:45 of course I want to see something cool
2:47 are we talking meta AI or llama 3 this
2:49 evening we will do that for
2:52 sure we're even going to talk
2:55 about the speed with which this stuff
2:57 happens grock the hardware company
3:00 [Music]
3:02 has
3:03 already put llama 3 up on their um 800
3:09 token a minute demo server which is
3:12 amazing hey Tobias what's happening
3:15 [Music]
3:56 w
3:58 [Music]
4:28 yeah that's a minor he goes nuts a minor
4:31 is his his
4:34 key right
4:37 [Music]
4:46 champy I see he got his jar that's how
4:50 champy makes his money
4:53 [Music]
5:09 um yeah that's uh Amelio's wife did the
5:11 tip jar thank you Emilio's wife champy
5:13 appreciates
5:15 it quarters for chicken sticks
5:26 [Music]
5:31 all right people happy Friday night
5:33 let's get the uh looks like got some
5:35 likes going let's do the double tap see
5:37 if we can get some more people in here
5:39 um share the live if you haven't done
5:41 that and then we can start playing
5:44 talking about stuff luny stick what's
5:47 happening happy day happy
5:55 [Music]
5:59 day nice nice tapy Taps we're up to 1.7
6:04 th000 likes I like it solid we haven't
6:06 even started
6:08 yet just not just rolling just
6:13 rolling all
6:14 right um I had a I had a long day today
6:20 lot going
6:22 on had a good office
6:24 hours where lot of lot
6:29 of a lot
6:33 of people trying to figure out just
6:36 where does this AI [ __ ] fit in their
6:39 lives cuz it's not quite ready for prime
6:42 time there's a lot changing there's now
6:45 a lot of options for all of the
6:47 different tools right like if you want a
6:50 large language model there's a bunch of
6:52 free ones that are pretty much as good
6:54 as chat
6:55 gp4 if you want image generation models
6:58 there's a bunch of good ones there
7:00 um there's a bunch of new open- source
7:04 stuff Source Camp office hours was so
7:07 good today yeah we C we came up with
7:09 some really fun
7:10 ideas really fun ideas
7:16 about we're we're at an interesting
7:19 stage I well and this could just me be
7:21 me projecting on the world I'm at an
7:23 interesting stage I'd love you hear your
7:25 thoughts on how AI might fit into the
7:27 eye industry and network industry
7:31 um well
7:36 so so in a lot of ways you know Network
7:39 and it becomes more critical than ever
7:42 right because and I I also saw a really
7:46 interesting
7:47 thing that was
7:53 um that that data centers
8:00 data centers are not located where the
8:03 energy is they're located close to where
8:06 the people
8:07 are for Speed of distance they can't be
8:11 far away so so or something like that
8:16 but with training large language models
8:19 it's not synchronous you don't need to
8:21 do it in real time but you need massive
8:24 amounts of energy so you could say find
8:28 an area like a big desert that's full of
8:32 sun and just you know Park a bunch of
8:35 you know massive solar arrays and have
8:38 massive data centers that are training
8:39 sets because they don't need to be you
8:42 know proximity to cities I thought that
8:45 was really interesting today so so I
8:47 think I think networking and it and
8:51 security um are going to be really
8:55 critical so I don't think that goes away
8:57 that said I
8:58 think the management of those networks
9:02 is going to shift dramatically I think
9:04 the software development world for how
9:07 you manage those networks is going to
9:08 shift
9:11 dramatically you know where the bulk of
9:13 the work that gets done in those
9:16 networks is going to change dramatically
9:18 so yeah so I don't know so anyway so so
9:22 I think I think networks you know go up
9:24 in importance compute is kind of the
9:26 gold of the future so um you're going to
9:28 have to have your [ __ ] together with
9:30 with your with your data centers
9:35 um where where I've been going
9:42 is most of the tools that that I use
9:46 most of the time which is you know it's
9:48 mostly like image and word tools you
9:51 know I do some animation I do some music
9:54 but that those are not like what I do
9:55 the core of I work
9:58 on there are
10:01 at least three probably more like five
10:04 viable Alternatives of
10:08 tools and so it Almost Doesn't Matter
10:11 What tool you use anymore like it felt
10:13 like it was a bigger deal or you using
10:15 GPT 4 cuz that's the big good one now
10:17 doesn't feel like that's as critical
10:19 anymore and there's a bunch of free
10:20 options now what feels increasingly
10:23 important is like what are you doing
10:26 with them what are you producing with
10:28 them Jim Jim Ross today on on office
10:31 hours was like well I'm mostly producing
10:34 content you know for marketing and I'm
10:38 in a similar place where I'm mostly
10:40 using it for [ __ ] that I did before
10:43 right like I need to make I need to
10:45 respond to an email I need to right
10:47 that's all it's all [ __ ] we did before
10:49 and now we're using AI to do that more
10:51 efficiently which is fine it's
10:55 good but I just have this aching feeling
10:57 there's something more like we're build
10:59 we build these gpts and we put them out
11:02 in the world and no one uses them well
11:04 are we using them how are we using
11:07 them I made a virtual twin of myself in
11:11 this project with a bunch of other
11:14 people we've all got these digital twins
11:16 some of us are using them some
11:20 aren't I helped build the [ __ ] things
11:23 and I'm not really using them why not
11:25 what could we be using them for how are
11:26 they different how is interacting with a
11:28 virtual version of me different than
11:30 interacting with me like I should
11:32 understand how to do that I I should
11:33 understand what can this thing do that I
11:35 wouldn't have thought of before so the
11:37 the office hours talk today was just a
11:39 lot of talk about like how do we find
11:43 work what is the work how can we use
11:46 these tools in ways that are new and
11:48 novel and interesting not just
11:51 efficiency tools from the things we did
11:53 before Silver Fox I think gpts were a
11:56 good stepping stone about prompt writing
11:58 and results thereof yep
12:00 oops sorry I'm late it's okay Daniel we
12:04 forgive
12:07 you it's very cute though listen to
12:10 Ethan mik's new book co-intelligence on
12:12 the plane excellent read okay so here's
12:15 a question Brandon listen I love you
12:19 fantastic this this is a a pep peeve I
12:22 got with audio book
12:25 people you said it was an excellent read
12:27 you read it no was an excellent listen
12:32 wasn't an excellent
12:33 read an excellent read involves your
12:35 eyeballs doesn't it isn't this a thing
12:39 am I the only one that every time
12:42 someone who says I read a book and then
12:44 you're like oh how was it like great I
12:47 listen to it at the gym while I'm doing
12:50 Burpee
12:52 squats that I'm pretty sure that's not
12:56 reading do not talk down to the audio
12:59 people
13:00 there's no audiobook shaming in in here
13:02 is that what you're
13:12 saying Vicky okay listen Vicky Vicki is
13:17 my surrogate brain when my brain forgets
13:19 words Vicky is there for me Vicky has
13:21 come down on the side of huh not
13:25 reading it you did not read that book
13:29 was a good listen you could say it's a
13:31 good
13:32 book but a good read I don't think so
13:35 Brandon I don't think so listen just
13:40 saying are there
13:43 demerits for this kind of infraction on
13:45 this channel I don't think so we we
13:48 still process the same info I listened
13:50 you look I don't read I listen I'm not
13:53 saying reading's better I have ADD I
13:55 can't [ __ ] read I can't get through a
13:57 page without wondering when hell this
14:00 book's going to be over so I I don't
14:02 read but I also do not
14:05 listen do do do visually impaired not
14:08 read Braille oh that's a good call lunar
14:11 stick because it's a different sensory
14:19 thing I guess you are ingesting writing
14:22 I I I don't know I don't know I can't
14:24 listen or read I'm right there with you
14:26 Silver Fox I I am functionally
14:29 illiterate at this point I mean I can
14:31 read fine just fine for short amounts of
14:34 time but whole books just just doesn't
14:38 do it does AI read actually that's
14:42 really
14:44 interesting
14:49 huh I think in a lot of ways AI does
14:51 read right so if you read a
14:54 book your brain is encoding you know you
14:57 take actually this is kind of
14:58 interesting brand
15:01 oh wait but Vicky Vicky countered I
15:03 don't read podcasts okay all right so
15:06 see we've there are valid points on
15:09 there are very good people on both sides
15:11 of this
15:13 argument but but to the do AI do do
15:16 machine does machine learning read if
15:20 you think about how our brain works
15:22 whether you get it through brail your
15:24 ear holes or your eye
15:26 holes there are words in an order
15:30 that go into your brain and your brain
15:32 encodes them embeds them
15:36 right and then those words are sort of
15:39 clustered in little semantic groupings
15:43 in your brain and when someone
15:45 says something about how you use AI your
15:49 brain goes oh I read that thing from
15:51 Ethan mullik about you should treat the
15:53 AI like people and then
15:55 you pull that out of your large language
15:58 model down to your mouth hole and you
16:00 squirt out
16:04 that which is kind of what machine
16:06 learning's doing so yeah I think machine
16:09 learning I think it reads I think it
16:12 reads it might not store it quite the
16:14 way we store it but it's pretty close I
16:16 mean neural networks are designed on how
16:19 our brains are wired right huh
16:23 fascinating welcome back
16:26 Brandon I tokenized his content exactly
16:29 a processes and processes and analyzes
16:32 text it does not
16:34 read but it stores
16:40 it it stores it in an in an interesting
16:44 way
16:50 where H I don't know I don't know the
16:52 meth well enough I don't know what the
16:53 [ __ ] I'm talking
16:55 about it's not even text this number is
16:57 converted
16:59 no I understand that I understand how it
17:01 works like it's it's getting embedded
17:03 but like our brain it's not a word
17:04 that's sitting in our brain it's an
17:06 electrical you know it's an electrical
17:09 um synapse or an
17:11 Impulse that you know when fired in the
17:15 sequence decode into a word right that's
17:19 essentially what a large language model
17:20 is doing right a probability goes into
17:23 the latent space it it grabs a point in
17:27 mathematical space brings it out
17:29 converts it to a word and squirts the
17:31 word out that's essentially how our
17:32 brains work you're just talking the
17:35 information in through different holes
17:37 oh taking the information in through
17:38 different holes yeah I know but my issue
17:41 was can you call it reading if you're
17:45 listening to it I think I think it's a
17:47 good listen rather than a good read
17:50 that's all I was
17:51 saying I I didn't realize the most
17:54 controversial thing I would say on this
17:56 channel was an audio book isn't
18:02 reading there are very strong feelings
18:06 here I think this is for it apparently
18:10 this is the most interesting issue of
18:11 the 2024 election oh do you hang up on
18:15 people Kyle um May maybe I don't um I
18:22 generally don't pick up if I don't know
18:24 their
18:26 number all right MW to receive
18:30 oh maram Webster to receive or take in a
18:33 sense
18:35 of letters symbols especially by sight
18:38 or touch interesting all right Mariam
18:42 Webster is saying the ear
18:44 holes ear holes are not part of the
18:47 reading experience but fingers and eyes
18:49 are I wonder if actually you know it's
18:52 interesting I wonder if that's going to
18:53 change because audio books have gotten
18:54 so popular that might
18:57 change oh man why does stable diffusion
19:00 need a GP
19:02 GPU
19:04 um
19:09 well the way that the way that large
19:11 language models and diffusion models
19:13 work there are very specific kind of
19:16 math where the GPU is more efficient
19:20 than a CPU so so basically if P's in
19:24 here he can spank me later but but think
19:28 of it like this a CPU is a
19:32 CPU that's pretty funny almost as
19:35 controversial as the thumbs up emoji
19:37 that was a good one Papa JJ a CPU is the
19:40 most general of processing units Central
19:44 Processing
19:46 Unit it's designed to do lots of things
19:49 and and so in doing lots of things it
19:52 got it's got lots of different sort of
19:55 processing areas and then buses in
19:57 between those processing areas it's very
19:59 complicated because it's got to do a lot
20:02 of different things right it's got to do
20:04 networking and it's got to do processing
20:06 and it's got to do IO management I
20:08 whatever the [ __ ] it's got to do it's
20:10 got to do [ __ ] A GPU is faster because
20:14 it says hey we're going to limit the
20:16 amount of calculations we do to a
20:19 specific set of calculations that just
20:21 happen to be good for things
20:23 like doing Graphics processing right so
20:26 you're just you're just ramming a bunch
20:27 of data sort of through this
20:30 system um and then they found out oh
20:33 that also happens to be really good for
20:35 doing things like mining Bitcoin where
20:37 you're where you're having to sort of
20:38 run the the same kind of mathematical
20:40 operations over and over and over again
20:42 because it's more specialized it it's
20:44 good at that and then they found out oh
20:46 that happens to be really good for um AI
20:49 training as well right because it's more
20:52 simple and now there's a whole new class
20:54 of chips called tpus and I don't know
20:57 what the ones called that grock does grq
21:00 but they're even more specialized in
21:02 simplified and they go much much faster
21:04 let me let me show you
21:06 something
21:12 so so here's chat GPT 4 and we're going
21:15 to say
21:17 why uh uh let's see let's see make
21:22 a an
21:25 epic poem about
21:30 which
21:34 came first the chicken or the
21:39 egg and so if we pop that in you know
21:45 it's going well now it's going really
21:47 slow but even when it's going decently
21:50 fast it's like going slow enough that we
21:52 can read it this is actually
21:54 reminiscent of like a you know a 96 bod
21:59 modem back in the day when we went from
22:01 2400 bods which was like half this speed
22:04 to 9600 bods or then
22:07 144 you know when you saw text on the
22:10 screen it was kind of this
22:12 slow and so if I'm going to if I grab
22:14 this same
22:16 prompt and now I'm going to hop over to
22:18 this site called grock gq.com
22:31 and where do I choose the model I think
22:35 it's defaulting to
22:38 it so I'm going to put in the
22:41 same make an epic poem about which came
22:43 first the chicken or the egg now watch
22:45 this oh I gotta log in dang
22:49 it
22:51 okay did I sign in oh it's too late it's
22:54 already
22:55 done um here let me start over can I do
22:59 a new one so so we're using the Llama 3
23:02 Model so this is um this
23:05 is meta's new llama 3
23:10 model and in fact we'll do their we'll
23:12 do their uh 70 billion parameter
23:18 model and I want to do a new chat how do
23:21 I do a new
23:23 [Music]
23:25 chat how do I do it
23:29 all right uh scroll to the Top Copy
23:34 content all right let's pop in this
23:36 prompt
23:37 again all right watch it watch for
23:44 it
23:45 done the
23:47 egg the chicken so this is doing for the
23:51 70 billion parameter model 287 tokens a
23:55 second so that's about 200 words per
23:57 second and if I if I do with their
23:59 smaller model the the 8 billion 8
24:03 billion parameter
24:06 model that's doing 800 tokens a second
24:10 right where I don't know how fast chat
24:13 GPT was going but it was probably in the
24:14 neighborhood of 50 to maybe 100 so this
24:17 is eight times as fast how they're doing
24:20 that is the chip that grock makes is
24:23 even more specialized than GPU so you
24:26 got CPU super complicated GPU more
24:29 simplified and specialized and something
24:31 like grock it's only designed to do
24:34 large language model inferences so when
24:37 you do a prompt to chat GPT that's a
24:39 large language model inference and it's
24:41 designed just to do this and so it's
24:43 super fast because of that so that's why
24:47 efficiency or throughput or whatever the
24:50 kind of math it does I don't know
24:53 specifically what to do but in general
24:56 in
24:57 general that's how it works
25:03 it says clear chat in the upper leth
25:05 hand corner oh yeah oh yeah they buried
25:07 it underneath the
25:10 prompt that's the problem with interface
25:13 well especially when you go from sites
25:15 that are like relatively the same but
25:17 they're slightly different stuff's just
25:18 in slightly different positions your
25:20 brain kind of normalizes parts of the
25:22 interface so you literally can't see the
25:25 thing right in front of
25:27 you is it still worth paying for chat
25:35 GPT
25:39 um I don't think it's worth paying
25:44 for multiple
25:49 models I think it's worth
25:52 paying for either chat GPT or something
25:56 like Po where you can get access to a
25:58 bunch of different
26:00 models
26:02 um or something like Claude because
26:05 Claude 3 is very much a um gp4
26:10 competitor the the deal about chat GPT
26:12 though is no one's really no one's
26:14 really caught up with
26:16 gp4 and it's also got it's connected to
26:18 the internet claud's not um it's got
26:23 code interpreter Claude doesn't um meta
26:26 doesn't like there there's a bunch of
26:28 other things that don't have the
26:30 features that chat GPT has so I still
26:33 think chat GPT is the one to pay the 20
26:36 bucks a month
26:38 for I do chat GPT and co-pilot and
26:41 co-pilot's free do you do you pay for
26:43 co-pilot mud because co-pilot Pro you
26:46 can make gpts although I got a I got a
26:49 note from from Jim Ross today saying
26:51 that he tried to convert his GPT using
26:55 um was it Jim Ross no it was someone
26:58 else
27:01 I forget who it was but
27:03 anyway okay so you pay for pay for that
27:05 as well yeah he tried to he tried to
27:07 replicate his GPT in Microsoft co-pilot
27:12 and he said it's just not ready for
27:14 prime
27:15 time so yeah I don't know I don't
27:19 know I think chat gp4 is still still
27:22 worth paying for cuz quite
27:24 frankly it's been a year it's been
27:27 really two years since they've had it
27:29 kind of done and it was two years before
27:32 that they were working on it so gp4 is
27:34 four years
27:35 old and everyone is just now almost
27:40 catching up to it and they're red
27:42 teaming um chat GPT 5 right now so so I
27:47 think it's worth paying for chat GPT
27:49 just to get what's that
27:52 thing oh forever hook did that thank you
27:55 very much forever hook I thought I was
27:57 getting kicked off tikk St I thought it
27:59 was like 3 2 1 you're out of here you're
28:02 banned for
28:08 life paying $20 why why not pay this
28:12 small amount is a no-brainer no I get it
28:15 but the the the challenge is really how
28:18 many of those $20 a month is worth
28:20 paying for it used to be like I paid for
28:24 Claude and for chat GPT because Claude
28:27 had a big context window
28:29 well now I can go to Gemini Pro 1.5
28:31 which has a million token context window
28:34 and I can use it for free so I don't
28:37 have to pay for that anymore like I just
28:38 I just feel like there's enough free
28:40 tools out there that are at a level that
28:43 we can start trimming back on our
28:45 subscriptions I think I need to do a
28:47 subscription
28:50 audit I like meta AI answer formats
28:53 that's interesting I also noticed meta
28:55 seems to be meta's new um
28:59 chat GPT competitor seems to be a little
29:01 more sparse in its answers which is kind
29:03 of nice because you know chat GPT is a
29:07 little a little
29:10 wordy I pay $600 a month for10 models I
29:16 assume you're not kidding next
29:26 Vision have you tried Bard yeah I've
29:28 tried Bard a lot and Bard's now called
29:33 Gemini um
29:35 gemini.com both of them there's the
29:38 non-paid version and the paid version
29:41 one's Pro one's
29:43 Ultra they both
29:46 suck but there's a third one there's a
29:49 third super secret
29:51 Gemini that's at AIS studio. goole.com
29:56 not gemini.com
29:59 AI studio.
30:01 google.com and that's Gemini Pro 1.5 and
30:05 it's really good and it's got a million
30:08 token context
30:10 window and that's crazy that's crazy
30:15 it's also multimodal you can upload 11
30:18 hours of audio and it just listen has
30:21 listened to it or as as Brandon would
30:24 say it has read the entire 11-hour audio
30:28 [Laughter]
30:34 file sorry Brandon I don't think you're
30:37 living this one
30:39 down oh
30:43 man the free tools are not likely to
30:46 remain free I would absolutely agree
30:48 with that
30:50 although I would think
30:53 that meta will probably stay free don't
30:56 you think pate
31:00 because I would because I would assume
31:02 that if because in order to save your
31:04 prompt history in meta you have to log
31:07 in with your with your account and then
31:10 they're also using that llama model
31:13 across
31:15 everything IDR is coming out with 2.0 in
31:18 a few days oh interesting hope they keep
31:20 the free version they're coming out with
31:23 2.0 already they just they just launched
31:25 1.0 they went from like 0.1 to 1.0 like
31:29 a month
31:32 ago are they really coming out with it
31:34 that soon I didn't see that
31:36 today uh uh uh let's go look at the
31:42 X ID o g
31:51 2.0 idag
31:54 2.0 has launched 2.0
31:58 has
32:17 launched I I did not B I did not
32:21 B is it true it's launched id.
32:47 um busy do not disturb I can't I can't
32:51 look at these public sites I'm G to get
32:53 kicked off Tik Tok dang
32:56 it stop Mak taking Nookie pictures you
32:59 pervs oh come on oh I can't okay I'll
33:02 have to make my own thing she get got
33:07 everywhere okay wait hang on 70s muscle
33:11 cars you're like oh god with the [ __ ]
33:14 cars again 70s muscle car in
33:19 abandoned
33:24 Factory let's see do we have anywhere
33:26 that says what vers version this
33:30 is oh I just did I just lose my thing
33:33 God damn
33:38 [Music]
33:41 it 25 of 25 prompts left Discord
33:46 Community maybe let's hop over to the
33:48 Discord announcement
33:58 we're depreciating I okay that was the
34:01 17th
34:11 yesterday I don't see 2.0 in here
34:15 anywhere all right I don't know I don't
34:22 know yeah I don't know what that
34:24 is I already bore my students making
34:27 them listening listen to my my lectures
34:28 maybe I should bore super secret Gemini
34:31 yes I got an email
34:34 really Genie in a Bottle huh let's
34:38 see let's go to the Gmail and see if
34:45 ideogram maybe I got something too I've
34:48 I am woefully behind on my email
34:51 so uh what's it called uh ideogram
35:01 introducing 1.0 on March 2nd introducing
35:07 1.0 I don't have it in
35:12 here I would
35:15 say being conservative I would go check
35:18 that email Genie and make sure it's
35:20 actually from
35:22 id. if it's from like id.com or
35:26 .xyz or or anything other than id. a it
35:30 might it that might be a spoof cuz it's
35:33 it's not in their Discord it's not
35:35 announced in their
35:40 Discord i max out Claude every day I
35:44 have I have three accounts wow next
35:47 Vision that's
35:49 crazy it will make sense if tied into a
35:52 larger ecosystem with tons of money and
35:53 a need for data
35:58 hey Pate were you here when I was
36:00 describing the difference between a CPU
36:02 a GPU a TPU and an lpu I hope I did you
36:05 proud I think I did
36:08 good converted some yo
36:12 GPT wait converted someone oh to GPT
36:16 that thought AI was scary good job jimc
36:19 I I'll tell you it's it's one of the
36:21 things that I am most
36:24 um that sort of gives me the most
36:26 satisfaction with all this stuff is when
36:29 someone is oh no I missed it caught what
36:33 was probably the very end of it oh yeah
36:35 yeah I I I didn't I didn't pretend to do
36:39 any math you'll be happy to know but I
36:41 just described that you know sort of
36:43 increasing specialty and decreasing
36:45 complexity allows them to be faster
36:50 um but yeah but but getting someone
36:52 who's you know kind of actively hostile
36:55 to AI but hasn't tried it to to jump
36:58 over and you know try new things is
37:01 really cool um Let me let me share a
37:03 couple of things with you since we're
37:05 since we're grooving in here go to even
37:08 if you've even if you're a member of the
37:09 AI Salon go to the salon.
37:12 A and it's going to take you to a link
37:14 tree which I've I've updated the link
37:16 tree there's there's two sort of time
37:18 sensitive things here both of these top
37:20 two links the first one's really
37:22 important for the state of Colorado But
37:24 ultimately for the country um so
37:29 um it's a if you click on this it's it's
37:32 a it's a it's a letter that I'm putting
37:34 out uh on behalf of the AI Salon Along
37:39 with Dan Murray on behalf of the Rocky
37:41 Mountain AI interest group which is a
37:45 boulder-based
37:47 company um and it's it's a letter
37:51 basically going to the Colorado Senate
37:54 to say hey this bill that you just
37:58 dropped 10 days ago without oh you
38:01 filled out the letter great thank you
38:02 Cindy
38:04 um it's it's Bill
38:07 24205 which regulates artificial
38:09 intelligence but it just does
38:11 ridiculously stupid stuff like like
38:13 people have to reverse engineer the
38:16 systems they use to determine if there's
38:18 any bias in them and they have to get um
38:21 you know ethics lawyers and and things
38:24 like that to it's just it's ridiculous
38:27 it Al says like anything you produce
38:30 with AI has to be labeled I'm like well
38:33 okay to what degree like if I use
38:36 sentence completion in a Word document
38:38 that's AI does that have to be labeled
38:40 like it's just it's just not a wellth
38:42 thought out bill so this letter you can
38:44 go read it that there's a link to the
38:45 bill itself at the top um and then here
38:48 are some specific you know pieces of the
38:50 legislation that were you know pointing
38:53 out to them where they have their heads
38:55 firmly planted so especially if you live
38:58 in Colorado but even if you don't um go
39:02 to the
39:04 salon. go to this link tree click on the
39:07 first link that will take you to this
39:09 form which has the letter and if you're
39:12 down with it um just you know write you
39:15 know fill in sign sign up uh to support
39:19 this we're we've got a press conference
39:22 on Tuesday um that that I'm going to be
39:26 at this [ __ ] press conference cuz
39:28 again I know how to make Good Life
39:30 Choices I am I am about the least
39:32 political
39:34 person you know other than I just want
39:37 it all to stop and go away um so so I
39:40 lean heavily on on Daisy Thomas lunatick
39:44 and other people that are way smarter
39:46 about this [ __ ] than I do but you know
39:49 anyway so that's going to be happening
39:50 so if I could get your support on that
39:52 that would be great and then the other
39:54 link on the AI Salon is the second link
39:57 is tomorrow starting at 9:00 a.m.
39:59 Pacific I'm doing a Play Shop not a
40:02 workshop play shop with John Leland and
40:08 uh it's going to be a lot of fun so
40:09 that's it's a 4our thing get your hands
40:12 uh dirty with AI and and dig in and it's
40:15 going to be really good so if you go
40:18 there um you can sign up for it
40:20 tomorrow's the day and I'll give you a
40:22 little secret if
40:23 [Music]
40:25 you why is that not working oh there it
40:28 is if you go to the AI Salon
40:32 announcements area and scroll down just
40:37 past me looking at the solar
40:41 eclipse there's a discount code for 70
40:45 bucks off so if you type in Salon fun as
40:49 your discount code you'll get 70 bucks
40:51 off so you technically don't need to go
40:52 to the salon I just gave it to you
40:54 because it's tomorrow so anyway
40:57 pardon me 420 came early ideogram is
41:01 depreciating 0.1 on 422 I got my twos
41:05 mixed up ah okay Genie a bottle that
41:07 makes sense yeah they're depreciating
41:09 the old 0.1 that makes sense okay cool
41:13 great fantastic 420 came
41:18 early that's what I was saying to John
41:20 we were talking today about the the play
41:22 shop we're doing tomorrow and I was like
41:24 oh it's it's a it's on 420 we should
41:27 just show up [ __ ] stoned it's like yo
41:31 you want to do some cool [ __ ] in AI yeah
41:34 man Chin
41:36 [Laughter]
41:41 chong oh
41:43 man did AI write it you got my support
41:46 actually that letter was written by
41:51 humans humans did that
41:54 letter we don't use the word delve once
41:58 once we we we do not mention the
42:01 tapestry of innovation across
42:09 Colorado um no it's just normal
42:13 words it's a little fancier than my
42:16 taste it was funny the initial draft of
42:19 it was was basically like do you want to
42:21 have a conversation with these people or
42:23 you just want to burn this bridge so
42:26 anyway
42:37 I could just Echo to and
42:40 from Salon love no Salon fun s s a l o n
42:45 f u n fun or Salon play both of those
42:49 work Salon fun or Salon play or the
42:53 discount for the um for the play shop
42:59 so if you got nothing to do on a
43:01 Saturday come hang
43:06 out depreciating not depreciating wait
43:09 oh deprecating not depreciating yeah
43:11 right I knew that Matthew Craig
43:14 official it's you know do you know do
43:17 you know how small the letters on Tik
43:19 Tok are and then do you know how small
43:21 they are for someone my age
43:24 [Laughter]
43:27 pinch their
43:29 heads yes
43:32 deprecating ideogram 0.1 is deprecating
43:37 so that they can focus their resources
43:38 on
43:43 [Music]
43:51 1.0
43:53 [Laughter]
43:55 Del I use delve whilst before AI made it
43:59 cool oh
44:04 [Music]
44:07 man all
44:21 right hang on I'm looking at something
44:24 here people hang with me
44:28 what's
44:35 [Music]
44:40 that which had chat bot would you
44:43 recommend the most um chat gp4 still
44:48 chat gp4 by a lot
44:55 [Music]
45:09 everyone missed the significance of what
45:10 you
45:13 said I don't think I said anything
45:17 significant what was I talking about
45:27 um let's go look
45:33 at let's go look at meta is this it is
45:36 this
45:37 the so I don't know if you've seen it
45:40 but the new meta is out meta aai if you
45:44 haven't gone to it
45:46 so yesterday meta launched llama 3 it
45:50 might have been the day
45:53 before
45:55 and they've and they op sourced it like
45:58 they do but they also launched this site
46:01 which is quite close to chat GPT or
46:05 Claude or Pi or Gemini right so they're
46:10 all in the same neighborhood they're all
46:11 kind of copying each other's interface
46:13 you can you can either do um a
46:16 conversation or you could can put it in
46:19 image generation mode and it's it's
46:22 doing like a sdxl turbo where you could
46:26 put in let's see see um
46:31 rocket um taking
46:40 off from a
46:45 cupcake yeah there you go where the
46:51 frosting is the
46:53 smoke or steam or whatever the hell it
46:56 is so got that sort of
47:02 [Music]
47:05 right you can now animate oh yeah and
47:08 you can animate images right so wait so
47:11 if I submit this it's going to now
47:14 generate
47:22 these and then if you click on one of
47:25 them let's see which one we like best
47:27 that one's pretty
47:31 cool yeah I think that one I think the
47:34 original one was cool but I can just
47:35 click animate right here and it'll make
47:38 it into a
47:40 movie and this is
47:43 free that's
47:45 interesting oh look it animated all of
47:48 them that's kind of cool in fact I I
47:50 like that animation a lot
47:53 better that's pretty
47:56 slick huh
47:58 H look at that
48:08 [Music]
48:16 [Music]
48:28 all right what other questions you got
48:31 what questions you got Korea video is
48:33 about to go into beta testing yeah Korea
48:36 Korea looks pretty interesting Korea's
48:38 doing some very good stuff
48:41 lately um see if they're interface I the
48:45 problem with Korea is because it's real
48:47 time it it can do some some nasty
48:52 stuff that is going to get me kicked off
48:54 this place
48:57 here I'll just I'll just keep the uh I'm
49:00 going to keep the camera off to the side
49:02 here so let's
49:11 [Music]
49:16 see that's kind of cool let's see let's
49:20 turn this into a sun
49:46 H yeah I don't know I don't know what's
49:49 going on there all right virtual
49:52 assistant Bots I've been away for a
49:53 while glad to be back what is this AI
49:55 video that we were looking at so what we
49:57 just looking at was the
49:59 new Meta Meta doai so if you go to
50:06 meta and make an image so you can you
50:11 can be in chat mode so you can just say
50:14 let's see um
50:17 write a creative
50:20 brief
50:24 about um
50:31 um car
50:34 wax made of
50:38 uh uh turtles there's actually one
50:41 called Turtle Wax already made of meal
50:46 worms all right so it's going to make a
50:49 creative brief and then I'll go make an
50:55 add my
50:59 mockup with a clever
51:03 headline and it should yeah should make
51:06 us the
51:08 images and then I can just pick one of
51:11 those images or actually I can animate
51:13 one of those images and it'll animate
51:15 all four worm
51:23 shine worm shine
51:33 that's pretty
51:34 funny worm
51:36 shine
51:39 ah where' the other ones go did I lose
51:43 my other ones I guess
51:45 so very
51:49 funny worm
51:52 shine it's like Turtle Wax only
51:56 different
52:01 I mean not for nothing that's that's
52:03 pretty slick the the uh the words on the
52:07 back of the can are pretty bad
52:10 but for a free tool now they are
52:12 watermarking everything so you know I
52:15 don't know if there's going to be some
52:17 paid version also apparently this tool
52:20 is going to be incorporated in Instagram
52:22 and Facebook and everywhere right it's
52:24 going to be everywhere so I don't know
52:26 how many many millions or billions of
52:28 people use Facebook but it's a lot so
52:31 the fact that this is so simple um and
52:34 yet still pretty powerful is you know
52:37 there's probably a good chance that that
52:39 this brings a lot of people into AI you
52:42 know ask meta anything let's see um
52:46 explain quantum
52:49 mechanics as an
52:53 &m WP
53:04 pretty good verse one yo listen up I'm
53:06 about to spit some facts about a theory
53:08 so mind blown no need to react quantum
53:11 mechanics the rules of the small where
53:12 particles get weird and the laws Stand
53:16 Tall yeah and then it understands verses
53:19 and choruses I mean you know it's a it's
53:21 a good large language model right so you
53:23 can up vote it down vote it copy it to
53:26 the clipboard but super simple right and
53:30 this is I think this is actually a super
53:32 smart move on their part
53:34 because
53:36 it's you know this is the a this is not
53:39 a super technical audience right this is
53:42 why chat GPT took off because it was so
53:44 simple everyone understood how to use it
53:48 and now you know it's
53:50 funny chat GPT now has gotten way more
53:54 um complicated it's got all these
53:56 different modes you can upload things it
53:58 can do data analysis it's got custom
54:00 instructions this is just this it's just
54:03 like here's a box type crap in and let's
54:07 see fun play 90s music
54:12 trivia what 1995 album by Oasis became a
54:16 critical commercial success and is often
54:18 credited with helping shape the Brit pop
54:21 movement what's the story Morning Glory
54:24 I think that's it isn't it
54:27 that is
54:30 correct which 9s grunge band formed in
54:33 Seattle known for its hit Black Hole
54:36 Sun Black Hole Sun is Pearl Jam isn't it
54:39 or is it sound
54:42 garden come on all you old people know
54:44 it's sound
54:46 garden c yeah I think it's
54:48 [Laughter]
54:52 C
54:54 correct which '90s pop icon album The
54:57 Velvet Rope included the hit Got Till
55:00 It's Gone and together again I don't
55:02 [ __ ] know but it's got to be Mariah
55:05 Carey Velvet Rope A A is that right I
55:10 think that's wrong no damn it Janet
55:12 Jackson all right all right people with
55:15 your music
55:17 trivia there's no shaming the
55:23 host we gave up on shaming
55:28 yes no I'm Johnny I'm not going to be
55:30 asking that no sorry I don't want to get
55:32 booted off of Tik Tok sorry you can but
55:35 feel free to go ask
55:38 that wait for our
55:41 help all right what else do they have
55:44 here is their little simulate a mock
55:46 interview that's
55:49 interesting let's conduct a mock
55:51 interview what type of job or industry
55:54 what level oh that's kind of cool
55:55 actually
55:57 what type of job are you preparing for
55:59 what's your level of experience I think
56:01 the
56:02 the if these large language models could
56:05 crack everyone I know that's looking for
56:07 a job right now like they just can't get
56:08 an interview like much less prepare for
56:11 one like how do you get an interview
56:13 wait that was fun Kyle wait which one oh
56:15 the uh the interview I can go back to
56:17 that um okay so what type of job are we
56:21 preparing for software engineering
56:23 marketing I'll do marketing because I
56:25 know it
56:29 marketing what level of experience do
56:31 you have uh let's see
56:40 CEO what type of company are you
56:42 aspiring to lead an agency God know a
56:45 brand of Technology nonprofit Let's Go
56:52 technology oh the music oh the music was
56:55 interesting the music trivia I was going
56:57 to say this this is yeah well let's go
57:01 do something more interesting like where
57:03 can we do let's
57:07 see this might be kind of fun
57:14 uh was thinking we could we could flip
57:17 back and forth between two between two
57:19 large language models but I feel like
57:20 that would get
57:23 boring not sure I'm not sure people AI
57:28 trivia yeah we could do AI
57:31 trivia let's is is meta connected to the
57:34 internet um what is the
57:38 weather
57:40 in Denver right now
57:57 why is this not working
58:12 oh consist of snow with overcast Skies
58:16 view
58:17 sources WF from alpha huh all right well
58:22 I guess it is connected to the internet
58:24 um tell me about story vine.com
58:36 link
58:39 unavailable
58:43 um
58:45 who
58:47 uh please note websites are subject to
58:49 change I encourage you to visit link
58:53 unavailable uh how else would I know uh
58:56 let me see if I can go uh here's what
58:57 we'll do we'll go we'll go
59:11 archive okay so here's a PDF file let's
59:15 go back to meta and
59:17 say
59:20 summarize the PDF here
59:28 says
59:32 searching can't access the PDF you're
59:34 talking about summarize the PDF
59:41 here oh that's wild it gave us a Google
59:46 link so it sent us back to Google if it
59:49 couldn't do
59:52 it huh
59:58 I don't get meta's business model right
1:00:00 now this AI [ __ ] I don't have a [ __ ]
1:00:04 clue what they're doing they're just
1:00:05 open sourcing it they're giving it away
1:00:08 it's this is now pointing to Google I
1:00:11 guess are they
1:00:16 partners get lyrics for yud AI trivia
1:00:20 let's do AI trivia all right we'll do
1:00:22 we'll do AI trivia here
1:00:24 um let's
1:00:28 play and AI
1:00:34 trivia game with the latest
1:00:39 news about
1:00:44 AI meta is
1:00:46 tanking yeah
1:00:56 [Music]
1:01:02 here are five trivia questions okay
1:01:05 let's see uh no I want it to
1:01:11 be
1:01:14 interactive ask one at a time all
1:01:19 right I'm going go plus we make this a
1:01:21 little
1:01:25 bigger all right all
1:01:32 right what's the name of the AI model
1:01:34 recently gained attention for its
1:01:35 ability to generate realistic images of
1:01:37 artwork including portraits of fake
1:01:41 people Dolly stable diffusion mid
1:01:44 Journey
1:01:46 llama well everything but
1:01:50 llama what's the name of it that gained
1:01:52 attention for its ability to generate
1:01:54 realistic images
1:01:57 oh this is weird see I guess
1:02:01 see mid
1:02:04 Journey all of the
1:02:08 above the correct answer was indeed mid
1:02:10 Journey question two which Tech Giant
1:02:12 okay this is a
1:02:14 shitty AI trivia thing so let's go try
1:02:17 chat GPT and see if it's any
1:02:20 better um
1:02:27 make me an interactive trivia game about
1:02:31 the latest news in
1:02:40 AI oh which by the way I got to show you
1:02:45 something let's design a simple okay
1:02:48 game outline 10 questions per round
1:02:50 multiple choice topics covered recent
1:02:54 advancements all right it's asking a
1:02:56 multi together hang
1:03:02 on okay that's great I want to play it
1:03:04 with some friends so just ask one
1:03:06 question at a time and don't move on to
1:03:09 the next one until we answer the
1:03:16 first okay which AI model released by
1:03:19 open AI in 2023 is known for its ADV
1:03:23 Advanced language capabilities
1:03:27 see this is going to be boring I
1:03:33 [Music]
1:03:35 think
1:03:37 [Laughter]
1:03:39 correct which technology company
1:03:42 announced significant improvements in AI
1:03:45 based Health Care
1:03:48 Imaging Google
1:03:51 Amazon Microsoft and Apple it's got to
1:03:53 be Google with Deep Mind
1:03:57 a you think is that
1:04:01 right Danielle's going with Amazon I'm
1:04:05 going
1:04:08 Google correct they announced
1:04:11 significant advancements and AI based
1:04:13 Health Care
1:04:14 Imaging question three what ethical
1:04:17 issue is most commonly associated with
1:04:19 the deployment of AI in public sectors
1:04:21 bias reduction and employment energy
1:04:24 consumption privacy
1:04:26 I think the loudest one right now is
1:04:28 bias but I
1:04:33 think
1:04:36 correct as of 2024 which country leads
1:04:40 the number of AI research papers
1:04:42 published
1:04:43 annually the US China Germany UK
1:04:49 a I think we're leading here
1:04:54 yep all right that was good but let me
1:04:57 show you something notice how my chat
1:04:59 GP chat GP he notice how my chat GPT is
1:05:05 behaving each of my responses is
1:05:08 numbered see it says response number six
1:05:10 and then it
1:05:13 um it summarizes the the response and
1:05:17 then it gives me the date and time well
1:05:19 actually didn't give me the time did it
1:05:21 give me the date though so and then it
1:05:23 gives me the uh token context window so
1:05:27 how many tokens has it used out of
1:05:29 4,000 um how I'm doing that is in my
1:05:32 custom instructions if you say customize
1:05:35 chat
1:05:37 GPT the the there's two places there's a
1:05:40 who are you so I just put you're a
1:05:42 helpful assistant who's not afraid
1:05:45 afraid to criticize as well as support
1:05:48 if it is appropriate and then how would
1:05:50 you like chat GPT to respond so I say
1:05:53 for every response give us the
1:05:56 sequential response count followed by a
1:05:59 response title that summarizes the
1:06:01 response also include a token context
1:06:05 window uh counter as a ratio of the of
1:06:10 the oh tokens tokens used versus tokens
1:06:15 available let's see and the
1:06:19 date comma
1:06:22 time and then I'm going to make it in
1:06:24 MST in
1:06:27 Mountain Standard
1:06:29 Time because that's where I am all right
1:06:33 so let's go chat J and
1:06:38 say create a
1:06:44 charter oh now it didn't give
1:06:48 me now it didn't give me date in time so
1:06:51 I'm going to fix that let's see let's go
1:06:54 in here I'm going to get rid of the
1:06:56 token count cuz I think it's
1:06:58 [ __ ] also
1:07:00 include the
1:07:06 date and
1:07:08 time and I'll put this in parentheses
1:07:12 in
1:07:16 MST it is MDT you know someone said that
1:07:19 to me the other day I'm going to just
1:07:22 put in mountain time cuz I'll let it
1:07:26 figure out in my MDT or MST Mountain
1:07:30 [Laughter]
1:07:33 Time Dam see yes see see y
1:07:37 palas all right uh explain Nostalgia to
1:07:41 a kindergarten Here We Go
1:07:44 response 3:45 all right that is totally
1:07:48 hallucinated the time time is totally
1:07:51 hallucinated but the date is correct so
1:07:54 we can fix that we will get rid of the
1:07:59 time also include the date there we
1:08:05 go but how cool is that now so all of my
1:08:11 responses have titles to them and a date
1:08:15 and then it gives you the response and
1:08:17 the purpose of this is if you're doing a
1:08:20 really long conversation you can kind of
1:08:23 you know remember oh that was like that
1:08:24 was response number s
1:08:26 that was the good one is that slick
1:08:29 somebody gave me that idea in here like
1:08:31 a week ago so I've been I've been
1:08:33 slightly refining it ever since and I
1:08:35 had it turned off and I'm like Hey
1:08:37 where's my title and I was like all
1:08:38 right that's not the way this works but
1:08:41 it is
1:08:42 now
1:08:44 sweet sweet sweet
1:08:48 [Music]
1:08:51 sweet Kyle I keep having to take
1:08:54 screenshots is that a good thing thing
1:08:56 or a bad
1:09:08 [Music]
1:09:23 thing let's go make a song
1:09:26 all right so here we are at udio if you
1:09:28 don't know
1:09:29 udio oh for those custom instructions
1:09:32 hang
1:09:33 on I'll go I'll go hang on them long
1:09:36 enough so that you can see them so
1:09:39 here's where here's where I
1:09:42 landed all right screenshot that for
1:09:45 every response give it a sequential
1:09:47 response count followed by a response
1:09:50 title that summarizes the response and
1:09:53 there's probably a way that you could
1:09:55 write that that's more like English but
1:09:57 that was all I could get out of my head
1:09:58 at that moment and then also include the
1:10:04 date and then it does
1:10:06 that oh you got it okay cool beauty
1:10:09 beauty beauty beauty all right let's go
1:10:12 make a
1:10:13 song so if you don't know
1:10:16 yudo let's go to let's go to chat
1:10:20 jpt do we want to go there let's go
1:10:23 let's go give meta a try
1:10:26 since meta is the new kid on the Block
1:10:28 let's
1:10:33 say I want you to act like a Grammy
1:10:37 award-winning
1:10:42 songwriter and give
1:10:44 me 10 memorable
1:10:48 concepts for songs
1:10:56 with the basic story the
1:11:01 title and a musical style and
1:11:08 feel all
1:11:11 right title lost in the moment romantic
1:11:15 ballad ghosts of our past fire in my
1:11:18 soul whisky and regrets a melancholic
1:11:22 folk ballad about drowning out sorrow
1:11:24 searching for forgiveness
1:11:26 acoustic guitar driven with a hint of
1:11:28 Blue blueg Grass Allah Jason
1:11:31 Isbell this one's interesting
1:11:35 because I think it's going to reject it
1:11:39 because of
1:11:42 um I think it's going to reject it
1:11:45 because of
1:11:47 uh it says Jason Isbell but we'll see so
1:11:50 let's
1:11:51 say great write the lyrics for number
1:11:54 four and and
1:11:57 include all the description information
1:11:59 as
1:12:05 well now udio you can write
1:12:11 um it can write 30 seconds at a time
1:12:14 which is essentially a verse of
1:12:17 music so what we can
1:12:20 do we'll copy all of this and verse
1:12:24 one then we'll go to
1:12:27 yudo we're going to do custom lyrics
1:12:30 we're going to pop them in
1:12:39 here I don't think there's a place for
1:12:43 us well actually
1:12:45 maybe let me copy all of this stuff let
1:12:48 me see if I can put it all in there
1:12:55 well maybe we'll
1:12:59 see someone told me they turned off
1:13:01 manual mode Joker didn't you tell me
1:13:03 manual mode was turned off I still see
1:13:05 manual mode in
1:13:07 mine all right so let's see if it does
1:13:11 this it might kick it out because we use
1:13:14 the name Jason Isbell he might be famous
1:13:18 enough my Creations oh it's making
1:13:21 them now artist replaced Jason Isbell
1:13:25 replaced with male vocalist Americana
1:13:27 singer songwriter Northern American
1:13:29 Music Regional music Heartland Rock okay
1:13:32 that's not bad
1:13:34 actually I like I like what it's doing
1:13:38 there like it didn't just block it it
1:13:41 didn't just say hey you're using someone
1:13:42 else's name you can't do that it's
1:13:44 saying hey you're using someone else's
1:13:46 name you can't do that but I know enough
1:13:49 about this stuff to know that if I put
1:13:50 in these other terms it'll be in that
1:13:52 neighborhood I think that's actually
1:13:54 pretty smart
1:13:55 I like it wasn't you
1:13:58 oh other fillers okay let's
1:14:03 see I woke up this
1:14:06 morning with tears in my
1:14:10 eyes the memories of you leaving they
1:14:14 still linger and
1:14:16 Rise I reach for my
1:14:21 B those are really bad lyrics aren't
1:14:23 they oh meta
1:14:27 let's grab
1:14:28 this let's go to chat
1:14:32 JP and see if it do any
1:14:36 better AC is Grammy
1:14:42 awardwinning act as
1:14:45 a
1:14:47 songwriter and write me the lyrics for
1:15:00 in the dim light of a fading day Rusty
1:15:02 strings soft guitar plays I pour my
1:15:04 heart like this old whiskey neat Echoes
1:15:07 of view in every
1:15:09 Retreat
1:15:20 um the lyrics are a little simple and
1:15:23 predictable can you make them more
1:15:25 nuanced
1:15:28 metaphoric and non-
1:15:35 cliche The Dusk spills its secrets on
1:15:38 the dusty floor silent strings strum the
1:15:41 echo's lore in the Amber por a mirrored
1:15:43 Sky Reflections where the swallows fly
1:15:47 chorus whiskey and regrets weaving
1:15:49 through my veins a tap a tapestry
1:15:59 you canot use the word tapestry wait you
1:16:03 canot use the word tapestry or delve or
1:16:07 any ridiculous words that large language
1:16:10 models overuse this is an Americana
1:16:13 heartfelt
1:16:15 song and they don't use words like
1:16:17 tapestry and delve you weirdo
1:16:25 it didn't even
1:16:28 [Music]
1:16:32 apologize all right let's
1:16:35 see even if falls does okay whiskey and
1:16:39 regrets Partners in my song melodies of
1:16:42 right and wrong each sip unfolds stories
1:16:45 Untold in this drink our lives unfold
1:16:48 yeah it's still kind of shitty isn't it
1:16:49 but let's let's it it at least it's got
1:16:51 to be better than what we had before so
1:16:53 let's go grab this we'll go back to
1:16:58 yudo and we are going to blast this one
1:17:01 out of
1:17:06 here all
1:17:11 right a tapestry slams keyboard AI Karma
1:17:16 Lords will get you for that one day I
1:17:18 know when these things get memory
1:17:19 they're going to be like you know when
1:17:20 you were a smartass to
1:17:23 me yeah I
1:17:26 now are these free versions um Udo is
1:17:30 free right now ud.com and it's really
1:17:34 good and they give you 600
1:17:36 Generations a month and they do two
1:17:40 songs per generation so it's it's a lot
1:17:43 it's a lot of music they're giving you
1:17:44 for nothing
1:17:49 [Music]
1:17:50 honey evening
1:17:53 Falls dust settles in in the light hate
1:17:59 it sounds better evening Falls dust
1:18:03 settles in the light this is
1:18:06 good Jason is belly quiet cords float
1:18:10 through the fing
1:18:12 [Music]
1:18:14 night in the glow of Amber Skies
1:18:19 Retreat this is
1:18:21 good mirrors of the past we can't
1:18:25 [Music]
1:18:30 okay
1:18:31 [Applause]
1:18:35 beautiful all right so now we're GNA so
1:18:38 the way you do this oh you can't see on
1:18:39 my screen hang on let me dim it a little
1:18:42 so I can either remix it which I guess
1:18:44 just rolls it I can extend it or I can
1:18:46 publish it so I'm going to extend
1:18:48 it and then what you do is I'm going to
1:18:51 pop in the chorus now and you can either
1:18:54 add an intro a section before this
1:18:57 section a section after the section that
1:18:59 we just heard or an outro so I'm just
1:19:02 going to yeah the gravity voice was
1:19:04 really good Chris Kristopher yeah it was
1:19:06 really good all right so let's see what
1:19:08 this does with the chorus hopefully it
1:19:10 it does a good Chorus for
1:19:13 us isn't this amazing that you can just
1:19:17 and this is one of the things that I
1:19:20 love doing is is just being able to like
1:19:24 jump from from tool to tool to Tool and
1:19:27 you know I don't know 2 years from now
1:19:28 when these things are all singing all
1:19:30 dancing multimodal tools everything will
1:19:32 just happen in the same interface but
1:19:34 right now it's really fun to be able to
1:19:36 just you know well let me go try this in
1:19:38 meta let me go try this over here maybe
1:19:40 because because they all have such
1:19:42 different personalities in fact let's do
1:19:45 another experiment let's go
1:19:48 grab the description of this and go to
1:19:52 Pi because pi
1:19:55 is designed to be conversational and I'm
1:19:58 wondering I'm wondering if Pi might do
1:20:01 lyric writing better than these other
1:20:04 tools pi.
1:20:10 a all
1:20:14 right so act as a Grammy awardwinning
1:20:17 singer songwriter whiskey and regrets
1:20:20 there's a bottle on the table glass half
1:20:22 full of Shame that's really good haunted
1:20:24 by the memor
1:20:26 I can can out run I can outrun A
1:20:29 Thousand Dreams forsaken drowned in
1:20:31 amber rain each sip a bitter reminder of
1:20:34 all that I've become it's pretty good
1:20:36 whiskey and regrets are all that I have
1:20:38 left an empty bottle of broken heart a
1:20:40 soul
1:20:41 B I'm searching for forgiveness in this
1:20:44 bottle's depths but all I find are
1:20:46 Whispers of mistakes and missteps this
1:20:49 is really
1:20:51 good this is really
1:20:54 good you know what you know what P got
1:20:57 here better than anything else I've seen
1:21:00 is the chorus like it understands that
1:21:03 the chorus has to tell the whole
1:21:05 story whiskey and regrets are all that I
1:21:08 have left an empty bottle a broken heart
1:21:10 a soul bre I'm searching for forgiveness
1:21:13 in the bottle's depths but all I find
1:21:15 are Whispers of mistakes and
1:21:17 missteps that's good we'll probably miss
1:21:20 that too Sizzle all right I'm gonna
1:21:25 I think I think we're GNA I think we're
1:21:27 going to punt on this other other
1:21:28 version but let's see how it
1:21:31 [Music]
1:21:34 did evening Falls dust settles in the
1:21:39 [Music]
1:21:46 light whiskey and
1:21:49 regrets Partners in my
1:21:53 song Melo
1:21:55 of right and
1:21:57 [Music]
1:21:59 wrong each zp
1:22:04 [Music]
1:22:07 unfold whiskey and
1:22:10 [Music]
1:22:14 regrets here you know what we're going
1:22:15 to do we're GNA also have
1:22:18 piie um give
1:22:22 me
1:22:24 10 alternate
1:22:30 titles empty bottles Broken Dreams
1:22:33 that's not bad drowning in amber
1:22:36 rain how about just amber
1:22:44 rain let's call it amber
1:22:49 rain
1:22:51 the
1:22:53 um chorus should still start with
1:22:56 whiskey and regrets but then you should
1:22:59 incorporate amber rain into the chorus
1:23:08 appropriately Amber rain's pretty
1:23:11 good cloudy amber
1:23:13 rain amber rain for the win yeah right
1:23:18 Corey I got a thumbs down I got a thumbs
1:23:22 up but can it do rap yeah it can
1:23:28 rrap whiskey and regrets my constant
1:23:31 companions in the
1:23:35 night amber rain falls from
1:23:50 the I'm going to
1:23:53 say keep your original first
1:23:58 verse but only start the chorus with the
1:24:02 line whiskey and
1:24:04 regrets
1:24:12 period there's a bottle on the table
1:24:15 glass half full of
1:24:18 Shame whiskey and regrets are all that I
1:24:20 have left an empty bottle of broken
1:24:21 heart a soul bere and searching for
1:24:24 forgiveness
1:24:25 rest of the lyrics wait okay damn
1:24:33 it okay but the song is called Amber
1:24:36 rain so the chorus has to have amber
1:24:38 rain in it possibly even more than once
1:24:42 period get with the
1:24:44 [Laughter]
1:24:52 program never meant to cause you any
1:24:54 trouble and R Ember rain whiskey and
1:24:56 regrets flow and Ember
1:25:11 rain uh okay we got to go we got to fix
1:25:14 this [ __ ] that was good who did that
1:25:16 Corey oh I on China wait yeah never
1:25:20 meant to cause you any trouble wait
1:25:22 whiskey and regrets
1:25:30 see
1:25:31 this okay this is the problem
1:25:35 people is is
1:25:38 uh when you actually try to do real work
1:25:41 and you pay attention to this [ __ ] it's
1:25:44 like it's like it they're just janky
1:25:48 these things are [ __ ]
1:25:51 janky RI them frig them all right
1:25:55 let's see copy okay let's go over to
1:25:59 notepad we'll go here we'll
1:26:02 here we'll go so you can see it there's
1:26:06 a bottle okay whiskey and regrets are
1:26:08 all that so now let's go back and
1:26:11 get one of the choruses that didn't suck
1:26:25 whiskey and regrets are all that I have
1:26:33 left
1:26:39 whet all right this is fine we'll just
1:26:42 grab this
1:27:05 B once you start correcting AI gets
1:27:08 worse it totally gets worse when you
1:27:10 start correcting it Whiskey Whiskey
1:27:12 regrets are all that I have
1:27:19 left wait whiskey and regrets
1:27:33 um
1:27:36 deepening the
1:27:40 pain
1:27:43 um deepening the
1:27:46 pain I see
1:27:51 it in your eyes
1:27:56 and in the
1:27:59 amber rain
1:28:04 okay and then we'll repeat this
1:28:07 one for the last
1:28:09 one we're gonna have a hit song Here
1:28:14 people okay
1:28:18 so so it's now called Amber rain okay
1:28:21 copy this
1:28:24 oh it put amber rain in the first verse
1:28:27 God dang
1:28:28 it okay wait a Thousand Dreams
1:28:33 forsaken
1:28:36 um hit like afraid uh
1:28:40 hit me like a
1:28:43 train okay there we go good Lord verse
1:28:48 one okay we have a verse back to
1:28:52 yudo okay we want to read reset this oh
1:28:55 crap I didn't want to do that well too
1:28:57 late now all right so lyrics go
1:29:04 here this thing
1:29:11 goes where's my prompt
1:29:18 edit strange strange
1:29:26 okay title
1:29:28 okay melancholic folk
1:29:32 ballad copy all right
1:29:36 udio paste this in
1:29:39 here I'm leaving Jason Isbell in there
1:29:41 I'll let it correct it for
1:29:43 me all right there we
1:29:50 go when are you going on tour I I got to
1:29:53 finish the song first I can't get
1:29:55 through these [ __ ] lyrics check my
1:29:57 previous two posts Kyle
1:30:00 [Music]
1:30:03 okay it all just sounds the
1:30:06 same I'm drowning here under the amber
1:30:10 rain yeah I know this good I'm tired of
1:30:12 hearing myself complain over and over
1:30:15 again whiskey and regrets yeah these are
1:30:17 good
1:30:19 Corey um
1:30:24 yeah these are
1:30:31 good yeah we gota we gotta oh you know
1:30:34 what Corey do me a favor go go post them
1:30:37 either DM me DM me on um on Mighty
1:30:42 networks so just post it to a regulars
1:30:44 and then I'll go grab them because I
1:30:45 don't want to do it I don't want to do
1:30:46 it here live I just kind of want to make
1:30:47 the song and get it
1:30:50 extended but I like I like the the
1:30:53 changes that you made artist place all
1:30:54 right let's see if we
1:30:59 got there's a B on the table glass half
1:31:04 full of
1:31:06 Shame haunted by the
1:31:09 memories that I can out run a Thousand
1:31:14 Dreams
1:31:15 forsaken hit me like a train ex simp a
1:31:20 bitter
1:31:21 reminder of all that I've
1:31:26 [Music]
1:31:35 become there's a bottle on
1:31:39 table glass half full of
1:31:42 Shame haunted by the
1:31:45 memories that I can't out
1:31:48 run Thousand Dreams
1:31:51 forsaken hit me like a train each sip of
1:31:55 bitter
1:31:57 reminder of all that i'
1:32:00 become yeah that's
1:32:02 [Music]
1:32:08 good see it in the glass and see it in
1:32:11 the Ember rain yeah yeah that's pretty
1:32:13 good all right let me go grab
1:32:18 the Le
1:32:21 lorus Flo amber rain whiskey regrets
1:32:24 Ember and flows again okay
1:32:27 copy Udo
1:32:31 extend paste those
1:32:35 lyrics extend
1:32:42 it and this is you know it's funny like
1:32:46 you know I was talking earlier about
1:32:47 this this shitty bill that that the
1:32:50 Colorado Senate's trying to
1:32:53 pass and like the their thing about you
1:32:55 know stuff that's been produced with
1:32:57 ai's got to be labeled and it's like
1:33:00 well
1:33:02 like if I spend like even if this is the
1:33:05 tool that's that's creating the music if
1:33:08 I'm like rewriting the lyrics and doing
1:33:11 27 versions of the song and and
1:33:14 potentially even taking it into some you
1:33:16 know DAW and you know splitting the
1:33:19 tracks out and fixing them and doing all
1:33:21 that [ __ ] like at what point yeah this
1:33:23 is very John Prim as
1:33:25 at at what point does this not when it
1:33:28 is it not Ai and like why does it
1:33:30 [ __ ]
1:33:33 matter you
1:33:36 [Music]
1:33:40 know there's a bottle on the
1:33:43 table glass half full of
1:33:47 Shame haunted by the
1:33:50 memories that I can't out
1:33:53 run that Thousand Dreams
1:33:56 forsaken hit me like a train each sip of
1:34:00 bitter
1:34:01 reminder of all that I've
1:34:04 [Music]
1:34:10 become that's very John
1:34:16 Prime whiskey and regret FL the Amber R
1:34:21 nope
1:34:27 whiskey and regrets flow in the amber
1:34:32 rain in each bitter drop by taste a
1:34:37 thousand
1:34:38 stolen sips of Dreams
1:34:42 foren heart
1:34:45 nope try it
1:34:50 again Market a relabeled whiskey amber
1:34:53 rain
1:34:54 it's gone folky it
1:34:57 has time for a duo with Brazilian
1:35:01 Dylan posted in in IR regulars just a
1:35:04 diddy definitely an acoustic guitar song
1:35:07 it is all right let me go to regulars
1:35:09 let me get Corey's lyrics because
1:35:11 they've got to be better than what's
1:35:13 what this thing squirted out and I'm
1:35:14 editing in real
1:35:20 time I'm so tired of hearing myself
1:35:22 complain over and over again
1:35:29 oh yeah that's good this is a good
1:35:30 chorus this is really good okay so let's
1:35:32 go back here screw that [ __ ] that I put
1:35:34 in there so we'll take
1:35:36 the 32nd one this is it right yeah so we
1:35:41 extend this one we'll throw these in
1:35:45 we'll we'll call them A
1:35:52 Chorus add the section all right there
1:35:55 we
1:35:55 go I think this is going to be much
1:35:58 better that's mo better cor Sandler
1:36:01 that's mo
1:36:02 better MOA
1:36:08 [Music]
1:36:11 Blues it needs will frell in the
1:36:14 Triangle
1:36:18 Ting hey and Murphy what shaking we're
1:36:21 just sitting here writing some Jason
1:36:23 isbel
1:36:25 but not really we can't use his
1:36:28 name it matters because people have to
1:36:30 make money to live
1:36:32 marketing oh markeing AI is an attempt
1:36:35 to preserve the profitability of
1:36:38 creative
1:36:39 art yeah
1:36:41 but the the whole creative art um the
1:36:45 whole create but that's not going to
1:36:46 that's not going to save it ad agencies
1:36:49 are not going to not use this um it it's
1:36:54 it's not going to save it to label it cu
1:36:56 cu artists are going to use these tools
1:36:59 even musicians are going to use these
1:37:01 tools and it's it's not like they're
1:37:03 just going to push a button and out
1:37:05 comes the song they're going to be
1:37:06 collaborating with other artists with AI
1:37:09 it's just going to be a tool in the mix
1:37:11 just like you know just like a Daw is a
1:37:13 tool in the
1:37:14 mix um yeah but don't get me wrong I I
1:37:19 empathize and it's a [ __ ] mess right
1:37:20 now
1:37:22 but let's see
1:37:28 I'm so
1:37:29 tired of hearing my self complaint it's
1:37:35 good over and
1:37:38 over and over
1:37:42 again whiskey and
1:37:45 regrets and it all just sounds the
1:37:49 same don't watch me
1:37:51 drowning in my Amber
1:37:54 yeah that's good there you go that's the
1:37:56 chorus that's not a I get in like that
1:37:58 version but let's hear this
1:38:01 [Music]
1:38:06 one I'm so tired of
1:38:10 hearing my self complain that's not
1:38:14 right all right let's
1:38:18 [Music]
1:38:20 see when it said artist replace did it
1:38:23 give me
1:38:26 we've replaced isbel with male
1:38:39 vocalist let me grab all that and go up
1:38:43 here and replace it
1:38:56 write a jingle for your channel that's
1:38:57 not a bad
1:39:03 idea edit out so tired to just tired all
1:39:07 right
1:39:27 I'm tired of
1:39:37 here I'm so wait I'm so I No it should
1:39:42 be I'm so sick and
1:39:45 tired of hearing myself complain over
1:39:49 and over and over
1:39:58 well let's see let's see what it did
1:40:00 with the last one maybe it maybe it did
1:40:02 it
1:40:03 [Music]
1:40:10 okay I'm so tired
1:40:14 ofing I'm so freaking tired who just
1:40:16 said that Corey
1:40:18 yeah how's the Musical coming along uh
1:40:21 it's good I didn't work on it a ton this
1:40:24 week I worked on a couple of
1:40:26 songs I'm
1:40:28 so fraking
1:40:32 tired I'm so freaking tired of hearing
1:40:34 myself
1:40:38 complain I'm so freaking tired if hear
1:40:41 myself complain
1:40:56 let's try this one extend it all right
1:41:00 it's going it's going people
1:41:14 [Music]
1:41:28 CU I think what would be really cool to
1:41:30 do is just write a bunch of these and
1:41:33 then just start singing them and just
1:41:35 you know what I mean like learn them on
1:41:37 guitar and just see which ones see which
1:41:40 ones you internalize and and feel real
1:41:44 right
1:41:50 [Music]
1:41:56 I'm so fing tired
1:42:03 thicking I'm so freaking tired okay
1:42:06 can't do freaking
1:42:08 tired did I spell it wrong oh I
1:42:13 did user
1:42:20 error that's really
1:42:22 funny oh freaking H it's all right
1:42:25 freaking freaking I'll change freaking
1:42:27 for the next one cuz this probably won't
1:42:29 be good I'm so freaking
1:42:34 tired
1:42:36 ficking we just we'll just blast another
1:42:38 one in there there we
1:42:40 [Laughter]
1:42:44 go
1:42:46 ficking I'm so freaking tired
1:42:55 [Music]
1:43:02 I only want to see you laughing in the
1:43:05 amber rain amber rain amber rain oh
1:43:08 [Laughter]
1:43:12 Joker
1:43:14 oh you could say now see if you say
1:43:17 something like drinking my amber rain it
1:43:19 makes it too obvious
1:43:23 [Music]
1:43:29 I'm so freaking tired of
1:43:32 [Music]
1:43:33 hearing
1:43:36 [Music]
1:43:42 nope I'm so freaking tired there of
1:43:47 hearing complain there you
1:43:49 go over and
1:43:52 over oh over
1:43:55 again whiskey and
1:43:57 regrets and it all just sounds the
1:44:01 same don't watch me
1:44:06 drowning in
1:44:08 my oh that's good that one's good all
1:44:12 right now can I heart that one oh they
1:44:16 fix their heart their
1:44:19 interface okay let's try this one
1:44:25 I'm so freaking tired
1:44:34 nope there's a bottle on the
1:44:37 table glass half full of
1:44:41 Shame haunted by the
1:44:43 [Music]
1:44:46 [Applause]
1:44:48 memories I'm so freaking tired of
1:44:51 hearing nope okay all right all right so
1:44:53 we'll do this one we're going to extend
1:44:55 it I'm going to add an Intro
1:44:58 instrumental go so let's see if we can
1:45:00 get a decent beginning to this
1:45:06 thing don't just stand there and watch
1:45:08 me drowning in my amber rain that's
1:45:10 pretty good yeah that's pretty good
1:45:12 don't just stand there all right but but
1:45:15 wait I've got
1:45:16 a I see now I forget where I've got to
1:45:21 let me go fix it in let me grab
1:45:26 uh this
1:45:31 one let's
1:45:44 see don't just stand
1:45:47 there watching don't
1:46:04 oh hang
1:46:11 on let's go put this over here
1:46:47 all right got it there's I think there
1:46:50 needs to be something else before over
1:46:51 and over and over again but let's go see
1:46:54 if we got anything decent for the
1:46:55 beginning of this
1:47:06 thing like
1:47:11 that that's a
1:47:15 [Music]
1:47:18 little ah champy
1:47:23 [Music]
1:47:24 [Applause]
1:47:28 [Music]
1:47:33 yeah you like that right
1:47:36 champy
1:47:38 yeah it's a
1:47:42 hit Champion
1:47:44 [Music]
1:47:50 approved there's a bottle on the table
1:47:54 glass half full of
1:47:57 Shame haunted by the
1:48:00 memories that I can out
1:48:03 run Thousand Dreams
1:48:06 forsaken hit me like a train each sip of
1:48:10 bitter
1:48:11 reminder of all that I've
1:48:14 [Music]
1:48:21 become I'm so freaking
1:48:23 tired of hearing myself
1:48:28 complain over and
1:48:30 over over
1:48:33 again whiskey and
1:48:35 regrets and it all just sounds the
1:48:39 same yeah course needs to be fixed but
1:48:41 that's all right we're going to extend
1:48:42 it we're going to add another
1:48:44 verse all right we're going to add
1:48:47 another verse there's a picture in my P
1:48:49 Pocket faded and worn a relic of the
1:48:51 life I want had
1:48:55 known eyes full of wonder Hearts so
1:48:57 young before this road of Sorrow that
1:48:59 I've stumbled
1:49:01 [Music]
1:49:04 on I guess that's
1:49:09 okay anyway if you're not playing with
1:49:11 yudo you should be playing with yudo
1:49:13 because it's free and it makes cool
1:49:17 music we have a
1:49:21 winner champion
1:49:23 approved all right you'll hear one more
1:49:25 champy
1:49:31 [Music]
1:49:47 [Music]
1:50:02 all
1:50:08 right over and
1:50:11 over over
1:50:13 again whiskey and
1:50:16 regrets and it all just sounds the
1:50:20 same don't watch me drowning
1:50:23 [Music]
1:50:26 in my
1:50:26 [Music]
1:50:36 am there's a picture in my
1:50:40 pocket faded and
1:50:44 worn relic of the life that I once had
1:50:50 known
1:51:02 am there's a picture in my
1:51:05 pocket faded and
1:51:09 worn a relic of the life that I once had
1:51:15 known eyes full of
1:51:18 wonder and a heart so
1:51:21 young before this road of Sorrow that i'
1:51:26 stumbled on all right we like that one
1:51:29 it's good enough it's got it's gotten
1:51:32 itself a little confused with the Rhythm
1:51:34 but that's
1:51:35 okay we're going to throw this chorus
1:51:39 back in there let's see
1:52:17 I'll grab an outro we'll do an outro
1:52:26 oh did I just close that window I did
1:52:28 I'm an
1:52:29 [Laughter]
1:52:34 idiot you big dumb
1:52:37 dumb like Creations all right the last
1:52:40 one we're going to extend it we're going
1:52:42 to do an outro we're going to do lyrics
1:52:44 we're going to put the outro lyrics
1:52:47 here let's see I walk toward the future
1:52:51 with with with hope in my
1:52:57 heart leaving whiskey in regrets forever
1:53:00 in the
1:53:03 dark look toward my
1:53:15 future
1:53:17 pouring amber
1:53:19 rain I walk toward the future
1:53:40 [Music]
1:53:43 wait the bottle's almost empty my soul
1:53:46 is almost free yeah that's better almost
1:53:51 free that's better
1:53:53 the past just a
1:53:55 shadow deep in my
1:53:58 memory I walk toward the future we need
1:54:02 like um oh how about I steam toward the
1:54:08 future a locomotive
1:54:12 train there we go that's very John Hyatt
1:54:16 I steam toward the future a locomotive
1:54:18 train
1:54:33 um let's see leaving leaving whiskey and
1:54:36 regrets now that's not where um pouring
1:54:39 embran let's see I steam toward the
1:54:41 future oh wait
1:54:48 um um I turn
1:54:58 toward the
1:55:10 past I
1:55:18 beckon moments
1:55:24 past pouring amber rain and there wasn't
1:55:27 a dry eye in the house all right outro
1:55:30 extend let's do this
1:55:42 [Music]
1:55:58 [Music]
1:56:02 version 2.2.2
1:56:05 point1 oh that's nice they're numbering
1:56:07 them now that's nice the the
1:56:10 extension oh that's really good they
1:56:12 added that so look here's extension
1:56:14 version two here's extension version
1:56:18 2.2 and
1:56:20 2.1 and then 2.2.2 and 2.
1:56:24 2.2.1 that's really good cuz that was
1:56:27 one of the hard things all right before
1:56:31 this road eyes full of
1:56:35 wonder and a heart so
1:56:38 young before this road of Sorrow that
1:56:42 I've stumbled
1:56:44 [Music]
1:56:51 on the B bot's now
1:56:54 empty my soul is almost
1:56:58 free the past just a
1:57:01 shadow deep in my
1:57:05 memory I steam towards the
1:57:08 future a locomotive
1:57:11 train I beckon moments
1:57:15 past por
1:57:20 and yeah not not hard
1:57:29 [Music]
1:57:38 the bottles now empty my soul is almost
1:57:43 free the past just a
1:57:46 shadow deep in my memory memy I steam
1:57:51 towards the future
1:57:53 a locomotive
1:57:56 train be moments
1:57:59 pass amber rain yeah that's not good all
1:58:03 right well we'll leave it with that this
1:58:06 song didn't quite work it was close but
1:58:09 it was just there was something off on
1:58:11 the Rhythm on it the whole time and it
1:58:13 never quite it never quite got it but
1:58:14 that's all right that's all right this
1:58:16 is how we learn people all right I'm
1:58:19 getting out of here hope you had fun
1:58:21 tonight um happy Friday night everybody
1:58:24 follow my
1:58:26 channel um subscribe to the lives keep
1:58:29 showing up keep coming back let's see
1:58:31 who is in here
1:58:36 tonight Papa JJ
1:58:39 Joey steo peace out good to see you
1:58:42 Corey Sandler thanks for the uh lyric
1:58:44 rating fun random thoughts by TK good to
1:58:47 see you pop JJ Dr J Silver
1:58:50 Fox good night
1:58:54 laughing Thunder was in here who else
1:58:58 unraced two block Tom laughing
1:59:02 Thunder Lily
1:59:08 [Music]
1:59:11 bigons Danielle oh no good
1:59:17 night good night Danielle all right
1:59:20 everybody Apple user good to see you
1:59:23 Robert Rossy thank you so much as always
1:59:25 Brian Whitney great to see you virtual
1:59:27 assistant Amelio's wife thumbs
1:59:29 up thumbs up it's a good
1:59:32 thing come to the play shop oh yeah come
1:59:35 to the play shop tomorrow uh if you're
1:59:37 not uh if you're not registered you can
1:59:40 be if you go to there go to the
1:59:44 salon. there's two important links there
1:59:47 above the AI Salon Community there's a
1:59:49 letter to the Colorado Senate that I
1:59:52 want you to sign and then there's the
1:59:54 play shop that I'm giving tomorrow from
1:59:56 9:00 a.m. Pacific
1:59:59 until 1 Pacific 4
2:00:02 hours um there's a link to that and if
2:00:05 you use the code Salon fun you got money
2:00:09 off a lot of it all right peace
2:00:12 out have a good night everybody