Feb 21, 2024 - Embracing AI: Transforming Creativity and Community

Description

Chapters

Transcript

0:02 [Music]
0:09 [Applause]
0:13 [Music]
0:20 [Music]
0:29 oh
0:31 [Music]
0:35 AR you
0:37 boy with a quiet
0:40 voice and impeccable
0:43 [Music]
0:47 style don't ever let them steal a
0:51 joy and gentle
0:53 ways in your face dress you down and
0:57 tell you that your place is in the
0:59 middle when they hate the way you
1:03 [Music]
1:05 shine I see you tugging on your
1:08 shirt trying to hide inside of it hide
1:13 how much it
1:15 hurts you let them
1:18 laugh while they
1:20 [Music]
1:22 can let them
1:25 spin let them scatter in the Wind
1:32 the movies I've
1:39 seen on
1:42 [Music]
1:57 them you done you just barking at out
2:00 you're just done just whoop done that's
2:10 it you get discouraged don't you
2:15 girl it's your brother's world for a
2:18 while
2:23 longer we got to dance with the devil on
2:26 a river to beat the
2:29 Stream
2:33 call a living a
2:35 dream call a kicking a
2:39 ladder they come to kick d in your face
2:43 dress you down and tell you that your
2:45 place is in a middle when they hate the
2:48 way you
2:50 [Music]
2:53 shine now that was the wrong that was
2:56 the wrong lyrics but you're
2:57 [Music]
2:59 singing
3:02 [Music]
3:10 they let them
3:13 spin let them
3:15 [Music]
3:18 scatter I've been to the movies I've
3:21 seen
3:24 how
3:25 [Music]
3:29 jok
3:32 [Music]
3:46 he's pissed he's pissed he's not happy
3:48 now he's like I'm not singing
3:53 [Music]
3:59 anymore
4:08 [Music]
4:17 oh good evening good people what the
4:19 hell day is today Wednesday oh right I
4:21 had a board meeting board meeting ouch
4:23 board meeting went fine showed them some
4:25 of our new AI stuff they were like well
4:27 that's that's
4:28 fancy that's very fancy what you've
4:31 created
4:33 [Music]
4:54 there yeah this is a good comment shirt
4:57 isn't it yeah sometimes I do better than
5:00 others I try to do good every
5:03 [Music]
5:14 [Applause]
5:17 [Music]
5:20 night that shirt looks kind of square
5:24 thanks amethyst Jen you mean it's got
5:26 squares or like I'm
5:28 boring is the hair ha's the hair boring
5:31 yeah hair's a little boring what are you
5:32 going to
5:34 [Music]
5:59 do
6:01 so do me a favor tell all your
6:04 friends tell all your friends you're
6:05 like Hey listen listen I just canceled
6:08 by Netflix account I'm watching this
6:10 fascinating Tik Tok Channel I watch it
6:12 two hours a night three hours a night
6:14 night seven nights a week yeah you
6:17 should come watch it come watch it now
6:20 what's it about uh puppies it's about
6:22 puppies in ai ai yeah a little bit just
6:26 a little bit of a I hate II you should
6:29 come watch it
6:30 [Music]
6:41 I love the acoustic guitar I have one
6:43 but need to practice more you know what
6:46 helped
6:47 me well this this is I don't know if you
6:50 have ADHD but if you have ADHD this can
6:53 help one of the one of my symptoms of
6:56 ADHD is sometimes I'll get a song stuck
6:58 in my head for [ __ ] in
7:03 months uh especially if it's a song I
7:05 like like I did it with avit Brothers uh
7:08 uh I and you and me or whatever the hell
7:10 that one's
7:12 called I did it with a bunch of Martin
7:14 ston songs I did it with some John
7:15 hayatt songs I did it with some Brandy
7:17 Carlile songs I did it so what helped me
7:20 get better with
7:23 [Music]
7:27 guitar is I found someone
7:30 that I really liked this playing and it
7:32 was
7:32 really significantly
7:36 beyond my skill and then I picked a song
7:40 of his that was also significantly
7:42 beyond my
7:43 skill um and so the the artist for me
7:47 was Martin
7:48 ston and so I just played the same song
7:54 for like I don't know 3
7:58 months and just sucked a little bit less
8:01 each
8:02 [Music]
8:22 time sitting in this lonely
8:25 Town wonder when things are going to
8:28 change
8:32 dream my life
8:34 away seems these dreams have turned to
8:37 Bunch D
8:40 clouds getting my nerve up but my past
8:44 is pulling me
8:48 down wondering how
8:51 long this black sheep going to stick
8:58 around somebody told me once before you
9:03 can never go home
9:05 again won't you
9:07 leave say things to steer me away yeah
9:13 from the truth of who I am and what I
9:16 believe so I thanked him for his two
9:19 cents with a
9:20 handshake and some sympathy yeah I
9:24 packed up my blue
9:27 jeans headed for this big
9:31 price of my
9:35 freedom
9:36 byebye Black Sheep to the black sheep of
9:41 the
9:43 family
9:45 [Music]
9:47 byebye oh that means so very much to
9:51 me
9:53 by black sheet to my friends and my
9:58 family
10:03 by going off and set my
10:07 soul set it
10:12 free
10:13 free times they were
10:17 changing did a little bit of
10:22 rearranging take a couple
10:25 chances my progress I know advances is
10:30 to the
10:32 time of my
10:35 freedom
10:37 y
10:39 [Music]
10:44 y
10:46 [Music]
10:53 y and then Martin Sexton goes on from
10:55 there and [ __ ] slays it I can't sing
10:57 what he does next but like that was one
11:00 that took me months to get the guitar
11:04 the singing the like like anywhere close
11:07 to not
11:09 embarrassing do you know that do you
11:11 know the Ira Glass if if you've got
11:14 taste thing do you all know that one
11:17 that might be a nice way to start off
11:20 tonight let's see if I can find
11:26 it I think this one of the most
11:28 beautiful
11:32 the most beautiful articulations of why
11:35 art is hard and why it's
11:57 important it's called irlas the creative
12:06 process let's see that's 155 that's 2011
12:09 I think this the whole respecting the
12:11 makers comes back in reversal thriving
12:14 and growing Together full
12:18 circle nobody uh tells people who are
12:21 beginners and I really wish somebody had
12:23 told this to me is that um all of us who
12:25 do creative work like you know we get
12:27 into it and we get into it because we
12:28 have good taste but it's like there's a
12:30 gap that for the first couple years that
12:32 you're making stuff what you're making
12:34 isn't so good okay it's not that great
12:37 it's it's trying to be good it has
12:38 ambition to good but it's not quite that
12:39 good but your taste the thing that got
12:42 you into the game you're your taste is
12:43 still killer and your taste is good
12:45 enough that you can tell that what
12:46 you're making is kind of a
12:48 disappointment to you you know what I
12:49 mean a lot of people never get past that
12:51 phase a lot of people at that point they
12:53 quit and the thing I I would just like
12:55 say to you with all my heart is that
12:58 most everybody I know know who does
13:01 interesting creative work they went
13:02 through a phase of years where they had
13:05 really good taste they could tell what
13:06 they were making wasn't as good as they
13:08 wanted it to be they knew it felt short
13:10 it didn't have this special thing that
13:12 we wanted it to have and the thing I
13:14 would say to you is everybody goes
13:15 through that and for you to go through
13:17 it if you're going through right now if
13:18 you're just getting out of that phase
13:20 you got to know it's totally normal and
13:22 the most important possible thing you
13:24 could do is do a lot of work do a huge
13:26 volume of work put yourself on a
13:28 deadline so that every week every month
13:30 you know you're going to finish one
13:31 story because it's only by actually
13:33 going through a volume of work that
13:35 you're actually going to C catch up and
13:37 close that Gap and your the work you're
13:39 making will be as good as your Ambitions
13:42 in my case like I I took longer to
13:45 figure out how to do this than anybody
13:48 I've ever
13:49 met it takes a while it's going to take
13:52 you a while it's normal to take a while
13:54 and you just have to fight your way
13:56 through that
13:58 okay
14:00 that
14:01 beautiful a lot of people never get past
14:03 that phase a lot of people at that point
14:04 they
14:06 quit
14:08 that's that's the [ __ ] tragedy right
14:11 there
14:17 right that your work isn't shitty
14:19 because your work is shitty your work is
14:23 shitty because you have taste that's so
14:26 [ __ ] profound isn't
14:28 it
14:30 that you're just trying
14:34 to express yourself
14:38 into your
14:43 taste and so many people quit before
14:46 they get
14:50 there and that's [ __ ]
14:56 [Music]
14:58 tragedy
15:00 I think it's I think if there's a single
15:02 thing a single thing on this planet that
15:06 I'm most passionate about it's
15:11 this that it's working with people who
15:14 have taste that it's working with
15:17 people who are trying to live up to
15:19 their
15:25 taste it's working with people that are
15:28 willing to put in the work
15:33 only going through a huge volume volume
15:36 of work do you ever have the chance of
15:39 catching up to your
15:46 taste I just love that idea it's not
15:48 that you're not talented it's not that
15:50 you're not skilled enough it's that you
15:53 just haven't caught up to your taste
15:55 isn't that [ __ ]
15:58 brilliant
16:01 having high ideals and perfectionism
16:03 prevents me from
16:06 creating right that goes back to
16:12 this where is
16:16 it where is
16:19 it is kind of a disappointment to you
16:21 you know what I mean a lot of people
16:23 never get past that phase a lot of
16:25 people never get past that
16:28 phase so the trick
16:32 is
16:35 celebrate that you're not there yet
16:38 because it means you have taste and your
16:42 job is to just keep doing the work until
16:44 you get there here's the other thing
16:47 here's the other thing what's what's
16:48 your name Ridge oh wait where's my
16:50 where's my hang on people I need I need
16:54 my black piece of paper I can't read the
16:56 lettuce hang on we got Construction
16:59 paper for
17:02 this $20 $20 million Production Studio
17:06 and we're going with black construction
17:07 paper to read the the screen yes so
17:10 what's your name Ron
17:12 Diggity
17:15 um having high ideals and Perfection
17:18 perfectionism prevents me from
17:20 creating celebrate that Gap cuz the Gap
17:24 means that you have the
17:26 taste you're already there you're
17:29 already an
17:31 artist now it's just about doing the
17:33 work here's the thing I'm most excited
17:36 about with generative AI one of the
17:38 reasons I
17:39 dump hours a night onto this Channel and
17:44 and many more during the week into the
17:45 AI salon and other
17:51 endeavors is
17:57 that we're living in a Time in
18:02 history I I recorded an an Murphy and
18:06 and uh Cindy Co and I and a couple other
18:08 people
18:09 Suzanne and two other folks um are
18:13 creating a video series about gpts and
18:16 the first chunk that we recorded today
18:18 was just the five of us
18:20 talking and it was it it it was just
18:24 amazing like like realizing
18:28 that
18:30 we're living in a
18:31 time where people who uh three M storage
18:36 management love that I need to watch
18:38 wait oh crap I just lost it I I need to
18:40 watch that video again just put it put
18:41 it back up there again for
18:46 Jim by the way if you don't know Jim
18:48 Ross thre mile management you should get
18:52 to know him he's an amazing
18:56 guy I don't know what that what that
18:59 comment was if you can repin repin it if
19:01 you can find it oh wait here it is I
19:03 need to watch that video again on the
19:04 hard days yeah yeah exactly let's watch
19:07 it again I I
19:09 just Ira Glass has such an Elegance in
19:12 how he talks and how he articulates this
19:14 idea and it's what I love about this
19:17 quote I I assume he's not reading this
19:19 like this sounds like he's in an
19:21 interview right and he's just talking
19:22 this you can tell that this is an idea
19:25 that he
19:27 has processed
19:30 10,000 times you know it's there's
19:34 something about the Elegance of How It's
19:36 articulated just as he's talking
19:39 it nobody uh tells people who are
19:42 beginners and I really somebody had told
19:44 this to me is that um all of us who do
19:47 creative work like you know we get into
19:48 it and we get into it because we have
19:50 good taste but it's like there's a gap
19:52 that for the first couple years that
19:53 you're making stuff creative work like
19:55 you know we get into it and we get into
19:57 it because we have good taste I love
19:59 that to get to do creative work we get
20:02 into it because we have good taste I
20:04 want to do that I could do that right
20:06 you aspire to
20:10 something but it's like there's a gap
20:13 that for the first couple years that
20:14 you're making stuff what you're making
20:16 isn't so good it's not that great it's
20:19 trying to be good it has ambition to
20:20 good but it's not quite that good but
20:22 your taste the thing that got you into
20:24 the game you're your taste is still
20:26 killer your your taste is still killer
20:30 the thing that got you into the game the
20:32 thing that made you say I want to do
20:35 this I can do that I want to write a
20:38 screenplay I want to write a novel I
20:40 want to write a poem I want to make a
20:42 GPT striving is living in action
20:45 absolutely lunatick again you your taste
20:48 is still killer oh by the way lunatick
20:50 when we when we were recording our
20:54 um our little conversation today one of
20:58 the things we were talking talking about
20:59 us talking about people who have
21:02 transformed themselves with with you
21:04 know just AI in some way and I was
21:07 telling your story and I started balling
21:08 like a baby so thank you for
21:12 that but it'll be recorded so you'll be
21:14 able to watch it over and over again um
21:17 I was telling the story about how you
21:20 know you were creatively stuck and then
21:22 you made these gpts and got really
21:24 excited and then you told me at one
21:26 point you went and got your index cards
21:30 and you were making you were basically
21:31 just pulling out old index cards and
21:33 that got me choked up and then I told
21:36 the story about when when um you created
21:39 the Alexander Hamilton GPT and you were
21:41 feeling hopeless about the future of our
21:42 company or country and you had a
21:45 conversation with Alexander
21:47 Hamilton and it restored faith for you
21:51 in our democracy and it just in telling
21:54 it I just [ __ ] broke down I'm getting
21:56 choked up now it's amazing
22:00 and your taste is good enough that you
22:01 can tell that what you're making is kind
22:03 of a disappointment to you you know what
22:05 I mean a lot of people never get past
22:07 that phase a lot of people at that point
22:08 they quit don't [ __ ]
22:13 quit and the thing I would just like say
22:15 to you with all my heart is that most
22:18 everybody I know who does interesting
22:21 creative work they went through a phase
22:23 of years where they had really good
22:25 taste they could tell what they were
22:27 making wasn't as good as they wanted it
22:28 to be they knew it felt short it didn't
22:30 have this special thing that we wanted
22:32 it to have and the thing I would say to
22:34 you is everybody goes through that and
22:36 for you to go through it if you're going
22:37 through it right now if you're just
22:39 getting out of that phase you got to
22:40 know it's totally normal and the most
22:42 important possible thing you could do is
22:45 do a lot of work do a huge volume of
22:47 work put yourself on a deadline so that
22:48 every week or every month you know
22:50 you're going to finish one story because
22:52 it's only by actually going through a
22:54 volume of work that you're actually
22:55 going to C catch up and close that Gap
22:58 and your the work you're making will be
23:00 as good as your Ambitions in my case
23:03 like I I took longer to figure out how
23:07 to do this than anybody I've ever
23:09 met it takes a while it's going to take
23:12 you a while it's normal to take a while
23:14 and you just have to fight your way
23:16 through that you just have to [ __ ]
23:18 fight your way through that that's it
23:23 man
23:27 okay yeah
23:31 that's a good
23:34 one and here's the good news we we are
23:37 living at a point in history where these
23:40 tools these AI
23:44 tools what if I told you
23:46 that to improve your videos
23:49 fast these tools are leveling people
23:53 up so one of the other things we talked
23:55 about in the in the conversation today
23:59 is
24:06 that what becomes possible for you so
24:10 like back to the irog glass taste
24:12 thing what becomes possible for you when
24:15 when in your soul you're like I know I'm
24:18 really good at writing and I'm sort of
24:20 good at
24:21 design and I don't really feel confident
24:23 about much
24:25 else what becomes possible for you
24:29 when now
24:31 you're brilliantly good at writing
24:34 brilliant at
24:36 design really good at illustration
24:38 really good at making music really good
24:40 at making video really good at project
24:43 management really good at
24:45 coding what happens when all of a sudden
24:47 you go
24:48 from a capability and a half where you
24:51 live up to your
24:53 taste to 10 of
24:56 them or 50 of them
25:00 that's the era we're
25:05 entering that's [ __ ]
25:11 insane the people that show up here
25:14 night after night they've nicknamed
25:16 themselves the Irregulars I just think
25:18 they're weird but they claim they're not
25:21 they claim they're not
25:22 weird they just say we're just we're
25:24 just
25:27 irregular
25:30 I'm watching
25:31 them in real time just [ __ ] go from
25:36 like I don't think I can
25:39 do to holy [ __ ] I just did a thing
25:42 that's kind of cool to oh this is really
25:45 amazing to oh this is cool so now I do
25:47 this and now I do that two you know what
25:49 I'm leading a seminar next week to oh I
25:51 just got promoted at work and now I'm
25:52 doing this and I'm I'm like watching
25:54 their energy thank you very much guys
25:57 I'm watching their energy go go
25:59 from kind of
26:02 retracted scared what do I do I don't
26:04 know
26:06 anything to just
26:09 blossoming and becoming contributing
26:11 members of the AI Salon community and
26:17 becoming
26:19 self-expressed humans on all sorts of
26:23 levels and all sorts of channels right
26:27 some of them are making music musical
26:30 with sunna some of them are just getting
26:32 killer
26:33 with images and and killer with images
26:36 is really interesting some are going
26:38 really artistic others are going like
26:41 funny others are going like you know
26:44 commentar and
26:46 satire some are going video thank you
26:49 Empower I appreciate that a lot that
26:51 means a lot oh thank you very much thank
26:53 you Robert Russy it's very
26:57 generous and they're just [ __ ]
26:59 Reinventing
27:01 themselves and like to a person that I
27:03 talk to that's that's that's going
27:06 through it's kind of like the Ira Glass
27:08 thing it's like with this AI [ __ ] it's
27:09 like just [ __ ] play with it
27:14 enough until it starts doing [ __ ] that
27:17 lives up to your taste because what's
27:19 happening with AI is it's getting better
27:22 while you're learning it so like you're
27:25 learning it you're like I'm not very
27:27 good at prompting and my image is to
27:32 suck and I know that's how you sound
27:34 when you're on your
27:36 [Music]
27:37 own it annoys the people around you I
27:40 can promise you that so you should just
27:41 stop that
27:44 [ __ ] and then but you're getting better
27:46 and you're getting better and you're
27:47 making images like maybe that's okay and
27:49 then all of a sudden someday just Sam
27:52 Alman goes hey there's a new thing
27:54 you're like oh okay and you put in the
27:56 same [ __ ] you used to do and all of a
27:58 sudden you're like
27:59 holy
28:02 [ __ ] so so so the irid glass thing wait
28:07 I need I need props
28:10 people I guess I'll use Legos Lego
28:14 pens the Ira Glass thing says that's
28:18 exactly how I
28:20 sound should leave me alone I'm I'm
28:23 trying to learn my prompts over here
28:27 just right
28:29 so you're you're you're you're
28:30 struggling you enter the AI learning lab
28:32 or you enter you know trying to figure
28:34 this [ __ ]
28:35 out and you're like ah sure would be
28:37 good if I didn't suck at writing you
28:39 know you're up here's Hemingway right
28:42 you're like okay I want to be like
28:43 Hemingway and you're like I can't form a
28:47 sentence and you you're playing around
28:49 with these tools and you sort of move
28:51 from green to light
28:53 green and then what's happening now like
28:56 in the olden timey days the move from
28:57 green to like green was like 6 months
29:01 and then the move to Yellow was like
29:02 another year and then the move to Orange
29:04 was like 3
29:06 years and you never [ __ ] get the red
29:09 and then you know there's Hemingway
29:10 right and it and it's like you still
29:12 Aspire for this [ __ ] but you're like if
29:15 I can be on the border between orange
29:17 and red okay I I'll I I'll be willing to
29:19 let someone read my work right but now
29:22 we're in AI world and you start out here
29:27 and AI sort of instantly takes you to
29:31 yellow and then as the tools get better
29:34 you're like blasting into orange
29:36 blasting into
29:40 red and what's fascinating for me is
29:44 that we've all got whatever skill
29:46 whatever is the thing that kind of
29:48 Lights us up that we
29:50 do and and there's a handful of people
29:53 on the planet PJ lockran who spoke at
29:55 the AI Salon two weeks ago um he's one
29:59 of these he's he's a guy he's like an
30:00 incredible musician an incredible
30:02 business person an incredible
30:03 illustrator just he's like an incredible
30:07 entrepreneur he's just he's
30:09 like if he tries it he's going to kick
30:11 ass at it and then there's the rest of
30:14 us right so we've got one or two things
30:16 we kick ass
30:18 at and maybe if you're like
30:22 Renaissance you kickass at three or four
30:25 but most people it's one or
30:27 two all of a sudden with this AI
30:31 [ __ ] you're kicking ass at stuff that
30:35 never in your wildest dreams would you
30:38 be any good
30:39 at like for me it's coding and and
30:43 making make not making pictures but like
30:45 illustrating like art making art like
30:48 I'm I'm a good photographer I've got a
30:49 good eye got good taste but I've got no
30:52 skills like
30:55 this and all of a sudden I can have an
30:59 idea for an image in my head and bang
31:01 it's
31:03 there and one of the things that I'm
31:05 noticing in the in the AI salon and I'm
31:09 trying to I'm trying to Tamp out as many
31:11 of these fires as I
31:14 can is
31:17 that people including
31:21 myself feel
31:25 guilty that the the speed with which
31:28 they're they're doing good work is too
31:32 fast and so they apologize for
31:35 it and I'm [ __ ] done with that I'm
31:39 [ __ ] done with
31:45 it I think it's a Remnant from the
31:48 Industrial
31:49 Age work hard work hard for the factory
31:52 you got to work hard hard work is where
31:54 it's at hard work is where it's at you
31:55 got to work hard okay I'll work hard
32:01 and up
32:04 until few years
32:07 ago a lot of the creative work and a lot
32:10 of the knowledge work like research and
32:13 writing and thought
32:16 leadership required
32:19 time because processing language in our
32:22 brains and getting it out of our
32:25 [ __ ] large language models
32:29 and into a form that didn't suck because
32:32 we have taste took
32:38 time and now what we can do is still
32:41 have that same
32:44 intention but instead of it taking you a
32:47 week to get to the first draft you can
32:49 get to it in five minutes that's with
32:52 some back and
32:54 forth and within an hour you can have
32:58 that thing refined to the point that
32:59 it's ready for
33:01 editing so the thing that used to take
33:03 you a week to get to the point that you
33:06 could do a second pass on it you're not
33:09 having an
33:11 hour and you can spend two or three
33:14 hours of that day taking that thing from
33:17 decent to really [ __ ] good and maybe
33:21 you use AI for that and maybe you
33:23 don't and you can do that for [ __ ]
33:26 you're not even good at
33:32 my exper what my experience has
33:42 been depends on your relationship with
33:45 your self-esteem if your relationship
33:47 with your
33:49 self-esteem tends toward the
33:52 shameful I used to joke on this channel
33:54 about shame is my love language I've
33:56 stopped that it was a joke and it was
33:59 mostly true right good comedy comes out
34:02 of
34:07 Truth and what I've noticed is
34:11 that
34:14 um AI is providing a capability boost
34:19 but it's also providing a confidence
34:22 boost because I can get to the ideas so
34:26 much faster and I can have more of
34:28 them and chat GPT is also like a golden
34:31 retriever it's like gee Kyle that was a
34:33 swell idea can you give me another let
34:35 me make it better for you oh yeah you
34:37 want a picture too here you go there's
34:40 there's like a there's like a there's
34:42 just like a there's not a lot of
34:46 room to go oh I don't have talent I got
34:49 writers block I can't do that I can't
34:51 come up with any ideas it's like I wish
34:53 I had an idea for a dog jumping over a
34:56 log oh [ __ ] there's 20 ideas oh oh that
35:00 one's not bad oh oh I kind of like that
35:03 oh well I guess I am pretty smart after
35:08 all and it's like like that happens
35:11 daily I wrote an outline in a and a in a
35:15 10-minute in a couple of minutes tweaked
35:18 it in a few more minutes probably had
35:21 enough to do the recording in about 15
35:24 minutes yeah yeah like oh I've got an
35:28 idea to make a video oh dear CH jbt dear
35:32 dear chbt very good to see you again
35:35 it's fantas you look good you look did
35:37 you lose weight anyway uh there's no
35:40 shaming in here okay anyway all right um
35:44 I would like an outline I'm thinking
35:46 about doing a video but you know I don't
35:49 have a lot of talent I don't have a lot
35:50 of time and here oh
35:53 my that's pretty good right and then few
35:57 minutes later oh that's really good and
35:59 then all of a
36:00 sudden you go from like the thing that
36:03 would have been your trap in the old in
36:06 the olden timey days you know like
36:09 2021 the thing that would have been your
36:11 trap which is if I want to get this
36:13 thing to 80% it's going to take me a
36:15 week and I've got to battle my shame
36:17 demons for the entire [ __ ] week to
36:19 get to that first draft how many first
36:22 drafts never survived the first
36:26 day many
36:31 I'm experiencing that less and less and
36:34 less and
36:38 less it's it's almost like it's almost
36:41 like before you could use oh it's too
36:45 hard as the excuse not to do the work
36:47 because what you were really feeling was
36:49 I'm not good enough and it won't live up
36:51 to my taste so I'm not going to do it
36:53 but it's hard and that's why I'm not
36:55 doing it it's just I'm too tired and
36:57 it's too
36:59 hard and now you don't have that excuse
37:02 anymore so now you're like oh [ __ ] well
37:05 it's not too hard so I can't justify it
37:08 that way oh [ __ ] am I just scared is
37:11 this just fear is this fear of what
37:13 other people are going to think of me oh
37:15 [ __ ] oh that's the sweat okay I'll wait
37:18 for a day I'm just going to sit back you
37:20 don't have the excuses anymore that you
37:22 can justify not doing
37:24 it cuz you can [ __ ] knock that [ __ ]
37:27 out now and and you can knock it
37:29 out you can get stuff to 80% so
37:34 fast and that was the drudgery
37:37 work and then you can do things like
37:40 this you're like well actually Paul
37:43 rater talks about this he says he
37:44 doesn't write to produce things he
37:47 writes to understand the thing he's
37:50 writing about right for him it's a way
37:52 of mentally processing the
37:56 thing okay so write your 80% draft like
38:00 that to process the thing and then use a
38:03 i to tweak it and you know edit it and
38:06 give you ideas and reorganize
38:11 it and then it's just really [ __ ]
38:16 good anyway hi my name's Kyle
38:20 Shannon welcome welcome to the AI
38:23 learning
38:25 lab um we talk about stuff here AI
38:30 life I have ADHD and AI helps keep the
38:34 hyperfocus and dopamine going without
38:36 the speed bumps totally Jane do Jane
38:41 qo
38:43 yeah AI is like a video game a AI is
38:46 like a creativity video game I have an
38:50 idea here it is
38:53 materialized I'd like to make that a
38:55 picture here's your picture
38:59 Vicky Baptist in here does 3D printing
39:02 as as a side business or as a business
39:04 business I think it's a side business
39:06 she's got all sorts of 3D printers and
39:08 stuff so she can like come up with
39:11 ideas visualize them like that and then
39:15 send them off to the 3D print
39:17 farm and start materializing
39:24 stuff holy
39:26 [ __ ]
39:29 so if you want to learn how to do that
39:31 stuff that's what we do here that's what
39:32 I talk about so if you have questions
39:34 about
39:35 AI pop it down Three Mile Storage
39:37 management AI is Flushing out the pro
39:39 procrastination and self-doubts for me
39:41 yep that's it for me
39:43 too it's you know it's funny it's funny
39:45 Jim
39:48 I I'm kind of
39:50 consciously I'm kind of consciously
39:53 procrastinating now meaning meaning like
39:57 it used to be I'd go oh it's too hard I
39:59 don't have the energy this that I'd have
40:01 all these excuses and now like I know
40:02 that's not reality anymore that it's not
40:05 going to take me that long like it's
40:06 literally going to take me 15 minutes
40:09 right for whatever the thing is and I'll
40:12 just go oh [ __ ] it I'll get to it like
40:15 you know I'll I'll I'll keep feeding my
40:17 ADHD brain where I got to have the
40:19 dopamine hit so I'll push it to the last
40:20 minute just to do it but I'm like I'm
40:22 much more conscious of it now it's much
40:24 more it's almost like I have an inside
40:26 joke with myself
40:28 like yeah you're going to be able to do
40:29 it in a few minutes doesn't [ __ ]
40:30 matter I'm loving this thank you you'll
40:33 help me reignite my AI passion Ron
40:35 Diggity yeah please do please do because
40:38 here's the thing with with the AI stuff
40:41 and and AI
40:46 passion one of the really lucky things
40:49 about being AI literate right
40:55 now is that there are essentially zero Z
41:01 Stakes if you decide to do something
41:03 with AI and you make the
41:06 worst the worst possible output like
41:09 like in the history of humanity if you
41:11 create the worst AI output in the
41:13 history of humanity and you show it to
41:17 people and you go I did this with AI and
41:20 I go oh
41:22 okay it doesn't matter because 3 months
41:26 from now the tools are going to be 10
41:27 times better and 3 months after that
41:29 they're going to be 50 times better 3
41:31 months after that we like our eyes are
41:33 just going to we're going to wake up in
41:35 the morning our eyes will start watering
41:36 tears as just [ __ ]
41:40 happens so so we have a we have a window
41:43 right now where all you have to do is
41:46 play just play don't put any pressure on
41:50 it's got to be good it's got to live up
41:52 to your expectations or anyone else's
41:55 expectations most people don't know that
41:57 this [ __ ]
41:58 exists and that's not a joke McKenzie
42:02 the McKenzie report that came out in
42:03 November 82% of Americans have not used
42:07 chat GPT
42:10 82% so that means 18%
42:16 have and that probably means that a huge
42:19 percentage of the
42:20 18% probably opened it once and
42:24 went you know describe a Dalmatian puppy
42:27 for me and it described it and then they
42:30 went it's stupid I don't like it and
42:31 I've never used it
42:33 since so let's call it half half of the
42:37 18% said
42:39 that have have used chat GPT once they
42:42 they still don't know what it is but
42:44 they said they used it so now we're down
42:45 to
42:47 9% how many of that 9% are paying the 20
42:51 bucks a month for chat gp4 I think it's
42:53 low even in this channel even in the AI
42:56 salon I am constantly surprised by
42:59 someone who I assumed had paid the 20
43:03 bucks for chat gp4 is like I think I'm
43:05 going to get chat gp4 now I'm like what
43:07 the [ __ ] are you waiting
43:12 for so so that so we're probably down to
43:15 5% have chat GPT
43:19 4 and then how many software
43:21 subscriptions do you have where you
43:23 bought the subscription and then never
43:25 really got around to using it a lot
43:28 right so let's call that half again so 2
43:31 and a
43:34 qu% of people in
43:38 America have any clue any
43:43 clue how powerful these AI tools
43:49 are 90 let's call it 95 95% of people
43:54 don't have a [ __ ] clue
43:56 so total freedom for us that are playing
44:00 right now to just [ __ ] around just [ __ ]
44:02 around and find out does it work is it
44:04 good is it as brilliant as I think oh
44:06 I'm going to go make a feature film oh
44:08 [ __ ] I can only make 4 second clips and
44:11 they're completely morphe and clamation
44:14 and weird I can't this is oh my God it's
44:16 all well maybe I'll try I right like you
44:19 can't [ __ ] up right
44:21 now if you make something remotely
44:24 interesting with video right now with
44:27 with Runway and
44:29 Pika people are like oh my God it's
44:32 glorious if you look at it objectively
44:35 compared to what filmmakers do you're
44:39 like well it's I mean it's it's okay for
44:43 AI but it
44:45 ain't you know it ain't 16 mm film with
44:49 a decent
44:52 editor but then open AI announces
44:56 Sora and all of of a sudden if you made
44:58 that Runway film and you know how to
45:01 make something decent out of something
45:04 crappy what are you going to do when the
45:06 Tool comes out where you can do
45:08 60-second
45:09 clips that are hard to distinguish from
45:14 reality well you're going to [ __ ]
45:16 kill it so so here's the 16 mm short
45:19 film here's your Runway film right
45:21 you're like yeah it's it's kind of janky
45:24 and embarrassing but maybe I'll show
45:25 some people and all all of a sudden
45:27 you're going to have tools that are
45:28 going to allow you to sort of execute up
45:31 here
45:33 overnight so there's no risk right now
45:36 zero
45:39 risk we all you all we have
45:46 permission to do
45:48 anything well I can't use it at work go
45:52 get yourself a $200 Chromebook and use
45:54 it at home on the weekends and nights
45:56 get it on your [ __ ]
45:59 phone start having conversations with pi
46:02 or chat gpt's mobile apps that have
46:05 conversational
46:07 ai go to cop pilot. microsoft.com if
46:10 it's not blocked at your work and start
46:12 making images and start experiencing
46:14 what it's like to interact with a large
46:15 language
46:17 model go to perplexity and see what it's
46:20 like to interact with a model that
46:22 actually does searches on the internet
46:25 and provides citations to the ANW so you
46:27 can go check the
46:31 work go
46:33 too. and say make me a song about being
46:37 frustrated that my boss is an
46:40 [ __ ] in the
46:43 Blues I just want chat GPT to let me
46:46 organize my chats oh my God wind rider
46:48 can you believe we still can't search
46:50 for [ __ ] we can't search our [ __ ]
46:52 chat history in the desktop app I'm like
46:55 what in the what in the [ __ ]
46:59 seriously okay I'll I'll give open AI
47:02 that that creating a folder High
47:05 hierarchy where you can stick your chats
47:07 in folders and save them I I get that
47:10 that might take three weeks but they've
47:13 already got the chat functionality in
47:15 the mobile app just put it on the
47:17 [ __ ] interface call the same
47:22 API I'm right there with
47:25 you which is again for evidence that
47:28 we're so [ __ ] early
47:30 here that you can do
47:35 anything so okay so your taste is up
47:38 here and your AI skills are down here
47:40 and the tools are down here well just
47:42 dick
47:43 around and your skills will little get a
47:45 little bit better and the tools are
47:47 going to [ __ ] make gargantuan
47:52 leaps and then you'll be one of the ones
47:55 that gets to say look at this thing I
47:57 made and people will lose their [ __ ]
48:05 mind I'm really excited next Tuesday we
48:07 have Kelly Bosch coming to speak at the
48:10 AI salon and she makes these
48:14 stunning videos with AI she doesn't
48:18 apologize for it she doesn't say it's AI
48:21 she just has a Tik Tok Channel and she
48:23 makes these stunning
48:25 videos people and animals and insects
48:28 and ballerina dancers that look like
48:31 insects they're they're just insanely
48:34 good Kieran wants to know about Story
48:36 vine okay I can talk about story
48:41 mine and when I reached out to talk to
48:43 her she's like
48:45 oh I'm not really a big deal I'm not
48:48 like I just I just it's not very good
48:53 like yes it
48:56 is well I don't have a lot of skill I
48:58 just like do it on my phone before I go
49:00 to bed
49:03 what that's even more
49:06 remarkable you do that quality of work
49:09 on your phone as you're going to
49:12 sleep
49:17 yeah holy
49:20 [ __ ] imagine what she's going to be able
49:23 to do when the tools get
49:26 good
49:35 we're going to have people winning
49:36 awards for for storytelling and for
49:39 creative and for things like
49:43 that that have created works that that
49:46 win awards that have never been in that
49:49 industry
49:54 before within three years watch you
49:58 watch there's going to be someone
50:00 accepting an Oscar or a special Oscar or
50:02 an
50:03 Emmy or a [ __ ]
50:09 Tony that was like well I used to clean
50:13 gutters and then I I started hanging out
50:15 in the AI learning lab and uh there was
50:18 this fell I don't know what his name was
50:20 he was an older fell anyway he talked
50:22 about AI he's like just go play with AI
50:24 so I started playing with AI and then I
50:26 you know before I knew it I I'd made
50:28 this masterpiece said what you know
50:31 anyway thanks for the award I I don't
50:33 know many of you people but I've seen
50:34 you in the movies it's pretty you're
50:36 pretty cute it's a this is
50:38 uh they you think they got Jack Daniels
50:41 back there
50:42 CU I'm I'm a wicked Tad
50:47 nervous we're gonna see that Awards
50:51 acceptance
50:54 speech I'm a wicked Tad nervous hope
50:57 they got Jack Daniels back
50:59 [Laughter]
51:04 dear
51:06 yes tell them Bob tell them what they've
51:09 won oh you've won the freedom to do
51:11 anything the [ __ ] you want
51:14 congratulations all right let me see
51:15 what you're saying in here I've been
51:17 rambling for a bit what's it been an
51:18 hour and a half an hour 50
51:22 minutes that's a that that's what we
51:24 call in the AI learning lab a 50-minute
51:26 hello hello
51:33 remember when open AI shut down
51:35 subscriptions I do I have my main code
51:39 and three others are close so I find
51:41 myself working all four for now my
51:44 husband use it used it once and said
51:46 Google's better yeah Jane do exactly you
51:49 watch your husband at some point so Jane
51:52 do you keep using chat GPT you keep
51:55 figuring out what the [ __ ] this thing
51:58 is and I guarantee you Mr Google I tried
52:03 chat GPT once Google's better I
52:05 guarantee you at some point he's going
52:07 to peek over your shoulder and go hey
52:09 Jane uh listen
52:12 uh you know you know Bob at work right
52:15 uh listen uh Bob Bob was telling me that
52:20 that the AI could uh it could do uh data
52:24 analysis do it and and you'll be like oh
52:27 yeah that's called code interpreter I do
52:28 that all the time what what do you mean
52:30 oh I use it to create spreadsheets of
52:33 all my seeds for the garden well how do
52:36 how do you do that oh you just ask
52:41 it Godfrey Lake hey Kyle greetings from
52:44 jamaicon what's Happening greetings
52:47 greetings thank you for checking
52:51 in yeah you just keep playing Jane let
52:54 let hubby let H hubby live in live in
52:56 the on the
53:06 sidelines it's
53:08 cuckoo it's cuckoo what you're you're
53:10 going to discover some things that you
53:12 can
53:14 do you'll be able to do [ __ ] better than
53:17 him at some point and you'll be able to
53:18 do stuff that he and you both
53:22 know there's no way in hell you can do
53:25 and you're going to do it and it could
53:27 be next
53:29 week just by just by messing around with
53:31 this stuff and all of a sudden you'll
53:33 stumble onto something like you're
53:34 you're like oh my God I'm really good at
53:36 this huh I wonder if I did this and that
53:40 and I put it together with that tool I
53:42 bet I could hm hey honey look what I did
53:45 here wait how'd you what do you mean you
53:47 did it who who did you work with I I
53:49 didn't work with any I just in the chat
53:52 [Laughter]
53:55 GPT what do you mean you're a published
53:58 author I just well I just I wrote the
54:00 book and then I Illustrated it and then
54:01 I and I edited it and then I and I edit
54:04 I did a pass on the pictures and then
54:05 came up with the names of the captions
54:07 of the pictures and then then and then
54:09 put it in what that chachy BT taught me
54:11 how to use canva because I well I used
54:12 the canva plugin it took me and then I
54:15 and then I didn't know how to publish
54:16 the Kindle but then all of a sudden I
54:18 went I went I went hey uh oh hey chat
54:20 gbt how do you publish the Kindle and it
54:22 told me it took me step by step and so
54:24 now I'm published I'm a published author
54:26 so sorry sorry I know I know you were
54:29 excited about being published at some
54:31 point two decades ago but anyway um
54:34 ain't that cool what do you want for
54:42 dinner husband's face get red steams
54:45 comes out of his ears I'm so happy for
54:48 you
54:51 honey then three weeks later uh Jane
54:54 could you
54:55 uh
54:59 listen how do you how did you publish a
55:03 [Laughter]
55:10 book oh Corey Sandler I have a GPT
55:13 Gardener's Green Thumb it knows
55:15 everything and Designs Gardens and gets
55:17 seed zones yeah awesome awesome Corey
55:20 Sandler if you don't follow Corey
55:22 Sandler Pottery first of all you should
55:25 follow her Tik Tok Channel second of all
55:27 you should go buy her Pottery cuz it's
55:28 beautiful but you can also watch her
55:30 lives where she makes Pottery she's just
55:32 throwing mud
55:33 around and then occasionally you're like
55:36 oh there's Corey but she's cooking she's
55:39 like La singing cooking then sometimes
55:43 there's a guitar and she's singing and
55:44 then
55:45 lately she's like oh I'm a an
55:50 application developer now because of
55:52 gpts I just make
55:54 gpts and she makes [ __ ] kickass
55:58 gpts and not only is she making them and
56:01 putting them out in the
56:02 world she's
56:05 sharing with the world how she did it
56:09 why she did it how it
56:12 works what's cool about it look at all
56:16 these cool pictures I made in this
56:19 GPT I got to this one and it was really
56:21 cool and then I kept going and I made 30
56:24 more she she's sharing her her work
56:28 process and you know what's
56:30 amazing she she said this I think you
56:32 said this on one of your lives the other
56:34 night she's like I don't even know what
56:37 I'm going to do with this art I'm just
56:38 doing
56:40 it that's the IR glass
56:42 thing just [ __ ] do the
56:45 work just do the work she's making
56:48 remarkable art is she going to print it
56:50 out is she going to have a gallery
56:51 showing is it going to turn into a CD
56:52 ROM I don't [ __ ] know she doesn't
56:54 [ __ ] know it doesn't [ __ ] matter
56:59 but she's getting better and better and
57:01 better and better and then the tools are
57:02 getting better and better and better and
57:05 better and at some
57:08 point she's going to make something and
57:11 all of a sudden it's going to be crystal
57:12 clear I know what I'm doing with
57:15 this and it'll turn into an album or
57:17 it'll turn into a feature film or it'll
57:20 turn into some new form of pottery
57:22 that's never existed before on the
57:24 planet because she conceived of a thing
57:26 and she has some pottery skills and she
57:28 has some this skills and she needs some
57:30 sort of ceramic recipe that didn't exist
57:33 before but oh [ __ ] chbt can help me
57:35 formulate that and all of a sudden she's
57:38 got something that's completely new and
57:39 unique and
57:43 just and it might not even be Pottery it
57:46 might be something they use for the next
57:48 space
57:52 mission like that's the weirdness we're
57:54 in like the the the ren s explosion that
57:58 we're entering right now is just it's
58:00 going to get [ __ ] weird people are on
58:02 people are used to like we get on our
58:05 path we go to college we get on our path
58:07 and like gen Z's kind of [ __ ] that up
58:09 they're like I think I'm going to get go
58:11 get laid for a little bit and then maybe
58:13 I'll go to college and I don't like
58:14 working anyway and now that I got Co I
58:16 don't want to go into the okay maybe
58:19 they were right maybe they're right
58:21 maybe it doesn't [ __ ] matter we don't
58:22 need to be on this
58:25 path right maybe it's like [ __ ] around
58:27 [ __ ] around and then here's these tools
58:30 and up Blossom some talents and skills
58:33 and things that you never conceived you
58:35 have any interest in I really like
58:38 that
58:40 H who knew that I been being into or
58:43 going into organic chemistry in my 50s
58:46 huh who
58:51 knew
58:54 CNN ex accountant creates creates a new
58:58 organic chemical that's going to cure
59:00 this
59:01 disease well I was doing my accounting
59:04 business
59:06 and discovered a passion for organic
59:08 chemistry how'd you do that I don't know
59:11 I asked Chad GPT to make a poem about
59:13 organic chemistry and it did I thought
59:15 it sounded kind of cool and then like a
59:17 month later I uh I came up with my first
59:25 compound
59:28 [Laughter]
59:31 oh so Corey you're going to be creating
59:35 space Ceramics I don't know if you know
59:37 that yet but that's on your that's on
59:39 your little trajectory
59:43 there well I made a picture of a dog in
59:47 a ballet dress and
59:51 then that led to this and then and then
59:54 I went into chyp and then all of a
59:56 sudden Google Google came out with this
59:58 new model and I put the thing in the
59:59 thing and all of a sudden out came this
1:00:01 thing and it was a new Ceramics recipe
1:00:02 so I figured I'd fire it up and then it
1:00:04 had all these
1:00:08 properties and that was the weirdest
1:00:10 thing I hadn't told anyone about it and
1:00:11 NASA just called
1:00:13 me uh Miss Sandler yeah we'd like to
1:00:16 talk to you about your Ceramics
1:00:18 Explorations you want to buy some pots
1:00:21 no oh coffee
1:00:25 mugs
1:00:27 I can make one with NASA on the side
1:00:29 Cory doesn't talk like this by the
1:00:34 way this is just the news version of
1:00:36 Corey that is that has that has invented
1:00:40 a new ceramic for the space space
1:00:45 program I don't by the way I don't know
1:00:48 what the [ __ ] I'm talking about tonight
1:00:50 like I why anyone still here is beyond
1:00:53 me but I hope you're
1:00:55 enjoying
1:00:58 oh I just think of neuralink
1:01:00 interpreting our dreams into images and
1:01:02 videos you can you can essentially do
1:01:04 that right now Kat if you go to
1:01:07 um if you've got chat GPT 4 go to
1:01:10 explore gpts I created a GPT on this
1:01:13 live called Dream Weaver so look up
1:01:15 Dream Weaver by Kyle Shannon and it'll
1:01:19 ask you three questions about your dream
1:01:23 and then it'll visualize your dream and
1:01:26 then then it'll give you a yian analysis
1:01:28 of it so you can you don't need
1:01:30 neuralink for that but neuralink will be
1:01:31 cool because you'll just wake up in the
1:01:33 morning and it'll be like here's what
1:01:34 you saw last night would you care to
1:01:37 comment on
1:01:38 [Laughter]
1:01:41 it that's
1:01:44 crazy
1:01:46 oh it's like those modest singers in
1:01:49 American
1:01:51 Idol uh uh so how do you reach out to
1:01:55 these creators LinkedIn
1:01:57 DM who like Kelly Bosch I just reached
1:02:00 out to her on Tik Tok I just dm'd
1:02:04 her I I I don't I don't do it enough I
1:02:07 should reach out to more
1:02:09 creators
1:02:13 um but yeah you know I
1:02:16 mean most people don't get back to me
1:02:19 share your nightmares with friends
1:02:24 exactly hey Sally Sally want to see the
1:02:27 dream I had last night you show her
1:02:28 she's
1:02:32 like Jane I had no idea you were so
1:02:38 dark oh Gemini 1.5 is definitely better
1:02:42 but chat GPT 5 will be better than that
1:02:45 have you played with with Gemini 1.5
1:02:48 Joker Is it is it out widely yet Ethan
1:02:51 mullock is writing about he's writing
1:02:52 like four articles a day on it it sounds
1:02:55 [ __ ] amazing
1:02:57 but I don't think I have access to it
1:03:03 yet let me see I signed up for it um gem
1:03:09 ni
1:03:11 1.5 so you haven't
1:03:15 heard
1:03:16 [Music]
1:03:20 Gemini Gemini is
1:03:24 Google's chaty PT or it's their large
1:03:28 it's their large language model sitting
1:03:29 underneath
1:03:31 it and they they released Gemini
1:03:36 Advanced two weeks ago and it was it was
1:03:40 a flop it was a
1:03:41 disappointment and then shortly after
1:03:43 that they followed it up
1:03:45 with it was called Gemini Advanced based
1:03:48 on the Gemini Ultra
1:03:51 model and then like a week after it they
1:03:53 announced Gemini 1.5 Pro or Gemini Pro
1:03:58 1.5 and Gemini Pro was their middle
1:04:02 model and it sounds [ __ ] remarkable
1:04:06 so I don't I don't know what they're
1:04:08 doing with their like I don't know if
1:04:10 Gemini Pro 1.5 is actually Gemini Ultra
1:04:13 1.5 I don't I don't think it is but it's
1:04:16 got a million token context
1:04:19 window so you can upload an hour of
1:04:22 video to this thing and just say where
1:04:26 where's the where's the point where she
1:04:28 drops something in the sewer
1:04:31 grate and it will take you right to that
1:04:33 spot of the
1:04:34 video or or answer questions about it
1:04:38 where where it is in the video so it can
1:04:41 it can ingest like an hour of video so
1:04:44 it's multimodal
1:04:46 in and then he also did can ingest like
1:04:49 an hour of video so it's multimodal
1:04:52 in and then he also did like a 600
1:04:57 page multi-document
1:05:03 game you know when neuros
1:05:07 spicy OCD people make games and they
1:05:10 make like really hyper complicated games
1:05:12 it's like one of those games like 600
1:05:15 page
1:05:16 downloadable manual right and he
1:05:19 uploaded that to it and it plays the
1:05:22 game perfectly like it just knows all
1:05:24 the rules it pulls the [ __ ] from all the
1:05:25 do
1:05:27 doents and then the other example he
1:05:29 talked about is
1:05:31 um there's an obscure language that 200
1:05:35 people on the planet
1:05:38 speak and they
1:05:40 uploaded I guess whatever information
1:05:42 they had about this language like 200
1:05:44 people speak it so there was no there
1:05:46 was nothing on the internet so there's
1:05:49 nothing in the core the base model that
1:05:52 knew anything about this
1:05:54 language so they uploaded whatever they
1:05:57 uploaded about this language and it just
1:05:59 learned the language and they had a a
1:06:03 one of the 200 people that can speak
1:06:05 this language was like yeah it's
1:06:10 correct like people are playing with
1:06:12 that right
1:06:13 now and that's not Chad gbt that's
1:06:16 Gemini this this is a Google thing so
1:06:19 the capabilities of these tools are
1:06:21 going to you know 2023 they were going
1:06:24 like you know oh that's cool oh that's
1:06:26 cool oh that's cool 2024 is going to be
1:06:28 like nothing nothing nothing boom here's
1:06:30 Sora boom here's Gemini 1.5 boom it's
1:06:33 going to be these big like Donkey Kong
1:06:36 kind of climbing up the building kind of
1:06:40 moves crazy crazy
1:06:45 crazy Gemini to me is like a bull in a
1:06:47 china shop powerful and Majestic but
1:06:49 damn if you like the status quo I don't
1:06:52 quite know what that means my experience
1:06:54 with Gemini at this point is it's
1:06:56 it's so janky it's not usable like it's
1:06:59 just every other prompt just doesn't
1:07:02 work or it just shows you code instead
1:07:04 of an answer or it's like you ask it to
1:07:07 tell to tell you to show you some
1:07:09 YouTube videos on a certain topic and
1:07:12 then you have it do more and then it
1:07:15 says oh I can't search YouTube what do
1:07:17 you why would you I'm just a large
1:07:18 language model why would you ask me
1:07:19 about YouTube I'm like uh cuz you own it
1:07:22 all uh oh yeah Gemma Gemma was released
1:07:25 today finally so so Pate I didn't get a
1:07:28 I had a board meeting today and then I
1:07:29 had I I've been busy I saw Gemma was
1:07:33 released so so Google released Gemma
1:07:36 which is they're they're basically out
1:07:39 meaing meta so they released a bunch of
1:07:41 open- source Foundation models called
1:07:46 Gemma I guess like Gemini but Gemma
1:07:49 Gemma is the free [ __ ] or the I don't
1:07:52 know what the what the license is in it
1:07:55 is it are they commercial commercial
1:07:56 licenses pate and also like what scale
1:07:59 are I think they're smaller and smaller
1:08:01 and more efficient right so the idea is
1:08:03 to be able to run them on small
1:08:06 devices anyway if you don't know Pate
1:08:08 Pate runs the the um AI mechanics Guild
1:08:13 inside of the AI Salon which I'll talk
1:08:16 about yep woo open source I know and I
1:08:20 oh so so commercial commercial licenses
1:08:22 that's great that's great so so now
1:08:25 we've got most of most of the open
1:08:28 source world right now works on meta's
1:08:32 llama 2 models cuz they were big and
1:08:35 they were good and so there's all these
1:08:36 derivative models you know spun off off
1:08:39 off of that but they're all flavors of
1:08:40 the same thing so having someone like
1:08:43 Google release a competitor to that is a
1:08:47 really big deal are they at the same
1:08:49 level is is are the Gemma models are any
1:08:52 of them at the same level as llama 2 70
1:08:55 billion
1:09:06 P it has two oh okay it's it's got two
1:09:09 two billion and S billion win versions
1:09:11 okay so they're smaller right now yeah I
1:09:13 read that they were
1:09:15 smaller all right well there's going to
1:09:18 be cool [ __ ] coming and and the thing
1:09:20 about um the thing about small models
1:09:24 like that the the open source Community
1:09:26 is going to pounce on those and just
1:09:28 kick the [ __ ] out of them and optimize
1:09:30 them and there's there's cool stuff
1:09:33 coming there's also um Apple released
1:09:37 this thing called mlx which makes it
1:09:39 easy to run
1:09:42 um efficiently run large language models
1:09:46 on iPhones or you know mobile
1:09:48 devices so that's pretty
1:09:51 slick my AI learning mut is snoring in
1:09:54 my lap right now hey know what's
1:09:56 happening you put her to sleep Kyle I
1:09:58 know I know I'm not doing much tonight
1:10:00 but you know whatever what are you going
1:10:01 to
1:10:02 do keep in keep in mind one is not even
1:10:06 close to 1.5 yeah yeah that makes sense
1:10:10 Joker Joker did you say that you're
1:10:12 playing with it you have access to
1:10:24 it
1:10:26 [Music]
1:10:30 did you get Gemini Pro for workspace yet
1:10:32 no I saw that they announced it though
1:10:34 that they released it though did you
1:10:35 hear that Pate they Rel they released
1:10:38 Gemini Pro for for uh or Gemini Advanced
1:10:41 for for work
1:10:43 group Oh no just seen test and reviews
1:10:45 yeah yeah same
1:10:47 here yeah I want to play with right now
1:10:50 I'm I'm jonesing to play with Sora and
1:10:52 I'm jonesing to play with Gemini 1.5
1:10:57 yeah the the 1.5 being dramatically
1:10:59 better than one is kind of like like the
1:11:03 the RoR scale where a magnitude 2
1:11:06 earthquake and a magnitude 3 earthquake
1:11:08 they're not linear right the magnitude 3
1:11:11 is or the magnitude 31 is is
1:11:13 dramatically stronger than the two and
1:11:15 you know every every point is on an
1:11:18 exponential curve or whatever
1:11:21 logarithmic something I don't know some
1:11:23 math some math I don't know about
1:11:35 [Music]
1:11:37 just check the paper again they compare
1:11:39 it to mistal 7B and Lama 7B and 13B okay
1:11:44 cool all right
1:11:46 good that'll be
1:11:48 interesting very very
1:11:50 cool I like seeing Google having having
1:11:53 an open- Source Foundation model in the
1:11:55 game
1:11:58 why
1:12:00 because one of the one of the I think at
1:12:03 this point unfounded fears is that all
1:12:06 of the control of AI is going to rest
1:12:08 within two or three big
1:12:11 companies two or three big companies are
1:12:14 definitely going to have the big Iron to
1:12:17 run the really sophisticated models
1:12:21 but there's increasing competitive
1:12:23 pressure to get this [ __ ] out out of
1:12:25 those companies and so having
1:12:30 meta put
1:12:32 their um foundational models into open
1:12:35 source and now having Google do it as
1:12:37 well puts pressure for other people to
1:12:39 do that and and hopefully we'll have two
1:12:42 or three major companies putting out
1:12:44 significantly powerful models that the
1:12:46 open source Community can build really
1:12:48 sophisticated [ __ ] on now the big tech
1:12:52 companies are likely going to keep the
1:12:54 the sexiest highest end stuff um but
1:12:57 there's a point at which you know how
1:12:59 you remember when computers were not
1:13:02 powerful enough and every time a new
1:13:04 model came out everyone would flock to
1:13:06 buy it because the the perceived
1:13:09 performance jump was so
1:13:11 dramatic that's what we're going to be
1:13:13 doing with with generative AI right
1:13:16 where each next model is going to be so
1:13:18 much better that's like we got to have
1:13:19 the new one we got to have the new one
1:13:21 we got to have the new one a lot of sort
1:13:24 of so so if this this is
1:13:27 gp5 I guarantee you at some point gp4
1:13:30 gets open sourced and put out there or
1:13:31 some you know some gp4 equivalent gets
1:13:34 open sourced and and but we'll we'll go
1:13:37 to open AI or Google or Apple for the
1:13:41 gp5 level [ __ ] and then we'll come out
1:13:43 with GPT 6 and then this [ __ ] will get
1:13:45 open source there's going to be a point
1:13:47 at which the open source stuff is so
1:13:50 powerful that it's just incrementally
1:13:53 more capabilities or speed or something
1:13:56 like
1:13:57 that so but I but but what this says to
1:14:00 me is this AI [ __ ] ain't going away it's
1:14:03 not going away if you're sitting on the
1:14:05 sidelines waiting for this stuff to go
1:14:06 away it ain't going away ain't going
1:14:14 away do I lie to
1:14:17 you I might lie to
1:14:22 you it ain't going away and it's not
1:14:24 going to be it's not going to be
1:14:25 completely controlled by a handful of
1:14:30 companies and you're like well
1:14:32 government might shut it down all right
1:14:34 which government
1:14:40 where cuz the US ain't going to shut
1:14:44 something down if the rest of the
1:14:47 world is creating a competitive
1:14:51 Advantage because of the
1:14:53 stuff the us will go well we'll just
1:14:55 justify it in the following ways we'll
1:14:57 pass these laws and these Scoundrels
1:14:59 can't use it but you know you the rest
1:15:01 of you
1:15:05 can
1:15:06 oh my neighbors speak zoto and I can't
1:15:10 find a translator oh crap I just lost
1:15:12 that where was it wait
1:15:15 ah I just lost it can anyone find that
1:15:19 my neighbor oh here it is Jane Doe I'm
1:15:22 doing that tomorrow hang on wait Jane
1:15:24 Doe let's
1:15:26 go I'm going to okay so I'm on my I'm on
1:15:30 my
1:15:33 iPad if you don't know this for all of
1:15:37 you um if you download the official chat
1:15:41 GPT app you got to be really careful the
1:15:43 official chat GPT app is from open Ai
1:15:47 and it doesn't cost anything okay
1:15:51 okay um so if you get anything that does
1:15:54 cost
1:15:56 you be careful for the app itself now if
1:15:59 you want to upgrade to GPT plus that's
1:16:01 20 bucks a month so that costs something
1:16:03 but the the the app itself to be able to
1:16:05 use GPT 3.5 is free and when you run the
1:16:11 IOS app it's got this I think the
1:16:13 Android app as well it's got these
1:16:14 little headphones here and if you talk
1:16:17 tap on those little headphones I'm going
1:16:18 to put it in gp4 mode cuz it's just
1:16:24 better
1:16:28 do you know the language
1:16:33 zoto I'm not familiar with a language
1:16:35 called zoto it's possible it could be a
1:16:38 very obscure or specific dialect or
1:16:40 perhaps even a constructed language not
1:16:42 widely recognized can you provide more
1:16:45 context or details about it that might
1:16:48 help me understand better or find
1:16:50 relevant
1:16:51 information well Jane Do's neighbors
1:16:54 speak zoto so I assume someone knew
1:17:03 it it seems like you're referring to a
1:17:05 linguistic or cultural context that
1:17:09 isn't widely documented in the major
1:17:11 sources I have access to if zodo is
1:17:13 spoken here let me try let me try Pi Pi
1:17:16 Pi might have
1:17:24 it
1:17:28 [Music]
1:17:30 wait it's not zoto it's zong let's try
1:17:34 zot
1:17:35 [Applause]
1:17:39 tongue I might have pronounced it wrong
1:17:42 is it zot
1:17:50 tongue it looks like there might have
1:17:52 been a mixup with the term zoton the
1:17:55 search results I found primarily
1:17:57 reference zon as a brand name for
1:17:59 lanoo a medication used to reduce
1:18:02 stomach acid rather than a language it's
1:18:04 used to treat conditions like
1:18:06 gastroesophageal reflux disease and
1:18:09 peptic ulcer disease if zoton was
1:18:12 intended to reference a language it's
1:18:14 possible that there's still some
1:18:16 confusion or miscommunication about the
1:18:18 name or spelling of the language in
1:18:20 question languages and dialects can
1:18:23 often have names that are spelled or
1:18:25 announced similarly but refer to
1:18:27 completely different things if you have
1:18:29 any more details or if there's anything
1:18:31 else you're curious about feel free to
1:18:33 share it might be a Malaysian dialect
1:18:37 called zot Tang or
1:18:46 [Music]
1:18:54 Zang
1:18:58 Kyle you have 1.5 just click off of
1:19:00 advanced back to
1:19:02 Pro is that
1:19:05 true is it the is it the million
1:19:08 token Mayan oh
1:19:11 Myanmar tap to
1:19:13 cancel how about Myanmar is there a
1:19:16 language called Zang or Zang in
1:19:23 Myanmar oh laughing Thunder yeah I'll
1:19:25 I'll demo grock it's
1:19:28 amazing oh this thing stopped working
1:19:31 all of a sudden hang
1:19:36 on can you check Myanmar is there a
1:19:39 language called zong or Zang in
1:19:51 Myanmar yes there is a language called
1:19:53 zot also known as Z chin Zia or zo
1:19:57 spoken in Myanmar specifically in the
1:19:58 chin State the zong people are one of
1:20:02 the ethnic groups residing in Central
1:20:04 and Southern chin State the language is
1:20:06 known to have multiple dialects and does
1:20:09 not have a standard written form mainly
1:20:12 because of the various dialectical
1:20:14 pronunciations a way to write zong using
1:20:17 the Latin alphabet was developed based
1:20:19 on the lungo dialect which has been
1:20:21 widely accepted among zotung speakers
1:20:24 interestingly a zong translation of the
1:20:26 Bible was published in 2016 indicating
1:20:29 efforts towards preserving and form
1:20:31 formalizing the
1:20:33 language do you know the
1:20:40 language I don't have the ability to
1:20:42 learn or know languages in the way
1:20:44 humans do so I can't speak or understand
1:20:47 zun chin directly my knowledge is based
1:20:50 on the information that's been made
1:20:53 of let's try pi
1:20:59 hello there's a language in Myanmar
1:21:00 called zotung and I'm wondering if you
1:21:03 know
1:21:07 [Music]
1:21:11 it hey Kyle it's true there is a
1:21:14 language spoken in Myanmar called zong
1:21:16 also known as Zia it's spoken by the
1:21:19 zong people in rzua Township chin State
1:21:21 Myanmar it's actually a Continuum of
1:21:23 closely related and accents and it
1:21:26 doesn't have a standard written form
1:21:28 since there are multiple variations in
1:21:31 pronunciation it's fascinating how
1:21:33 diverse and complex language can be
1:21:35 isn't
1:21:36 it yes if I give you a sentence in
1:21:38 English can you translate
1:21:45 it unfortunately Kyle I don't currently
1:21:48 have the ability to translate into dang
1:21:50 it all right we were
1:21:53 close so those two don't have it someone
1:21:56 will have it somewhere um all right
1:21:59 let's
1:22:01 see demo Gro can be amazed changes that
1:22:04 we're about to see so I think mine has
1:22:06 gone crazy trying to tell me about a
1:22:08 dialect called
1:22:14 Karen
1:22:15 Joker
1:22:19 Joker all right so here's chat gp4 right
1:22:25 we know it we love it it's been it's
1:22:27 been weird I don't know if you've all
1:22:28 have noticed but the last two days this
1:22:30 shit's been weird like the custom
1:22:33 instructions don't seem to work or
1:22:35 sometimes they work but not
1:22:38 always but I'll show you speed
1:22:41 here so we'll
1:22:43 say
1:22:45 um tell me about the
1:22:50 Mayanmar
1:22:51 people that speak Zo
1:22:56 Tong I don't know if that's how you
1:22:58 spell it but close enough that's one of
1:23:01 the nice things about Chachi now look
1:23:03 how see how fast it's typing all right
1:23:06 so we're watching it the zotung language
1:23:08 also known as
1:23:09 zotung is a language spoken by the chin
1:23:12 people in Myanmar the chin people are of
1:23:15 the major ethnic minority
1:23:19 groups all right and so it's writing and
1:23:23 pretty solid so I'm going to grab this
1:23:28 prompt and I'm going to go over and I'm
1:23:30 going to go to a new
1:23:33 site called
1:23:35 grock and you're like
1:23:38 oh that's that's Elon musk's company and
1:23:43 you'd be
1:23:46 wrong he named his
1:23:49 stupid chatbot grock after this company
1:23:54 had has established their so he's just
1:23:57 being a tool but if you go to
1:24:01 gq.com
1:24:04 you get to this site now grock
1:24:08 grq is a chip manufacturer but they've
1:24:12 got this demo site and so what I'm going
1:24:15 to do is I'm going to put in the same
1:24:16 prompt and it's probably going to be
1:24:19 delayed for a second because there's so
1:24:21 many people using it so I don't want you
1:24:24 to count the time it takes before it
1:24:27 actually submits The Prompt but once it
1:24:29 submits The Prompt watch how quickly
1:24:31 this thing generates the response so
1:24:34 same exact prompt as what we just saw in
1:24:37 uh chat GPT that took whatever 15 20
1:24:40 seconds to
1:24:43 write boom
1:24:46 done 530 tokens per second is is what
1:24:50 this thing's doing um let's see write me
1:24:55 an
1:24:56 epic
1:24:58 poem
1:25:00 about uh let's see
1:25:04 about Z tongue
1:25:12 Warriors that's not a very long epic
1:25:14 poem is
1:25:15 it took it a second to write all that
1:25:18 528 tokens per second so that's about
1:25:22 375 words per second
1:25:26 so this is so they're doing this with
1:25:28 Hardware um Pate said he doesn't quite
1:25:31 know what they're doing here what we
1:25:32 don't quite know is how many of these
1:25:34 processors they're throwing at the
1:25:35 process right so if you throw many many
1:25:39 processors in parallel you could
1:25:40 probably get it to do some tricks but it
1:25:42 sounds like they're doing something
1:25:43 interesting with their chips that's
1:25:46 generating a very quick return so so the
1:25:50 the the tools that are out there are
1:25:52 getting better really fast and and you
1:25:55 know coming from sort of all angles so
1:25:58 again my my advice with this stuff is
1:26:01 just keep
1:26:03 playing this morning I saw a
1:26:06 scary video I don't agree with it but
1:26:10 we'll share it with the salon for
1:26:11 discussion it claims AGI has been dis
1:26:15 has been
1:26:16 achieved who's it from cosmic lover is I
1:26:19 I assume it's not David Shapiro let's
1:26:21 let me go see did you share it yet
1:26:25 um AGI has been
1:26:29 achieved cuz if this is just rehashing
1:26:36 the this is qar from about two months
1:26:39 ago is it this
1:26:43 one if it's just rehashing the qar stuff
1:26:46 we don't we don't know that it's been
1:26:47 achieved if it's just
1:26:49 that artificial general
1:26:53 intelligence qar AGI within seven months
1:26:57 that's David Shapiro if David Shapiro
1:27:00 says it's been achieved I'll look at it
1:27:01 I'll I'll believe it but if it's just
1:27:04 like this
1:27:07 dude more than just beautiful
1:27:11 dance oh my gosh wow okay so we have
1:27:15 some breaking news today open AI has
1:27:17 just announced that they have achieve
1:27:20 yeah that was two months ago that's qar
1:27:22 that's that this is all this is all chat
1:27:25 TMZ rumor
1:27:27 stuff um this is actually helpful though
1:27:31 I've been looking for zoto so finding
1:27:36 the right language is awesome great
1:27:38 great Jane do yeah zong
1:27:40 Myanmar from the chin chin people Chen
1:27:43 people I think
1:27:46 chin which is a large minority within
1:27:50 Myanmar Kyle that language is translated
1:27:53 by a specialized
1:27:55 llm oh laughing
1:27:57 Thunder do you know do you know which
1:28:00 laughing Thunder do you know which llm
1:28:02 does that language because if you do
1:28:04 post it in here um for Jane Doe that
1:28:07 that would be cool if we could solve
1:28:08 Jane Doe's problem tonight we found her
1:28:11 language and now we just need to find
1:28:13 the translator and then she can talk to
1:28:15 her neighbors so okay how exciting is
1:28:18 that how exciting is it that we are
1:28:20 living in a world right
1:28:22 now where ah
1:28:25 I think that's a Corey Sandler
1:28:27 Masterpiece there by the
1:28:28 way um we are living in a world right
1:28:31 now where not only can we find the
1:28:33 language but there is likely a large
1:28:35 language model where we can speak it so
1:28:39 accessibility think think about that
1:28:41 that not only do the do the neighbors
1:28:42 not speak English they speak some really
1:28:45 obscure
1:28:48 language and she's going to be able to
1:28:51 communicate with them if we can find the
1:28:52 right model like holy
1:28:56 [ __ ] which by the way if you haven't
1:28:58 seen these things do their language
1:29:00 thing I mean it it it is pretty
1:29:08 remarkable I'm going to start a sentence
1:29:10 in English and then I want you to finish
1:29:12 it in
1:29:19 Japanese gotcha let's give it a shot
1:29:21 Kyle ready here goes nothing the weather
1:29:25 today
1:29:28 is how I do no I want to say the first
1:29:32 half in English and then I want you to
1:29:33 complete it in
1:29:41 Japanese oh I see my apologies Carl
1:29:44 let's try again English I'm going to the
1:29:46 store to buy
1:29:52 Japanese is that better
1:29:55 no you didn't quite get it but that's
1:29:57 okay but anyway it can speak many
1:30:00 languages and you know you can just talk
1:30:04 it on your
1:30:07 phone have have someone at the grocery
1:30:12 store you can't communicate with talk
1:30:14 into it and then say translate that to
1:30:16 English I pasted the little bit I got
1:30:19 into the Irregulars Channel um Let me
1:30:22 let me try this Gemini thing I don't I
1:30:25 don't buy that if I go to GE if I go to
1:30:27 the non-advanced Gemini that I'm
1:30:29 actually going to
1:30:31 get
1:30:33 um that I'm actually going to get 1.5 I
1:30:38 don't buy it people I don't buy it
1:30:41 Gemini unless Pate says it's
1:30:44 true wait she always sounds like you
1:30:46 woke her up from a nap she has oh hey
1:30:50 Kyle it's so true okay here's Gemini
1:30:54 Gemini was updated C
1:31:00 update available today the new business
1:31:03 G okay this is the workspace
1:31:16 thing see find
1:31:22 one.5 no it's it's not out
1:31:25 yet it's not out yet you
1:31:29 weirdos don't get me all excited like
1:31:32 that dang
1:31:34 it my GPT can speak Elven and and dth
1:31:39 rucky but that's all the zong chin and
1:31:45 nose uh I think they were a peaceful
1:31:48 people that it's
1:31:53 possible what how so dang fast so Debbie
1:31:57 the way grock is doing that 500 token a
1:32:00 second response is they've got a uh
1:32:03 they've got a new kind of chip let me
1:32:05 see if I can find the uh let me go to x
1:32:08 and see if I can find there there was a
1:32:10 a ill an image of it I think I saw hang
1:32:13 on um
1:32:18 [Music]
1:32:23 x.com
1:32:26 grock
1:32:43 chip that's a lot of tweets tonight
1:32:47 facts responses we're faster because we
1:32:50 designed our chip and
1:32:52 systems it's an lpu a language
1:32:55 processing unit not a
1:32:58 GPU and P's got a TPU over there at
1:33:03 Google a tensor processing unit for
1:33:06 their tensor uh framework I guess
1:33:08 framework is that right we use open
1:33:11 source models but we don't train them we
1:33:12 are increasing access capacity weekly
1:33:15 stay tuned so I guess yeah you can buy
1:33:18 these chips to run your own
1:33:20 [ __ ] basically basically what they're
1:33:22 doing is so gpus are are designed to be
1:33:27 general purpose processors and they do
1:33:29 whatever they do in a way that makes it
1:33:31 friendly for large language models and
1:33:34 or for for AI machine learning training
1:33:37 and it's it's good at dealing
1:33:41 with machine learning processing
1:33:43 inferences I guess I don't know enough
1:33:46 about this they're doing something in
1:33:48 their chip that basically says we're
1:33:50 just going to assume we're going to be
1:33:52 doing large language model process
1:33:54 procing so we don't need any of the
1:33:56 general capabilities we're going to
1:33:58 custom design a chip that's specifically
1:34:00 designed to take out all the bottlenecks
1:34:02 of a GP GPU that give it its generality
1:34:06 its ability to do lots of things and
1:34:08 we're just going to make a chip that
1:34:09 does one thing and it's going to do it
1:34:10 really well and really fast so if you
1:34:13 wanted to use one of these lpu chips to
1:34:16 run your gaming system it wouldn't work
1:34:19 right it's more similar to a TPU than a
1:34:21 GPU yeah that that makes sense Pate cuz
1:34:23 the way way you describe tpus to me they
1:34:25 were more they were
1:34:28 simplified versions
1:34:31 um they had a cool picture of one where
1:34:34 it kind of
1:34:51 showed I mean what's wild about Gro is
1:34:54 it takes longer to write your prompt
1:34:57 than for it to answer
1:34:59 it there you go here look look at
1:35:03 this so the Nvidia 1070 chip you know
1:35:08 complexity leads to more compute costs
1:35:10 right so there's all these different
1:35:13 things look how simple the architecture
1:35:15 of this is right so it because it's just
1:35:18 doing one thing I guess they just built
1:35:20 a thing that you know it's way more
1:35:23 simple so there's a lot less mileage to
1:35:26 cover for each little operation and
1:35:30 there's probably all sorts of crazy ass
1:35:32 Gates
1:35:33 and caches and just nonsense here where
1:35:36 there's just probably none of it
1:35:38 here the grock chip pretty slick
1:35:52 huh oh
1:35:55 Rebel said try wait grock says zong
1:35:58 translated by Google translate deep L
1:36:01 deep L Microsoft translator let's go try
1:36:04 Google
1:36:05 translate but can Google translate talk
1:36:08 it can right Google
1:36:14 translate uh enter text let's
1:36:18 see what's it called Z zong they've got
1:36:23 Zulu Z
1:36:25 [Music]
1:36:27 Tong no can't do zot
1:36:29 [Music]
1:36:36 tongue nope
1:36:39 well it's possible I don't know how to
1:36:41 use Google
1:36:42 Translate do you think GPT
1:36:45 5 will just be using voice to voice like
1:36:48 a real person no I don't think so Jason
1:36:51 but I think it will be multimodal so so
1:36:54 GPT 4 actually GPT 3.5 you can do voice
1:36:58 to voice right
1:37:00 so you know so in the in the mobile app
1:37:03 you can just put it into voice mode
1:37:09 and what is the speed of an unladen
1:37:18 swallow the question about the speed of
1:37:21 an unladen swallow is a famous One
1:37:23 thanks in part to its humorous use in
1:37:25 the film Monty Python and the Holy Grail
1:37:27 got that right when asked about um so it
1:37:30 can do that now I don't think it'll go
1:37:32 exclusively to that it doesn't make
1:37:33 sense but it will be multimodal so
1:37:35 you're going to be able to upload video
1:37:36 you're going to be able to upload audio
1:37:38 it's going to be able to generate video
1:37:39 it's going to be generate audio um it it
1:37:43 may have very different kinds of
1:37:44 interfaces I don't know I I have no idea
1:37:48 but but Sam Altman
1:37:50 said gp4 is best understood as a
1:37:55 preview meaning it's just sort of a
1:37:58 stupid janky piece of
1:38:02 [ __ ] and then he says progress here is
1:38:04 not linear so GPT 5 isn't going to be
1:38:09 like a time better than GPT 4 it's going
1:38:13 to be dramatically
1:38:15 different
1:38:18 so yeah yeah gpus were designed for
1:38:22 graphics which are based on linear
1:38:23 algebra you know you know it's funny P I
1:38:26 was just I was talking to my friends the
1:38:27 other day and I was saying you know you
1:38:29 remember linear algebra I'm like every
1:38:32 GPU out there it's based on linear
1:38:34 algebra it's like how are we supposed to
1:38:36 do AI inferences for doing linear
1:38:39 algebra I mean am I right good Lord and
1:38:42 then they were like A2 + B2 equal C2
1:38:47 huh it was I don't it's that it could
1:38:52 have been a quadratic
1:38:55 situation too with the reverse in
1:38:58 inference on
1:39:01 the the Buckle cache on the back end of
1:39:07 the yeah you
1:39:11 MC I know math
1:39:18 good I went to Stanford for math well uh
1:39:23 Joe Stanford Joe Joe Stanford he uh he
1:39:26 he went to the York Community College
1:39:28 where where he learned math and then he
1:39:30 he taught me math pretty
1:39:42 good I have a TPU but it's not that
1:39:45 great
1:39:46 o does does recognition and
1:39:48 classification
1:39:50 well Pate are you going to put up with
1:39:53 that from Jo
1:39:54 ER he has a TPU but it's not that
1:39:58 great I'm going to let you two fight it
1:40:00 out I do not want to see this go down
1:40:03 it's going to be
1:40:04 ugly there's going to be images
1:40:10 generated by the way the Irregulars
1:40:12 tonight are like what's Kyle doing the
1:40:14 past two nights I just I just been
1:40:16 talking I don't know what I'm doing I'm
1:40:19 just it's been my my day you know what
1:40:21 it is my days have been really uh uh
1:40:24 jam-packed so like I got home tonight at
1:40:29 710 then I run out to Walgreens to get
1:40:32 my kids some drugs got back here at 7:30
1:40:35 8 finished eating at like 5 minutes to 8
1:40:39 or like I don't know whatever so I just
1:40:41 my brain is my brain is
1:40:44 scrambled I got scramble brain
1:40:49 people
1:40:51 ah did you see the up dated Will Smith
1:40:55 spaghetti videos I did but what I didn't
1:40:59 know and this is actually interesting
1:41:02 because Will Smith posted them what I
1:41:06 didn't know was were the new videos that
1:41:10 he compared to the old video were those
1:41:13 him just shoving spaghetti in his mouth
1:41:15 or was it actually AI generated like
1:41:17 does anyone know what tools were used um
1:41:21 my board meeting was good I demoed
1:41:25 um I demoed our new AI authenticity
1:41:28 engine which which is now it's so slick
1:41:31 I should show it to
1:41:34 you it's so slick what it does we got a
1:41:38 formatting [ __ ] better it looks really
1:41:40 good so showed off
1:41:43 that showed them the video translation
1:41:46 stuff showed them the personas bringing
1:41:49 personas to life creating personas and
1:41:51 then bringing them to life
1:41:54 um and they were like oh yeah okay yeah
1:41:57 this is good this is this is good and
1:42:00 then one of our board members members
1:42:01 called us after and said you got to get
1:42:04 moving you got to get moving because
1:42:06 this stuff's good like what you've done
1:42:08 is
1:42:09 good but 18 months from now no one will
1:42:13 give a
1:42:13 [ __ ] which is exactly what I'm
1:42:21 thinking Nvidia stock took a dive this
1:42:25 week perhaps no no so Nvidia stock took
1:42:27 a dive because earnings were coming up
1:42:32 and generally stocks will dip right
1:42:33 before an earnings call um generally if
1:42:36 stocks I heard this on Tik Tok so this
1:42:38 is definitely true um before the
1:42:42 earnings call if they if they go up
1:42:43 you're you're looking at a at a crash
1:42:45 they went down before the earnings call
1:42:47 and apparently um my buddy told me
1:42:49 Nvidia blew their numbers out of the
1:42:51 water today so I don't know if that's
1:42:52 true but if it is I would assume their
1:42:54 Stock's going to go back up so far the
1:42:57 concerns I've seen are related to power
1:42:59 consumption and it's software stack I
1:43:01 don't know about grock I would imagine
1:43:04 that their software stack is is shitty
1:43:06 and and weak because they're
1:43:08 startup yeah Google can translate most
1:43:11 it's definitely not
1:43:14 solo my H Honduran friend said AI is
1:43:17 much better at
1:43:21 translating uh uh uh
1:43:24 [Music]
1:43:31 my Google home answered
1:43:35 20.1 for the unladen
1:43:39 swallow I need a free I need free job
1:43:42 scan is it
1:43:44 possible
1:43:46 huh what's a free job
1:43:48 scan gimu G is that do you mean you want
1:43:53 it to write you a resume or you want it
1:43:56 to check your resume or like what do you
1:43:58 mean a job scin I don't know what that
1:44:00 means but chances are it's
1:44:04 possible
1:44:09 yes check the news on Salon regulars
1:44:12 chat I posted the scary video okay cool
1:44:14 let me go look at
1:44:16 that oh I haven't talked about the salon
1:44:18 tonight I should talk about the salon so
1:44:20 how many people are in here 91 nice um
1:44:23 if you're new to this AI [ __ ] or even if
1:44:25 you're not but if you're into this AI
1:44:28 [ __ ] and you're the only one in
1:44:31 your friend and family group that's into
1:44:35 it if you if you're like hey hey guys
1:44:39 guys have you seen this Ai and they're
1:44:41 like no shut up Terry we don't want to
1:44:44 hear about your stupid AI Terry it's all
1:44:47 you talk about is a i Terry maybe you
1:44:50 could talk about Budweiser once in a
1:44:52 while like the rest of us you big fat
1:44:55 dumb Dum and then you're like uh but I'm
1:44:59 skinny and I'm not a dumb dumb cuz I got
1:45:03 Ai and they're like shut up Terry about
1:45:05 you dumb
1:45:06 AI so if that's
1:45:11 you you
1:45:12 should join the AI Salon which I'll talk
1:45:15 about in a second
1:45:22 news
1:45:28 Gemma I don't see that video about
1:45:35 AI
1:45:38 Reddit is it let's
1:45:42 see it's a small world a after all
1:45:45 Disney Picks 11 Labs audio shake and
1:45:49 Promethean AI for their accelerator
1:45:51 program oh that's interesting
1:45:55 huh I don't see the video about AGI let
1:45:58 me let me search uh search a
1:46:02 [Applause]
1:46:06 GI did did you put it maybe you put it
1:46:08 in water
1:46:13 cooler predicting the next dayi
1:46:22 winter
1:46:24 [Laughter]
1:46:29 sorry I just saw another news item
1:46:31 [ __ ] meta every [ __ ] time every
1:46:35 [ __ ] time except in
1:46:38 close meta's Chief AI researcher says
1:46:41 open ai's world simulator Sora is a dead
1:46:45 end every time open AI does something
1:46:48 that like breaks faces this guy comes
1:46:51 out and goes yeah it's not that
1:46:52 impressive we've got better it's like
1:46:57 dude just shut Up release something
1:47:03 better the [ __ ] belly aching in this
1:47:06 industry is staggering
1:47:11 anyway
1:47:13 unbelievable thoughts I think he's full
1:47:15 of it for now oh
1:47:22 God that's good at least people in the
1:47:24 salon are trolling him it's not just me
1:47:26 people in the salon are trolling him too
1:47:29 oh good
1:47:31 lord here's the thing that blows
1:47:33 everyone's mind meta yeah it's not that
1:47:35 great whatever did you see the [ __ ]
1:47:38 Zuckerberg thing you know you know I uh
1:47:42 I went to look at the Vision Pro and I
1:47:44 figured well you know because we're
1:47:46 seven times cheaper you know it's going
1:47:48 to be a better a better value and so I
1:47:51 went and looked at uh at The Vision Pro
1:47:53 what did I learn but the The Meta Quest
1:47:56 is actually a better
1:48:00 product
1:48:02 Mark just just tell people it was an
1:48:05 avatar that it wasn't really you saying
1:48:07 that it's
1:48:09 [Laughter]
1:48:13 embarrassing it's it's not the same
1:48:17 thing you you you weirdo Irregulars
1:48:20 Channel I think oh world is not ready
1:48:22 for GP T 4 or five four oh
1:48:26 gptt Irregulars
1:48:30 okay there's the grock tensor Factory
1:48:33 Gemma okay that's Gemma here we
1:48:37 go open ai's new statement is
1:48:42 concerning when how old is
1:48:45 [Music]
1:48:48 this so there was a statement made by an
1:48:51 open AI employee that TR shocked me to
1:48:54 my
1:48:55 call what
1:49:00 what watch on
1:49:03 YouTube I want to see
1:49:06 Studio Freedom meets absolute efficiency
1:49:10 intelligent design scal by default and
1:49:13 even Advanced layouts adapt with respond
1:49:15 so there was a statement made by an open
1:49:18 AI employee that truly shocked me to my
1:49:21 core because of the ramifications of of
1:49:23 the statement and once I explain to you
1:49:26 where this statement comes from and what
1:49:28 it means you're going to understand why
1:49:30 we are clearly not ready for gbt 5 and
1:49:32 why open AI are basically saying that
1:49:34 look we're really not ready for any
1:49:36 advanced AI system just yet yeah yeah
1:49:38 yeah this is this is just them talking
1:49:40 about the [ __ ] that came out before so I
1:49:44 it's not here like okay it is possible
1:49:48 that within open AI they stumbled upon
1:49:50 something that is good enough that
1:49:52 scared the [ __ ] out of people um and
1:49:54 they're going to make really sure that
1:49:56 they do whatever safety [ __ ] they need
1:49:58 to before they put it out in the world
1:50:02 um their releasing of
1:50:06 Sora showing that Sora is so much better
1:50:09 than the current state-ofthe-art is
1:50:12 probably them just saying okay like
1:50:15 here's probably what the video
1:50:18 capability of this thing is going to be
1:50:20 and then they hint at it's good at video
1:50:23 because it actually understand and Sim
1:50:25 simulates the real world so there's
1:50:27 they're probably going to like tease out
1:50:29 a bunch of piece parts of GPT 5 so that
1:50:33 we can kind of prepare for it because if
1:50:35 they just dumped all that [ __ ] at once I
1:50:37 think people would lose their [ __ ]
1:50:38 mind so um I'll I'll watch that video at
1:50:41 some point but this just sounds like
1:50:43 speculation this is chat TMZ kind of
1:50:46 [ __ ] so I wouldn't worry too much about
1:50:48 it
1:50:52 um
1:50:59 [Music]
1:51:01 since you actually did the iris
1:51:04 skin do you invest in World ID
1:51:09 no CU I don't I don't have I don't
1:51:11 really have a
1:51:17 a here's the thing about world ID and
1:51:20 things like um and things like it
1:51:26 is the world right now is not clamoring
1:51:33 for
1:51:35 tokenization and if you don't know what
1:51:37 tokenization is
1:51:39 basically so what the what the question
1:51:42 is what what Debbie's asking about is
1:51:45 Sam waltman's got this company called
1:51:47 worldcoin and part of worldcoin is a
1:51:49 cryptocurrency and part of worldcoin is
1:51:52 this thing called World ID and World ID
1:51:56 is
1:51:56 a um deidentified so they don't know
1:52:01 what your who your name is but you go to
1:52:03 this physical
1:52:05 orb which is the most unfortunate name
1:52:08 like if you want to make
1:52:10 people if you want to make people
1:52:12 freaked out about the future name the
1:52:14 thing that scans your iris to prove your
1:52:17 identity or your retina the orb um so
1:52:21 they've got this thing called the orb
1:52:23 and you get your retina scanned and then
1:52:25 what it does is it Compares your retina
1:52:28 to all the other you know hashes in the
1:52:31 database the the the basic ID in the
1:52:34 database to make sure you're a unique
1:52:36 human if you're a unique human then you
1:52:39 get this world ID it doesn't have your
1:52:41 name on it but as the holder of this nft
1:52:44 they're not going to call it that but
1:52:46 it's an nft as the holder of this
1:52:50 nft it says you're real
1:52:54 human
1:52:56 now they've got a whole development
1:52:58 environment where you could create or
1:53:02 let's let's say Elon Musk all of a
1:53:04 sudden you know ped up with Sam Alman
1:53:06 again and liked him and
1:53:08 said Twitter is going to adopt World ID
1:53:12 because X is taking Bots really
1:53:16 seriously and we're going to create a
1:53:18 corner of Twitter that is only for
1:53:21 people that have verified human
1:53:24 identity oh well that would be pretty
1:53:26 cool like it' be cool it would be nice
1:53:28 to know you've got a community there
1:53:30 where you could do something like that I
1:53:32 think there's a potential experiment we
1:53:34 could do in the salon about like you
1:53:37 could opt into doing something it
1:53:39 doesn't need to be World ID I don't give
1:53:40 a [ __ ] what it is but some sort of
1:53:43 blockchain
1:53:44 verified humanness
1:53:48 right
1:53:51 because what we're going to find out out
1:53:53 in 2024 is
1:53:55 that the Deep fake Technologies and the
1:53:58 Deep fake tactics during a presidential
1:54:01 election year are going to make 2016
1:54:03 look like Child's Play so it's possible
1:54:07 at the end of 20204 the world loses
1:54:10 their [ __ ] and says we've got to know
1:54:12 when I'm talking to a real person and
1:54:14 when I'm looking at real content that
1:54:16 was produced by a real
1:54:18 person and you need some you need some
1:54:21 kind of tokenization you need some sort
1:54:23 of thing blockchain is one technology
1:54:26 there's others just like
1:54:28 databases but you could create a system
1:54:30 that says
1:54:31 okay if you're on this site that looks
1:54:34 like Twitter or looks like
1:54:37 Facebook and a person's got a green
1:54:39 check mark by their name it means that
1:54:41 green check mark doesn't mean they paid
1:54:43 15 bucks to Elon Musk that green check
1:54:46 mark means they're human and they're
1:54:48 uniquely human so they're not a spoof of
1:54:51 someone else
1:54:54 and then we'll be able to track on the
1:54:57 blockchain that this real human wrote
1:55:00 this article so we can maintain that
1:55:03 relationship and then we can even say if
1:55:05 the article's got a green check mark it
1:55:07 also means it hasn't been
1:55:09 modified right so that's that's what's
1:55:13 needed and I I or that's what's going to
1:55:15 be needed and I think it's going to be
1:55:17 needed kind of in the next year or two
1:55:20 but until there's a
1:55:21 compelling public outcry that we need
1:55:24 this [ __ ] no one's going to give a [ __ ]
1:55:26 about it because you'd basically have to
1:55:29 re you'd have to re architect or you
1:55:31 you'd have to you'd have to build some
1:55:33 sort of system like this into Facebook
1:55:36 into Twitter into payment platforms and
1:55:40 right all that sort of stuff what Sam
1:55:41 waltman's trying to do is he's saying
1:55:43 here's the development environment Go
1:55:46 integrate this into a bunch of different
1:55:48 things like right now I think
1:55:50 there's 15 different apps that it that
1:55:53 it um that world ID is connected into
1:55:56 like like you can you can use it on
1:55:59 Reddit so I don't know how it works on
1:56:01 Reddit I haven't played with it I don't
1:56:02 give a [ __ ] right now cuz it's not it's
1:56:04 not an issue but it's going to
1:56:07 be does it check for blood flow in the
1:56:10 eyeball I have no idea what it did it
1:56:11 was just all it did was you kind of had
1:56:13 to open your eyes wide and then it was
1:56:15 like a circle and it it it sort of Drew
1:56:18 the circle from the bottom up it was
1:56:20 like bing and then just all of a sudden
1:56:24 on my phone it said you've been
1:56:26 verified and then I had this little
1:56:28 passport looking thing and it says
1:56:30 United States it has the date on it has
1:56:32 my number and then it's got a you know
1:56:35 like a it's a it's like a an nft wallet
1:56:39 address right for the
1:56:42 ID and then whoever supports that
1:56:45 protocol you'll be able to log in with
1:56:48 your ID and it'll know it's you and it
1:56:50 and other people will know that that's
1:56:52 who I interacted with before and that's
1:56:54 a real
1:56:55 person
1:57:04 so oh look at that I was talking about
1:57:06 the US elections the bottom of the
1:57:08 screen right now learn more about the US
1:57:10 elections Tik tok's not
1:57:13 listening he said boobies he said
1:57:19 boobies I took an exam that had an image
1:57:22 on the of the veins of my hand oh that's
1:57:24 fascinating geez that's
1:57:27 creepy I've seen lots of movies that pop
1:57:30 out an eyeball to to to get it thrown
1:57:34 into the retina scanner well here's
1:57:36 what's fascin so Dina what's fascinating
1:57:38 about world ID is they SC they scan your
1:57:41 retina once create a
1:57:44 hash with so a hash is just a a long
1:57:47 string of numbers they take whatever
1:57:49 data they get from your retina they turn
1:57:50 it into a unique long string of numbers
1:57:54 and then they throw out the retina data
1:57:55 so they claim right possible that's not
1:57:58 cuz you
1:57:59 know tech companies but in theory they
1:58:04 throw out the retina data and then
1:58:05 you've just got that that unique hash
1:58:07 and then all they're doing is comparing
1:58:09 that hash to the other hashes of people
1:58:11 that scan their retinas in the database
1:58:12 and if it's Unique then they write to
1:58:15 the blockchain that this hash that this
1:58:19 person you know this person's account
1:58:23 signed up for is is a unique human being
1:58:26 and that's all World ID is you know when
1:58:30 I scan the retina it it generated this
1:58:32 number is that number unique compared to
1:58:34 all the other numbers we've created if
1:58:37 yes then give them an
1:58:39 ID and so that's the only time so so
1:58:42 you're not having to rescan each time
1:58:44 it's if you have the password to your
1:58:48 world ID account then you you your ID is
1:58:52 good
1:58:53 that's
1:58:54 it that's that's the cool thing about
1:58:56 the
1:58:58 blockchain like blockchain Technology's
1:59:00 cool it's unfortunate that the nft Bros
1:59:03 [ __ ] it all up by Ste stealing
1:59:05 billions of dollars from people but
1:59:08 they're in the corner where they belong
1:59:10 right now cuz they were bad boys and a
1:59:12 few girls but mostly bad
1:59:15 boys scaming little old
1:59:18 ladies and hedge
1:59:21 funds
1:59:26 and uh yeah so when people start
1:59:30 clamoring for that [ __ ] when there's an
1:59:31 actual use case like it would be swell
1:59:34 to know if Kyle is Kyle and not some
1:59:36 [ __ ] Avatar cuz he's been playing
1:59:39 around with
1:59:41 avatars we'll be able to do that I've
1:59:44 seen a lot of movies oh yeah we just did
1:59:46 that
1:59:47 one do you have your did it have your
1:59:49 owner's name too no it's it's anonymous
1:59:58 [Music]
2:00:05 jimc I'm confirming now that I'm human
2:00:08 football
2:00:10 emojis it puts the retina in the basket
2:00:13 or else it gets the hose
2:00:14 [Laughter]
2:00:17 again
2:00:19 H share bear what's happening good to
2:00:22 see you thank you very much can we
2:00:25 compose music with AI yet Kayla come on
2:00:28 of course we
2:00:30 can go to let's go have some fun shall
2:00:33 we go too. a s.
2:00:39 a and we will say make a
2:00:44 song I think I got to log in here hang
2:00:46 on is this my
2:00:50 stuff uh uh uh sign
2:00:54 up here we
2:01:09 go so when you're here you go to create
2:01:12 and then we'll say
2:01:16 um I would
2:01:19 like like a pop
2:01:24 ballad in a minor
2:01:27 key
2:01:29 about a love named
2:01:33 Kayla
2:01:36 who's more
2:01:39 synthetic than
2:01:42 real I normally do this about Sydney but
2:01:45 since you ask Kayla we'll do it for you
2:01:47 synthetic then
2:01:48 real all right so you just describe it
2:01:51 and then it's going to write the lyrics
2:01:54 it's midnight Turtle
2:01:56 up it's going to write the lyrics first
2:02:00 and then it will write a song and then
2:02:01 we get to listen to it sing the
2:02:04 song oh it's done already they they've
2:02:06 sped this thing up this used to take a
2:02:08 lot longer all right synthetic love here
2:02:10 we go minor key pop
2:02:13 [Music]
2:02:17 ballad
2:02:20 it's but there something about
2:02:24 her that just doesn't
2:02:26 [Music]
2:02:31 feel
2:02:33 [Music]
2:02:34 laugh in the
2:02:39 cold I hate this all right let's try
2:02:42 this
2:02:44 [Music]
2:02:46 one Stand
2:02:47 [Music]
2:02:50 Out touch
2:02:56 that's shitty all right so let's go pop
2:02:58 ballad let's go uh let's go
2:03:02 Americana Americana
2:03:05 ballad let's see I'll kill the
2:03:09 key air
2:03:12 Iana
2:03:13 ohir Iana
2:03:18 okay I just hated those two songs but
2:03:21 yes you can make music now oh no I mean
2:03:23 compose oh give me sheet music um I'm
2:03:26 sure there are some uh composition tools
2:03:29 that have thrown AI layers on it I would
2:03:31 say go check go check uh um future
2:03:48 [Music]
2:03:51 pedia.com comes a girl shine so Bri her
2:03:55 Smiles like sunshine on a Cy
2:04:00 day but there's something about her that
2:04:03 just doesn't feel the
2:04:08 [Music]
2:04:11 same it it it tried to say Kayla was
2:04:14 like oh
2:04:17 Ella in awn
2:04:21 where there's a girl named Kayla a heart
2:04:25 of she's not quite like the rest she
2:04:29 she's got a soul it's more
2:04:33 synthe with your plastic smile you're a
2:04:43 masterpiece all right yeah so anyway so
2:04:45 I don't know I don't know about the
2:04:47 Sheep
2:04:48 music I I would I would be surprised if
2:04:51 not
2:04:53 I would be
2:04:59 surprised waiting on Quantum
2:05:06 distribution oh yeah a human being's
2:05:08 good we know she's human because it's in
2:05:10 her name see you get it verified she
2:05:14 wrote it
2:05:15 there you can't fake
2:05:19 that listen Linda Kyle I'm a toddler in
2:05:23 this AI
2:05:24 world no you know what Kayla you come
2:05:27 here a lot so you are not a toddler in
2:05:29 this AI world that's this is this I'm on
2:05:32 a mission I am on a
2:05:35 mission especially so if you join the AI
2:05:37 Salon I'm going to talk about the AI
2:05:39 Salon I I'll wrap up with a with a
2:05:41 talking about the AI salon I'm on a
2:05:44 mission to get people in the AI Salon to
2:05:46 stop apologizing stop apologizing for
2:05:48 not being an expert in this cuz no one's
2:05:51 [ __ ] an expert in this this I am so
2:05:54 clueless about this
2:05:56 stuff so
2:05:59 clueless like M KY you talk about it all
2:06:01 the time I know
2:06:03 exactly
2:06:04 exactly if there's anyone out there that
2:06:07 tells you confidently they know what the
2:06:09 [ __ ] is going on come back here and I
2:06:12 will remind you they're full of
2:06:15 [ __ ] and that you shouldn't feel bad
2:06:17 about where you are in this journey if
2:06:19 you're paying attention to this stuff in
2:06:21 any way like you're asking questions
2:06:22 like can it compose yet can it do sheet
2:06:24 music can it do this can it do that
2:06:28 you're way ahead of where most people
2:06:29 are right now right the only job I think
2:06:33 any of us have right now is sort of feel
2:06:35 in the dark room and feel where the
2:06:37 furniture is and where the walls are and
2:06:39 just know that they keep moving the
2:06:40 furniture and the walls keep pushing
2:06:42 farther
2:06:44 out which is why it's so overwhelming
2:06:47 you're like okay I think I got to figure
2:06:49 oh [ __ ] there was a couch here wait wait
2:06:51 where's the wall oh god well I'm I'm all
2:06:53 alone
2:06:54 [Laughter]
2:06:58 again that's what's going on right
2:07:02 now want to be able to write some actual
2:07:05 sheet music not just the loops that they
2:07:06 do yeah
2:07:08 I I'm sure there's some stuff out there
2:07:10 that's doing
2:07:12 it because training training a large
2:07:15 language model on music notation is no
2:07:19 different than training a large language
2:07:21 model on on Text data or image data or
2:07:25 the other kinds of data it's it's all
2:07:28 getting embedded into
2:07:29 this into this latent space and then the
2:07:33 output data is just whatever it was
2:07:35 trained on right it's you're just it's
2:07:36 it's all the same [ __ ] which is which is
2:07:39 one of the remarkable things about it is
2:07:43 you want it to be good at something just
2:07:45 give it a different data set and it will
2:07:47 be I mean it's not that easy but it kind
2:07:50 of is
2:07:52 you just you just need lots of money
2:07:54 let's hit 30,000 shall we all right
2:07:56 we're at 28,500 for the tappy
2:07:59 tappers my name is Jeff movie 22nd
2:08:02 Street my name is Jeff movie 22nd Street
2:08:07 I lost I don't know the reference no I
2:08:09 mean
2:08:10 compos absolutely mindblowing
2:08:13 brother I did a heavy heavy rapso baby
2:08:17 llama mama drama that's good I like it
2:08:19 solid oh my God sounds like the radio I
2:08:22 know
2:08:24 right the suo stuff and here's the thing
2:08:27 about the sunno
2:08:29 stuff it's janky right I I would say
2:08:32 that the quality of the sunno music is
2:08:36 is about the same from a quality
2:08:39 standpoint as Runway ml or pabs video
2:08:43 sometimes it's brilliant most times it's
2:08:45 like it close good try good try there
2:08:50 buddy but the next iteration of it is
2:08:54 just going to be stupid it's going to be
2:08:58 like and then someone's going to come up
2:09:00 with a killer interface where you can
2:09:02 just turn dials and Sliders and [ __ ] and
2:09:05 just generate songs it'll automatically
2:09:08 put them in tracks it'll automatically
2:09:10 do the
2:09:12 notation would you like that for a
2:09:14 symphony or a three piece you
2:09:17 know emo band with a cello a violin and
2:09:22 a
2:09:24 Bas yall know the band Hugo
2:09:28 Largo I saw them in New York in the 80s
2:09:31 they were one of my favorite bands Mimi
2:09:33 Goose was the singer the lead
2:09:35 singer and they had
2:09:37 a a violin a cello a base it might have
2:09:42 been two
2:09:45 violins Kayla and so chat GPT is pulling
2:09:49 fromo or do you have to use each tool
2:09:51 separ Seely uh no they're all separate
2:09:53 right now so sunno is a is a standalone
2:09:57 tool thank you Sher
2:09:59 bear
2:10:01 um a lot of the tools right now are
2:10:03 Standalone
2:10:06 tools
2:10:08 um chat GPT 4 is multimodal so it can
2:10:12 make images and it can make um
2:10:17 code and it can do data analysis but it
2:10:19 can't make music and it can't do voice
2:10:22 synthesis and it can't
2:10:25 do um video although live tough It's
2:10:30 stuer when you're stupid um it can't
2:10:34 do video although it you know it will be
2:10:38 able to soon as soon as they do their
2:10:40 red team testing on the video stuff see
2:10:43 I have a
2:10:44 feeling I don't think we're going to see
2:10:46 the sunno I don't think we're going to
2:10:47 be able to play with the sunno video
2:10:49 anytime soon I my guts telling
2:10:52 me every two weeks they're going to drop
2:10:56 some new
2:10:58 capability like look what else is coming
2:11:00 oh it can do music now and it's going to
2:11:02 be really [ __ ] good music and then
2:11:05 it's going to be like oh it can
2:11:07 do then it's going to be like oh it can
2:11:10 do um animation now oh it can do 3D
2:11:14 objects now oh it can do 3D World
2:11:16 creation now I just I just have this
2:11:19 sneaking suspicion they're going to just
2:11:22 plop out little
2:11:24 capabilities that at some point chat GPT
2:11:28 is just going to like level up and it's
2:11:30 just going to have all this [ __ ] in it
2:11:33 but I don't know I don't [ __ ] know
2:11:35 chat GPT uses Dolly three yes it does
2:11:39 can we hit 35k of course we can Sher
2:11:42 Bear all right 35k we're going to get
2:11:44 there we're 324 right now we'll hit that
2:11:46 we'll we'll hit 40 before I finish
2:11:48 saying goodbye because I haven't even
2:11:50 talked about the salon yet Sora
2:11:53 yes songwriters are [ __ ] well so
2:11:56 here's the thing where's the
2:11:59 bottom every time someone sees something
2:12:02 like Sora the immediate response is
2:12:04 songwriters are [ __ ] when you see or
2:12:07 or sunno you think songwriters are
2:12:09 [ __ ] you see Sora you think
2:12:10 videographers are [ __ ]
2:12:11 cinematographers are
2:12:14 [ __ ] they're either [ __ ]
2:12:18 or they get AI literate and they get
2:12:22 really good at using tools like
2:12:24 Sora and they just
2:12:27 replace their current way of making
2:12:31 film or their current way of making
2:12:33 songs with a new
2:12:35 tool but they're still the one in the
2:12:38 director's seat right and and you can
2:12:41 you can make all the judgments you want
2:12:43 about well it's not the same as making
2:12:45 music if I put it in a recorder in
2:12:48 layers and do it it's different than
2:12:49 that is it is different than that but
2:12:53 putting stems into a Daw is different
2:12:57 than recording on
2:12:59 tape right and recording on tape was
2:13:02 different than recording on wax
2:13:07 cylinders it's a new
2:13:09 tool and so I think
2:13:14 that some people in those specific
2:13:16 professions if they just sit back and go
2:13:19 well [ __ ] it that's not real music
2:13:22 they're going to they're going to be
2:13:24 they're going to face a lot of it's
2:13:26 going to
2:13:28 suck but there's going to be people in
2:13:30 those professions that are like [ __ ] it
2:13:32 I'm leaning in I'm going to go learn
2:13:33 learn into
2:13:35 tool and what they might discover is
2:13:37 they were really good at music with the
2:13:39 traditional music and they're really
2:13:41 good at music with AI music but they're
2:13:44 also really good at AI video and they're
2:13:46 also really good at AI images and
2:13:48 they're also really good at AI
2:13:50 writing and maybe now rather than just
2:13:52 writing the songs for a m movie they're
2:13:55 going to create the whole
2:13:57 movie and so maybe that that person that
2:14:00 had that incredible taste to make that
2:14:02 incredible music now they're making full
2:14:04 narrative Journeys for us because they
2:14:09 can so yes they're [ __ ] and there's a
2:14:13 massive opportunity for them to step
2:14:15 into whatever the new tools are going to
2:14:17 be I'm a songwriter and I think it's
2:14:19 useful as hell there you go Scott fate
2:14:22 Kevin Vasquez in here feels the same way
2:14:24 he uses suo to write songs pulls them
2:14:27 into the Daw splits them into separate
2:14:30 tracks and then uses it as part of his
2:14:33 songwriting process why not come up with
2:14:37 50 [ __ ] ideas pick the two or three
2:14:40 that are interesting pull them pull them
2:14:42 into a dah split them into tracks you
2:14:44 know comp your tracks add whatever else
2:14:47 you want to add and write a [ __ ] song
2:14:50 why not there's no [ __ ] rules go
2:14:54 play it's like writers that are like
2:14:57 I'll never use chat
2:15:01 GPT you'll never use the most powerful
2:15:04 writing tool in the history of the human
2:15:10 existence because
2:15:12 why
2:15:15 Pride you don't want to face the fact
2:15:18 that this thing might be better than you
2:15:19 at
2:15:20 words
2:15:25 what if it augments you what if you're
2:15:28 super good at words and you say here's
2:15:30 the world's most powerful writing tool
2:15:32 in the history of humanity what if I
2:15:35 took my Brilliance and augmented it with
2:15:38 some of that fine
2:15:40 [ __ ] you'll be better than people who
2:15:43 were not excellent
2:15:46 writers but if you just sit there on
2:15:48 your [ __ ] soap box with your arms
2:15:50 crossed
2:15:52 a bunch of people that were never good
2:15:53 with words are going to get really good
2:15:55 with words and they're going to [ __ ]
2:15:56 eat your
2:15:58 lunch shame on me so late oh quick take
2:16:01 Jake you you're gonna have to change it
2:16:03 to slow slow Toe
2:16:07 Joe just you can't you can't be running
2:16:10 around with quick take Jake as your name
2:16:12 and show up an hour and a half [ __ ]
2:16:14 later two 2 hours and 15 minutes later
2:16:17 it's it just doesn't
2:16:19 work you got you got brand in congruity
2:16:23 young
2:16:24 man all right I got to go people artists
2:16:28 are the same way in particular
2:16:30 tattooists and people who use standard
2:16:33 mediums I I mean honest to God
2:16:35 tattooists right
2:16:37 now tattooists that Embrace AI are going
2:16:40 to be able to do such Banger [ __ ]
2:16:44 insane
2:16:45 designs that are going to show up on
2:16:47 people's skin and it's going to it's
2:16:49 what's here's what's going to happen
2:16:52 CU it's not about that it was created in
2:16:54 AI it doesn't [ __ ] matter what tool
2:16:56 they used if a tattoo artist says here's
2:16:59 I've got a vision for how tattoos should
2:17:02 look and if that artist gets good at
2:17:06 working with generative AI to not only
2:17:10 take their Vision but expand it and come
2:17:12 up with some completely new twist that
2:17:15 is who they are but like an elevated
2:17:19 version of it and then they start
2:17:20 tattooing that and it blows people's
2:17:23 [ __ ] minds and people like holy [ __ ]
2:17:25 where'd you get that tattoo I got it
2:17:27 from
2:17:28 Bob [ __ ] Bob's amazing Bob's
2:17:32 amazing is anywhere in that conversation
2:17:35 AI probably
2:17:38 not why cuz who gives a [ __ ] if Bob drew
2:17:42 it like this or if Bob drew it like this
2:17:44 go make me
2:17:46 [ __ ] Bob got good with I know listen I
2:17:52 understand that Bob is not the name of a
2:17:55 tattoo artist I understand that that's
2:17:57 not the
2:17:58 point don't judge me you you know
2:18:02 there's a lot of judgment right now Bob
2:18:05 a tattoo artist named Bob a lot of that
2:18:08 questioning going on it's just it's not
2:18:09 appropriate it's not what we're talking
2:18:12 about my choice of names for my
2:18:15 fictional
2:18:16 characters sometimes I got it sometimes
2:18:19 I don't I'll admit it I slipped up
2:18:22 tonight tattoo artists are not named
2:18:27 Bob you know sure they may have been
2:18:31 given the name Bob but they're like
2:18:32 shark Bob right shark Bob's a tattoo
2:18:36 artist that's not the point that's not
2:18:38 what we're talking about here I don't
2:18:40 know why you brought this
2:18:43 up if you didn't bring it up I wouldn't
2:18:45 be talking about it all right fine I
2:18:49 don't consider myself less of a writer
2:18:51 for using AI to fill out my ideas
2:18:53 exactly I don't consider myself less of
2:18:55 a writer because I use spell
2:18:58 check I'm like great I didn't know how
2:19:01 to spell that word now I
2:19:04 do I just heard Beyonce use the kids
2:19:07 song from hey Franklin as her
2:19:08 inspiration oh fascinating huh you're
2:19:11 good oh thank you very much Ron
2:19:14 digity I'm good I'm I'm off my [ __ ]
2:19:20 rocker
2:19:23 oh we need another 24-hour live with
2:19:25 Kyle oh God no all right listen okay so
2:19:29 here's the deal there's 111 people in
2:19:34 here one of the things that I think you
2:19:38 need right now that not would be nice to
2:19:41 have one of the things I think you need
2:19:43 right now if you want to get AI
2:19:46 literate is
2:19:49 community you need to be hanging out
2:19:52 with people that are trying to figure
2:19:53 this [ __ ] out cuz no one can figure it
2:19:55 out right now it's too massive it's too
2:19:57 big it's moving too fast there's too
2:19:59 many disciplines involved nobody's got
2:20:02 the key nobody nobody zero people have
2:20:05 the
2:20:07 key so you need
2:20:09 Community a way you can get community is
2:20:11 just keep coming back here like the
2:20:13 community that formed here the the sort
2:20:15 of jam session that happens here every
2:20:17 night is absolutely [ __ ] inspiring
2:20:19 and cool so you can just keep coming
2:20:20 back here
2:20:22 but what I would strongly urge you to
2:20:28 do
2:20:31 is go
2:20:34 there the
2:20:39 salon. and that's a link
2:20:46 tree it's going to take you
2:20:49 here and you're like well that doesn't
2:20:51 look like all that and then the top
2:20:54 link is the AI Salon Community now I'm
2:20:58 not trying to sell you [ __ ] this doesn't
2:21:00 cost
2:21:02 anything but it's not
2:21:04 free so so here's what I mean by
2:21:08 that this community is really [ __ ]
2:21:11 special like it it is a really unique
2:21:14 place and so when I say it's not free
2:21:18 what I mean is this I don't want you
2:21:20 just to join it to join it I want you to
2:21:23 actually take a little time to
2:21:24 understand what we're about so there's a
2:21:27 section called Welcome to the salon
2:21:29 welcome to the AI
2:21:31 salon and it talks about who we are why
2:21:34 we do what we do how we do it and what
2:21:36 we're looking for like very short see
2:21:37 how short that is I have ADD I can fit
2:21:39 that on one scroll with pretty big type
2:21:42 because I'm old and I got my type set
2:21:44 big because that's how [ __ ] happens you
2:21:46 people with your young eyes [ __ ]
2:21:48 enjoy it while you got it yeah jerks
2:21:55 Anyway by the way we're up to
2:22:00 38,000 um the thing I really want you to
2:22:04 look at are the values so here's why the
2:22:06 values are really important and here's
2:22:08 what I mean this is not
2:22:10 free my humble request and I do I do
2:22:13 mean humble my humble request is that
2:22:15 before you join this community check out
2:22:18 our values and make sure they resonate
2:22:20 with you I'll go through them quick here
2:22:22 but it's a big deal if you don't know
2:22:25 anything about growing communities or
2:22:27 big businesses a lot of times you'll
2:22:29 hear businesses talk about values our
2:22:31 corporate values are blah blah blah blah
2:22:33 blah blah blah and they say it like it's
2:22:34 a successories poster and there's no
2:22:37 soul to it and there's no it's it's just
2:22:39 they're awful
2:22:41 right values are actually really
2:22:43 important whether you declare them or
2:22:46 not every Community has values sometimes
2:22:50 they're not declared sometimes
2:22:51 communities are awesome sometimes
2:22:52 they're toxic sometimes they're
2:22:55 funny sometimes they're
2:22:58 inconsistent this community right now is
2:23:00 really remarkable and so these values
2:23:02 were not just made up out of thin air
2:23:04 thin air about 6 months into this we
2:23:06 said hey this thing's starting to grow
2:23:09 and as it grows what do we not want to
2:23:11 lose what do we want to
2:23:14 amplify and so these values are what's
2:23:17 already here so my request is what what
2:23:19 this is going to cost you to be a part
2:23:22 of is that you just take a look at
2:23:26 these and commit to them like commit to
2:23:30 like okay yeah I can I can do that I'm
2:23:32 down with that if you read these values
2:23:34 and they turn your stomach and you're
2:23:35 like H that sounds awful you'll hate
2:23:38 this place if you look at these values
2:23:40 and you're like oo that gives me a
2:23:43 little tingly down in my we Willy wanker
2:23:45 then fantastic Bob tell him what he's
2:23:48 won so the first one is curios
2:23:51 yes I just said we Willy wanker why you
2:23:54 weirdos you you ain't never heard of
2:23:56 Comedy
2:23:58 before we want people in here who are
2:24:00 curious not cynical like you it's not
2:24:03 very good okay great it's not good why
2:24:05 is it not good is there something better
2:24:08 well how do you make it good could you
2:24:09 could you do better prompts if you give
2:24:11 it a shitty prompt does it give you a
2:24:12 shitty answer right curiosity we want
2:24:15 people that are rabidly curious we also
2:24:17 want people that are generous so not
2:24:19 just people that are curious and going
2:24:20 to go learnu and be all stingy with it
2:24:22 like ha I discovered the secret haha
2:24:24 [ __ ] you
2:24:26 [ __ ] no no we want people like oh hey
2:24:30 I was rapidly curious and I found this
2:24:32 thing and it does this you want to know
2:24:34 how I did it here's how I did it that
2:24:37 that oh that's really cool what did you
2:24:39 do to get the light to look like that oh
2:24:41 I used volumetric lighting and it just
2:24:43 did it oh cool I never knew
2:24:45 that that's who this group is already
2:24:47 exploration is one of our values like
2:24:49 people that are
2:24:51 choosing to be on this
2:24:54 adventure we don't know what's on the
2:24:56 other side of the Abyss we don't it's
2:24:58 scary yeah let's go
2:25:00 anyway collaboration we want people that
2:25:03 are you know willing and excited to to
2:25:05 say hey you want to do something
2:25:07 together I started this video series
2:25:09 today with Cindy [ __ ] and and an Murphy
2:25:12 who's just like they were just like hey
2:25:14 you want to do a cool thing together
2:25:16 okay you know they organized they put it
2:25:19 together there were other people there
2:25:21 like far
2:25:23 out inclusion people are coming from all
2:25:26 different professions and
2:25:29 backgrounds Embrace that because the
2:25:32 person that you think is the weirdo and
2:25:35 like oh well they a there a sanitation
2:25:38 worker in New York City yeah yeah that
2:25:40 sanitation worker is going to do
2:25:42 something with some tool that's going to
2:25:44 blow your [ __ ] mind because their
2:25:47 lens on the world is going to make them
2:25:49 look at that tool in a completely
2:25:50 different way than you would ever look
2:25:52 at it and this early in an industry when
2:25:55 things are this primordial you want as
2:25:57 many points of view as
2:25:59 possible next one is empathy people are
2:26:02 scared people are pissed
2:26:04 off like understand that have empathy
2:26:07 for people that they might not be in the
2:26:08 same place you are maybe you're super
2:26:10 excited and they're like yeah you're a
2:26:12 jerk okay why am I a jerk because you
2:26:14 like AI all right what is it about AI
2:26:16 right maybe you find out they don't
2:26:18 understand how it works maybe you find
2:26:20 out that that the reason they're really
2:26:21 scared is they just got laid off from
2:26:23 their [ __ ] job oh yeah that sucks
2:26:26 dude what did you do oh have you tried
2:26:29 any of this stuff no [ __ ] that I'm not
2:26:31 going to try that stuff I just got laid
2:26:32 off because of that [ __ ] oh that's cool
2:26:34 you want you want to see how it works
2:26:36 you want you want to see like what
2:26:37 you're up against no not really well let
2:26:40 me just show you a little thing okay
2:26:42 cool oh holy [ __ ] it can do that yeah it
2:26:44 can do that it can do other stuff too
2:26:45 you want to see that yeah maybe right
2:26:49 empathy empathy empathy empathy and then
2:26:52 the last one is
2:26:54 gratitude if people are being all these
2:26:56 other things like Curious and generous
2:26:58 and adventurous and collaborative and
2:27:01 inclusive and really like taking you in
2:27:03 as a human like the least you could do
2:27:05 is show a little appreciation
2:27:10 right we just added we just added
2:27:12 gratitude like two weeks ago we believe
2:27:15 gratitude is a practice gratitude is not
2:27:17 just a feeling but an action take the
2:27:20 time to acknowledge the efforts support
2:27:22 and successes of the extraordinary
2:27:23 people in this community especially if
2:27:26 someone inspires you go out of your way
2:27:28 to acknowledge them tag them in your
2:27:30 post reach out to them and let them know
2:27:32 they made a difference and no AI didn't
2:27:35 write that [ __ ] I wrote
2:27:43 that you're so bitter you're so
2:27:47 bitter and then here's what I want you
2:27:49 to do if you dig that if you dig what
2:27:52 we're about I want you to join and I
2:27:55 want you to go to the the first place I
2:27:57 want you to go is in the community
2:27:58 section to introduce yourself and I want
2:28:00 you to introduce
2:28:01 yourself danda gamble just joined and
2:28:04 introduced
2:28:06 herself that's
2:28:08 generosity 40,000
2:28:12 woohoo and then after you introduce
2:28:14 yourself go check things out there's a
2:28:15 water cooler there's a show andt tell
2:28:17 Channel where people are showing all
2:28:18 sorts of inspiring art there's a news
2:28:20 Channel there's AI tools and resources
2:28:23 this is a list if you're trying to
2:28:24 figure out what tools to start with I
2:28:26 personally curate this list you start
2:28:28 with chat GPT until further notice start
2:28:31 with chat GPT and yes you should pay the
2:28:33 20 bucks if you can afford it if you
2:28:35 can't [ __ ] it doesn't matter but you
2:28:37 know don't go to Starbucks three times
2:28:39 this month and there's your 20 bucks so
2:28:41 do
2:28:43 that
2:28:46 um and then there's a big long list of
2:28:48 tools here's eight other large language
2:28:50 models you can play with a lot of these
2:28:51 are free or have free options here's all
2:28:55 the image generation things that I think
2:28:56 are worth your while here's video
2:28:59 generation music and
2:29:02 audio 3D worlds and objects and then
2:29:05 interesting people to follow and then
2:29:06 directories like here's all the ones you
2:29:08 want all the tools there they are go get
2:29:10 them but you know these are the ones I
2:29:12 promise you if they're on this list it's
2:29:14 not a waste of your
2:29:15 time the other thing we've got here is
2:29:19 guilds so
2:29:21 AI 101 Guild and there's you know art
2:29:24 and writing and business and we've got
2:29:27 all sorts of exciting things planned for
2:29:29 the guilds coming up and these are
2:29:32 subcommunities of people that might be
2:29:34 into [ __ ] you're into and and you know
2:29:36 you can like like joy and Vicky just
2:29:38 started the ethics Guild so if you're
2:29:40 like wow we got to talk about the ethics
2:29:42 here okay great there's people that want
2:29:43 to do that go contribute to that
2:29:48 community and then there's clubs and so
2:29:51 I'll I'll highlight a couple of clubs
2:29:53 the Irregulars Club is us so the people
2:29:55 that show up to
2:29:57 this nightly they're called the
2:29:59 Irregulars and we have our own club here
2:30:01 and so we can like you know make inside
2:30:03 jokes and and share share images with
2:30:06 each other and and you know talk and and
2:30:08 sort of you know really explore what's
2:30:12 going on in
2:30:14 here in our own private space have our
2:30:17 own private chats um there's also a
2:30:20 group called weekend GPT jams uh hosted
2:30:24 by an Murphy so every Saturday from noon
2:30:26 to
2:30:28 2 you get to go learn how to build gpts
2:30:32 and you know learn from people that are
2:30:34 doing really cool [ __ ] with
2:30:37 them the one coming up this
2:30:43 Saturday I don't know oh I think this
2:30:45 Saturday is another Innovation
2:30:47 lab so some people from the community
2:30:50 are going to showing the cool stuff they
2:30:52 did um Jane and Leah have a cool prompt
2:30:55 generation thing where over the next six
2:30:58 weeks they're going to teach you how to
2:31:00 get creative with prompting like prompt
2:31:03 like a human human in the loop if you
2:31:06 haven't heard that so that's the AI
2:31:10 salon so go join go join the salon go
2:31:16 join the
2:31:17 salon and you know how we pronounce it
2:31:20 right it's the
2:31:22 Sal it's
2:31:24 French I don't know if it's
2:31:28 French but saying theong is
2:31:32 [Laughter]
2:31:36 [Applause]
2:31:38 funny how does Kyle prioritize his life
2:31:41 uh it just seems if it cracks him up he
2:31:43 seems to do
2:31:43 [Laughter]
2:31:47 it it's not it's not that far off
2:31:52 if I enjoy it if I'm passionate about it
2:31:54 I'm in if I'm not [ __ ] it I'm
2:31:57 out we're at 42,000 we we blew through
2:32:00 40,000 I knew that would happen all
2:32:03 right can you use the result legally yes
2:32:07 I don't know what result you're talking
2:32:08 about the output the output of large so
2:32:11 here's the here's the important thing to
2:32:13 know about these the G and GPT stands
2:32:16 for
2:32:17 generative it's not CP P PT copy paste
2:32:23 pre-trained Transformer model no it's
2:32:26 not copying and pasting it's not how
2:32:28 they
2:32:29 work the the the people that say chat
2:32:33 jpt is the world's largest the most
2:32:35 efficient plagiarism machine they don't
2:32:37 know how it works they don't understand
2:32:40 how it works they don't understand how
2:32:42 it works everything that's generated out
2:32:44 of this is
2:32:46 generated uniquely each time you know a
2:32:50 token at a time or you know a group of
2:32:51 tokens at a
2:32:53 time so yes you can use them um whether
2:32:56 or not you can actually copyright them
2:32:58 yourself that's a whole different thing
2:33:00 you've got to prove your human
2:33:01 contribution to the system and the
2:33:04 copyright office is in a [ __ ] show right
2:33:06 now and they're getting it wrong right
2:33:08 now in my
2:33:09 opinion um but I think copyright in
2:33:12 General's got to be completely rethought
2:33:13 because we're entering a world
2:33:17 where everything's going to be
2:33:19 personalized for every person is is
2:33:22 going to be an an option and so you're
2:33:24 not going to need to protect all that so
2:33:26 so I don't know I don't know where it's
2:33:28 going but you know if you want you could
2:33:32 go start a copyright and trademark Guild
2:33:34 in the salon and talk about this for the
2:33:36 next 20 years because the only people
2:33:38 excited about copyright and trademark
2:33:40 right now are the copyright lawyers who
2:33:42 are yacht shopping because they're like
2:33:44 oh people are going to be suing
2:33:46 themselves back and forth for decades
2:33:50 you want to get rich right now you can
2:33:53 make
2:33:54 money go get a law degree in in
2:33:57 intellectual
2:33:59 property and then just and then just sit
2:34:02 there collecting cash as both sides sue
2:34:05 the living [ __ ] out of each
2:34:07 other just flush the cash into your
2:34:12 yacht
2:34:17 bucket there you go there's your billion
2:34:21 idea oh Lordy all right show and tell
2:34:25 all right wait what am I supposed to
2:34:26 look at wait check check before filing
2:34:30 legal documents yeah
2:34:32 exactly so wait so what is all of GPT
2:34:36 and what are
2:34:39 gpts so so Kayla here's the deal funnel
2:34:43 fairy C couch or coach um come back
2:34:47 tomorrow and remind me let let's start
2:34:49 with that I'll start with the difference
2:34:51 between
2:34:53 gpts with GPT 4 chat GPT and gpts and I
2:34:58 I'll bring out the box if you haven't
2:35:00 seen the Box presentation oh it's good
2:35:03 so come back tomorrow and I'll start
2:35:04 with that just remind me to do that um
2:35:07 Corey wants me to see show and tell okay
2:35:09 so let's go Troy MAA oh that's
2:35:14 cool show me your birds of prey if
2:35:16 you're so inclined here are a few of
2:35:19 mine that's
2:35:26 [Music]
2:35:28 gorgeous very cool
2:35:33 oh look at that like like
2:35:37 that's let's see is this a song or is
2:35:40 this just an image I think it's just an
2:35:41 image it's just like that I feel like a
2:35:45 beer I mean imagine using
2:35:48 that throwing some typ graphy on top of
2:35:51 this as some sort of cool ad like look
2:35:53 how visually arresting that
2:35:55 is like you don't immediately go Oh
2:35:57 that's AI you write some clever copy on
2:36:00 that like actually and just like that I
2:36:03 feel like a beer like use that for a St
2:36:06 Patty's Day you know beer commercial
2:36:10 that's just [ __ ] gorgeous beer for
2:36:12 you absin for
2:36:17 me Mike knee that's gorgeous
2:36:20 gous oh that's a GPT creates visual
2:36:24 amalgamations so this is one of the cool
2:36:26 things you'll see here so what what Mike
2:36:29 put in here this is a a GPT so a GPT is
2:36:32 a customized version of chat GPT so you
2:36:36 can go play with this thing where he's
2:36:38 essentially pre- prompted it to make
2:36:40 super cool things what time do you start
2:36:43 your live Arizona time if you're
2:36:45 mountain time it's the same as me I
2:36:47 start them around 8:00
2:36:49 p.m.
2:36:51 um Corey Sandler
2:36:54 gorgeous H this is interesting to me GNA
2:36:57 explore this a bit more yeah the wood
2:36:59 carving thing so wood carving and
2:37:02 papercraft Corey are are always really
2:37:05 interesting to me I I'm I'm absolutely
2:37:08 enamored of images that that kind of
2:37:11 mimic [ __ ] that's done in the real world
2:37:14 that yeah things that have an analog
2:37:16 feel to them because I think it's going
2:37:18 to inspire artists to go make real world
2:37:22 versions of that stuff and and some of
2:37:24 the visuals are just amazing oh that's
2:37:28 funny Cory Sandler prompted three
2:37:32 times I wanted one image of asparagus
2:37:35 brussle Sprouts in an eggplant on a
2:37:37 cutting
2:37:41 board that's
2:37:43 hilarious grock
2:37:45 [Laughter]
2:37:47 chips oh man
2:37:50 all right let's go we'll take a let's
2:37:52 take a quick pass through a regular see
2:37:54 if we have anything fun in there
2:37:57 tonight Ian Murphy Oh register for the
2:38:00 weekend GPT
2:38:02 Jam so go do that the grock tensor
2:38:06 streaming processor Factory nice world's
2:38:10 not ready for gpts game
2:38:16 over I should be making space shuttl is
2:38:20 stuff well I've done that oh that's
2:38:21 really
2:38:24 cool
2:38:31 huh that's
2:38:34 great no excuses AI makes everything
2:38:37 possible yep that's there's your message
2:38:40 we'll leave we'll leave on that
2:38:42 message thank you Silver Fox thank you
2:38:44 so
2:38:45 much kind of late for
2:38:48 gpts yeah exactly exactly it's kind of
2:38:50 late for that maybe go to ask for help
2:38:52 yeah you can do that all right so let me
2:38:55 say goodbye to everyone all right if you
2:38:57 like what you saw here if you want to do
2:38:59 more of this keep coming back to this
2:39:01 channel I'm here
2:39:03 nightly um follow my channel you know
2:39:06 follow the AI learning lab um you can
2:39:10 subscribe to the lives that'll support
2:39:12 me you can buy the video series that's
2:39:15 in the bottom corners
2:39:17 there um and you can go to the AI salon
2:39:19 join the AI
2:39:21 Salon but join sort of embracing those
2:39:25 values cuz we want to keep it as special
2:39:28 a place as it is I'm telling you I mark
2:39:31 my
2:39:33 words there are going to be companies
2:39:36 spawned out of the AI Salon there are
2:39:38 going to be success stories spawned out
2:39:40 of the AI Salon that are going to be
2:39:42 eyew watering I promise you I we are
2:39:45 moving probably within the next 18
2:39:48 months 18 months to 3 years
2:39:52 where why people get hired is going to
2:39:55 shift from what they can do their skills
2:39:59 to who they are and how they're
2:40:02 trusted let me say that again
2:40:06 wait let me let the Tik tok's going to
2:40:08 get glitchy for a minute
2:40:12 here the next 18 months to three
2:40:16 years there's going to be a shift
2:40:20 in why people get
2:40:24 hired from the skills you have what you
2:40:28 can
2:40:29 do to who you are and how much people
2:40:33 trust
2:40:37 you I think there's a reckoning for
2:40:39 [ __ ] coming you know people that are
2:40:42 super talented but they're really
2:40:43 [ __ ] hard to work
2:40:45 with guess what a lot of other people
2:40:48 are going to get way more more
2:40:50 talented and The Talented
2:40:53 [ __ ] are going to stop getting phone
2:40:55 calls and the people who are trusted and
2:40:58 engaged and in community are going to
2:41:00 get jobs mark my
2:41:03 words joining a group like the AI Salon
2:41:06 is a career
2:41:07 move promise you may not seem like it
2:41:10 now it's going to seem like well it's
2:41:12 just people they're not very
2:41:13 professional they're just playing and
2:41:14 they're do and then they make funny
2:41:16 images and then they do stupid [ __ ] and
2:41:19 yeah yep you
2:41:25 watch I'm watching people in there right
2:41:27 now that I'm like okay these are these
2:41:29 people are going to make you
2:41:32 know 10 20 100 200$ 500 million
2:41:38 companies because they're [ __ ] driven
2:41:40 they're just like oh AI is
2:41:42 transformative I'm inspired [ __ ] it I'm
2:41:44 all in I'm
2:41:47 going it's it's [ __ ] Wild
2:41:50 and like the tools aren't even good
2:41:54 yet so get your ass in community and I
2:41:57 don't give a [ __ ] if you join the AI
2:41:58 Salon or some other one but if you want
2:42:00 to be AI literate I'm I am convinced
2:42:04 more every
2:42:05 day that who you're doing it with is as
2:42:10 important as what you're learning
2:42:12 because the skills are irrelevant
2:42:13 because everyone will have
2:42:17 them you ain't so special
2:42:22 anymore the only way you're special is
2:42:24 if someone else thinks you're
2:42:26 special and they would hire you because
2:42:29 of how special you are that's how you're
2:42:31 going to get
2:42:33 hired so if
2:42:35 you're isolating yourself if you're an
2:42:40 [ __ ] go buy some books on
2:42:44 empathy and [ __ ] buckle up
2:42:48 buttercup
2:42:51 cuz you're gonna have to start
2:42:55 connecting so anyway all right enough
2:42:59 Soap Box Rachel Sparkles good to see you
2:43:03 and multi-talented exactly a human being
2:43:05 Ron Diggity great to see you this woman
2:43:08 Works awesome fantastic thanks for being
2:43:11 a subscriber Joker here good good comedy
2:43:15 tonight good solid solid once again the
2:43:18 lines being delivered
2:43:20 um yes it will happen there you go
2:43:23 Elizabeth gem good to see you Source
2:43:25 Camp thank you Cory Sandler Pottery I
2:43:28 hope so of course
2:43:30 Corey cor Corey's gonna be a software
2:43:33 magnet she's gonna have materials in the
2:43:36 space station she's going to be a
2:43:39 business mag magnet and she's going to
2:43:42 go oh and I still sell
2:43:45 pottery that's going to be
2:43:48 awesome Silver Fox good to see you Janu
2:43:52 Tom uh laughing Thunder great to see
2:43:54 youve been seeing you a lot around here
2:43:56 lately uh Jane Doe um good luck with the
2:43:58 hubby hey tell tell hubby you spent some
2:44:01 time tonight learning about AI tell him
2:44:02 he was right Google's better then just
2:44:06 just keep learning it just show up a
2:44:08 month from now and just blow his
2:44:11 mind Source Camp I'm going to the woods
2:44:13 tomorrow to work for a week it's getting
2:44:15 real nice oh I like it that sounds great
2:44:19 that's sounds great I can't wait to hear
2:44:20 what you come up with why don't we
2:44:22 create our own entity um there's
2:44:25 actually riggity there's a bunch of that
2:44:26 going on in the AI salon right now so I
2:44:29 would say just get involved in the salon
2:44:31 there's a lot of why don't we create
2:44:32 this why don't we create that going on
2:44:34 right now there's some really cool
2:44:36 startup energy and not startup like
2:44:38 we're having a hackathon and the tech
2:44:41 Bros are getting together startup like
2:44:43 why are we doing this together what are
2:44:45 we doing together how are we doing it
2:44:47 like there's really cool stuff going on
2:44:49 that it almost doesn't matter what
2:44:51 they're building it's how they're coming
2:44:54 together I I I think there's some some
2:44:56 good stuff
2:44:58 happening all right uh I'm out of here
2:45:01 people it's
2:45:03 late my day starts early tomorrow I will
2:45:05 see you tomorrow night um I should be on
2:45:09 time it's Thursday 50 minute hellos and
2:45:11 goodbyes shut up shut
2:45:14 up codependence it's not just a river in
2:45:17 E Egypt all right every buddy peace
2:45:24 out