AI Learning Lab

Feb 4, 2024 - AI Learning Lab: Unveiling the Future of Creative AI with Kyle Shannon

635ASLhN6tg
Video2024-02-112:17:245 views

Description

In the latest episode of the AI Learning Lab, Kyle Shannon explores the intersection of creativity and artificial intelligence, emphasizing the importance of playfulness and curiosity in mastering AI tools. He shares insights on how to cultivate AI literacy through experimentation, application, and sharing knowledge within a supportive community. The discussion highlights the transformative potential of AI in creative fields, encouraging individuals to embrace their unique artistic voices without fear of judgment or apology. Kyle also reflects on the recent Grammy performances, drawing parallels between artistic expression and the evolving landscape of AI-generated content. This engaging dialogue invites viewers to rethink their relationship with technology and creativity, fostering a culture of collaboration and exploration. For more thought-provoking discussions, check out the AI Learning Lab on TikTok: [AI Learning Lab](https://tiktok.com/@aiLearningLab). #AI #Creativity #Art #Community #Learning #Innovation #Technology #AIArt #aicommunity Chapters: 00:00:00 Grammys Recap 00:12:00 Value of AI-Generated Art 00:15:00 Different AI 00:23:55 Sherry's Runway ML Project 00:31:50 Runway ML Tutorial 00:46:14 Introduction to AI Salon 00:51:55 Learning Chat GPT 00:54:24 Suno AI Music 01:04:57 Google Gemini Launch 01:08:02 Chatbot Arena Rankings 01:15:15 Deepfake Scams 01:26:30 Importance of Spatial Computing 01:32:13 Skybox Blockade Labs Demo 01:36:20 The Future of AI 01:41:19 Google's AGI Levels 01:58:12 Power Loom Riots 02:02:04 Essential AI Tools 02:07:02 The Uniqueness of Pi 02:11:51 Ways to 02:13:16 Storytelling with AI

Chapters

Transcript

0:05 [Music]
0:16 stand
0:18 between you hard places
0:21 [Music]
0:25 inane to
0:28 me you and
0:31 makes it
0:32 [Music]
0:35 clear your trouble to
0:41 me Rich trouble can't you
0:45 [Music]
0:48 see leaning in
0:51 clothes smell of your perfume scares me
0:58 most leaning away
1:02 it feels father every
1:08 day you're trouble to
1:14 me real trouble can't you
1:18 [Music]
1:29 see
1:30 [Music]
1:43 hey no wrapping no
1:45 [Music]
1:59 wrapping
2:02 [Music]
2:16 yes I said no
2:19 [Music]
2:23 rapping we're singing Americana stuff
2:26 here you got to sing it you got to tell
2:28 the story you can't just bark it
2:31 out I feel bad playing this level of
2:35 music with the Grammys on but you know
2:38 God dang did you see the first three
2:41 performances Tracy Chapman with Luke
2:44 Luke
2:46 com
2:47 um
2:49 Miley
2:51 and who's the who's the third one anyway
2:54 it was amazing no cuts to Taylor Swift I
2:59 know
3:02 [Music]
3:14 yeah you should watch Corey you should
3:15 watch the beginning watch the first
3:17 three performances the Grammys cuz cuz
3:20 Tracy Chapman singing with Luke Luke col
3:22 was just
3:23 beautiful and then who was the second
3:25 one I know Miley went was the third
3:29 one
3:36 this is still better than the Grammys
3:38 yeah you tell them Source
3:40 [Music]
3:49 camp Oh Billy ish that's who it
3:53 was like it was a very emotional version
3:56 of the Barbie song which is an emotional
3:58 song anyway but just beautiful
4:00 with Phineas at an upright
4:03 [Music]
4:06 piano so I would assume that she's going
4:08 to take best song for that
4:11 [Music]
4:15 thing and Taylor already won best pop
4:18 album so I've got it on in the
4:21 background here so if anything big
4:23 happens I'll or if anything big happens
4:25 and I'm rambling on about AI just let me
4:28 know um
4:31 they had one AI joke at the beginning
4:33 Trevor Noah had one AI
4:35 [Music]
4:59 joke you want to sing more
5:02 [Music]
5:13 champy you and I
5:17 on Sunday morning here at
5:22 home sky blue and a C strong it's
5:28 true
5:31 then I open my eyes to a dream realized
5:35 in front of
5:37 me Vicky champy sounds worn
5:41 out he's a little he's a little crispy
5:44 that
5:47 champy they're making fun of you chy
5:49 don't put up with that [ __ ] you just you
5:51 know what you tell them just stop being
5:54 irregular he'll sing when you stop being
5:56 irregular Jay-Z just won
5:58 something
6:02 I do not know
6:05 what all right let's get going here
6:08 people welcome welcome welcome my name
6:09 is Kyle Shannon this is the AI learning
6:11 lab this is the uh this may seem like a
6:14 humble um Tik Tok Channel and this might
6:18 seem like a shitty home office but this
6:20 is a $20 million Video Production Studio
6:24 and that's the AI learning lab private
6:25 jet so anything you give here just it it
6:29 just the the amount of I'll tell you
6:31 what you know what this channel really
6:32 is you make
6:35 money yeah
6:42 exactly oh man so we got some got
6:47 some some good uh some some interesting
6:51 stuff from from last night to
6:54 today so last night we went and played
6:56 with a bunch of tools bunch of image
6:58 tools um
7:00 today so this image was made by Corey
7:04 Sandler and today what did he oh he the
7:07 Dr Dre Global impact award yeah that
7:08 makes sense Jay-Z would win that all
7:10 right cool
7:12 um Corey Sandler was
7:15 live and she was just walking through
7:18 she got
7:20 inspired I think who was it steo who I
7:23 just saw in here steo posted an image on
7:26 Show and Tell in the AI salon and then
7:29 it ran Corey down a rabbit hole where
7:31 she just started with an image he made
7:33 and just and and she walked through her
7:35 whole process it was so
7:38 cool just
7:42 hearing how someone else does it right
7:46 like like what she did like
7:51 tactically was what we all do you type
7:53 in some words you're like N I don't like
7:56 that but like what blew me away about it
7:59 was
8:00 she got to images that if I got to those
8:03 images I stop I just stop I'm I'm like
8:06 oh that's brilliant let me go share that
8:08 and then I'm
8:09 done and she got to This brilliant [ __ ]
8:12 and then she goes well now just make it
8:15 a girl make it a boy make it old make it
8:17 young make it Angry make it sad and she
8:19 just kept digging and digging and and
8:21 then at one point it went way away from
8:23 what she wanted then she got it she keep
8:24 brought it
8:26 back and it was just a reminder to me
8:30 that a sharing what you know sharing
8:33 what you're learning just sharing how
8:35 you do
8:36 AI as obvious as you think it is and
8:41 that was one of the things you know I
8:42 hopeful I hope I got through to her a
8:44 bit at one point she was kind of
8:45 apologizing for not really knowing what
8:46 she was doing I'm like no no you know
8:49 what you're doing like like you're doing
8:52 it
8:54 um and and it's listen it's something we
8:58 got to work on people Irregulars and all
9:01 um something we got to work
9:04 on is I want us to all so so okay so
9:08 here's the thing if you're new here this
9:10 this is the this is the basic here's the
9:13 trick for get getting AI literate you
9:15 play with it you play with it a lot you
9:17 get really curious and you just play and
9:20 then you apply it you just figure out
9:21 something you want to do with it and you
9:23 do it more and like you apply it towards
9:25 something you solve a problem or you
9:27 come up with an art project you apply it
9:30 and then the third step is you
9:32 share you share what you're learning you
9:34 share what worked you share what didn't
9:36 work why because everyone's trying to
9:40 figure this out right now and it will
9:43 help you get better at what you're doing
9:46 by articulating what you're doing and it
9:48 will help other people who think this is
9:51 [ __ ] rocket science or jet
9:55 design to to maybe have the courage to
9:58 try it and then I want to add something
10:00 to it I want us to encourage each other
10:04 to stop
10:07 apologizing stop apologizing for not
10:10 being an artist stop apologizing for not
10:12 what knowing what you're doing about AI
10:14 stop apologizing about oh it didn't take
10:17 me that long so it doesn't have that
10:19 much value [ __ ] that it
10:23 does it
10:26 does just Corey sharing that she got to
10:30 this remarkable work and kept going like
10:33 10 levels
10:35 deeper was revelatory to
10:39 me and to her that's just how she works
10:42 right so the things that we think are
10:44 common sense are not
10:49 common thank you for saying this it's
10:51 such a great
10:53 reminder yeah this is this is part of my
10:56 you know I talked about my my you know
11:00 declaring my new who I am stuff and I
11:02 you know I said I'm going to leave shame
11:04 at the
11:08 door let's all do
11:11 that [ __ ]
11:13 this there's nothing to apologize for
11:16 there's nothing to apologize for for
11:18 using AI
11:20 tools they're absolutely art if you're
11:23 if you're making if you're
11:26 making if you're creating things with AI
11:29 with the intention of an artist with
11:31 artistic intention it's [ __ ]
11:36 art and if you generate something and it
11:40 only took you 3
11:42 minutes and in your mind it should have
11:44 taken 3
11:47 hours let's Let It
11:51 Go that spending more time on something
11:54 makes it more
11:58 valuable time equals money that's that's
12:01 what the the industrialists said at the
12:04 beginning of the Industrial
12:07 Revolution work work harder work more
12:10 work more hours so I think we have we
12:13 have we have
12:14 deep-seated apologies I missed what
12:16 Corey did what did you
12:18 do she well she she what she did and was
12:23 both non- remarkable and remarkable the
12:26 non- remarkable part was she she was
12:28 just showing like here's what I did last
12:30 night I got inspired by steo and show
12:32 and tell in the AI Salon he made this
12:34 cool image it inspired me and you know
12:37 she said she was up till like 2: in the
12:38 morning making
12:40 images but what she did on her live was
12:43 she just walked through she she had her
12:46 whole prompt history there for what she
12:47 did so she showed the image that she
12:50 started with with him and then she
12:52 started just running down these creative
12:54 rabbit holes and and and she would sort
12:56 of run out of where she wanted and back
12:58 into where she wanted and then she got
12:59 to these images like that one really
13:01 speaks to me and she just kept
13:05 going and I just I I enough enough
13:09 apologizing we don't have to apologize
13:12 for using AI we don't have to apologize
13:14 for not knowing what we're doing we
13:15 don't have to apologize for not putting
13:18 in enough
13:22 time so
13:24 yeah so and here's the thing I'd like us
13:27 to hold start holding other to account
13:30 and and it's been happening a little bit
13:32 on on your weekend GPT jams I've
13:35 noticed that sometimes the people that
13:37 are there speaking who are brought in as
13:40 experts because they're expert at
13:42 something apologize for not being a good
13:45 enough
13:49 expert um Kyle and the Gang going to
13:52 Grammys and show inel awesome okay great
13:56 uh that's awesome um and and I'll show
13:59 you what what
14:02 um what's going on oh that's really cool
14:05 steo oh you know what's funny steo so I
14:08 showed I showed Gabby you did a a mini
14:11 fig of me and champy I showed it to
14:14 Gabby and she thought it was real she
14:15 goes oh can he send us one of those it's
14:18 like it doesn't actually exist she goes
14:20 can we frame it so so we might have one
14:23 of your pieces of artwork hanging in our
14:25 house here
14:28 steo
14:31 um yeah yeah
14:34 yeah got thinking about I know nothing
14:37 means I have no B boundaries and can fly
14:40 yeah that was the that was the the
14:41 phrase you said uh Corey something
14:43 around I know nothing or whatever it was
14:46 but but it was just like so I'm glad
14:48 you're in here um okay great thanks um
14:51 steo said he'd be honored if I if I hung
14:53 his his Lego picture printed out his
14:55 Lego picture all right we'll do that but
14:58 no I'm glad you're in here Cory it was
15:00 just I it was really a pleasure just
15:02 just just watching you do what you did
15:04 and just just you
15:09 know just expressing yourself and but
15:12 but in in in your life you weren't you
15:16 weren't you know expressing yourself as
15:17 an artist you were just walking through
15:19 here's how I did what I did and here's
15:21 what lit me up and here's what didn't
15:22 light me up and then I kept going I kept
15:24 digging and I kept doing this and I kept
15:26 going
15:28 that
15:29 and it just like what hit me
15:34 was we we all have such different
15:37 working Styles right was this Wizardry
15:41 recorded um it probably was I don't know
15:44 if Corey so so Corey if you don't know
15:46 it in Tik Tock if you go to
15:49 your profile then click in the upper
15:53 right and then go to Creator tools and
15:55 go to live and go to
15:57 replays um
15:59 it Tik Tok saves the last 30 days of
16:02 your lives I don't know if you've ever
16:03 downloaded that but that and you might
16:07 do that a lot but I just I thought it
16:09 was a really good one so anyway it might
16:10 might be worth throwing up on YouTube
16:12 Somewhere so it it's got um some
16:15 longevity to it because it was just it
16:16 was just really good it was it was just
16:18 very simple it was a very
16:21 simple you know just good reminder that
16:24 we all work in different ways and we all
16:26 bring different things to this right
16:27 that idea of inclusion
16:30 um because the thing that's obvious to
16:33 you is revelatory to someone else I just
16:35 I keep running into it over and over and
16:37 over
16:38 again
16:40 so
16:42 anyway so thank you for that thank you
16:45 thank you thank
16:46 you
16:54 um wait what oh wait wait what shared
16:58 the live thank thank you wait what
17:00 that's a good name wait
17:04 what uh if you would be so kind share
17:08 the live um let's get some more people
17:11 in here share it particularly with
17:13 people who might be skeptical about AI
17:16 or have questions about AI or are pissed
17:18 off about AI or scared of AI or are sick
17:21 of you talking about
17:24 AI let's get them in
17:26 here all right
17:34 Jay
17:36 Z yeah it's been pretty good the the U2
17:39 the U2 performance in the globe was
17:42 pretty cool the sphere the sphere it's
17:45 called um all right so we'll talk we'll
17:48 talk a little bit about the AI Salon in
17:50 a second but let me uh let me
17:53 just see what everyone's saying
17:57 here
18:01 oh wow Sher D 12 hours left this is my
18:05 rough draft for the submission of the 48
18:08 hour contest for Runway thanks to
18:11 Danielle and Leah mad lives throughout
18:13 the months I was able to complete this
18:15 challenge there are the rules cannot
18:18 express the amount of gratitude for the
18:19 as salon and the Irregulars tears of joy
18:22 I am doing finishing touches but had to
18:24 share Sherry so Sherry I assume you're
18:26 live here hey hey there
18:29 um do you mind if we show it do you want
18:32 do you want to do do you want to do a
18:34 live a live premiere of your rough cut
18:38 if if you do I'll play it but if not I
18:40 don't want to play it if you don't want
18:41 to show something that's not where you
18:43 want it to be do it okay cool all right
18:46 cool so so here's what we're going to do
18:48 we're going to show so so Sher D is one
18:51 of the
18:52 Irregulars um has been coming here for
18:55 months about a month ago or something
18:58 like like that one of one of the things
19:00 that we talk about in here is talk about
19:02 what you're doing with AI so she was
19:04 talking about what she was doing with AI
19:06 she's gotten really good at the image
19:08 stuff and someone she went to high
19:11 school with has got a band and reached
19:13 out to her and said hey will you make a
19:14 video for us using Ai and so so she is
19:18 now a a music video producer um and then
19:22 so she just signed up for Runway ml's 48
19:26 hour video contest so we're going to
19:28 look at rough cut together here on Ye
19:31 Old Tik Tok this is
19:33 exciting um so thank you for sharing
19:35 that and allowing us to to put it here
19:38 that's pretty cool all
19:44 right all right let's go over to
19:48 YouTube here we are who just won
19:55 what Billy Isis just won for
20:05 hang on hang on
20:06 Sherry such a fan of Billy
20:14 ishes it's such a great song that song
20:17 is like
20:20 haunting hang on hold please
20:27 wait
20:37 y my goodness damn that's song of the
20:41 year who literally like I was looking at
20:44 everybody's faces and I lean over and I
20:46 was like not a chance yo I just want to
20:49 say everybody in this category
20:51 like that was a crazy list of incredible
20:56 people incredible artists she's so cool
21:03 she's so cool all
21:05 right you know one of the things I I
21:09 there's there's something about her
21:12 that there's something about Billy
21:15 eish
21:21 that she's got such a hang on let me let
21:24 me I'll I'm going to come back to you
21:26 Sher dve we're going to play this but
21:27 but let me let me just talk
21:29 there there's something about
21:32 her that she's got such a unique point
21:35 of view
21:38 creatively and and you know I'll give
21:40 I'll give her brother credit too he's
21:41 he's an amazing engineer and and a
21:44 songwriter but there's something about
21:46 her her particular point of view that
21:49 it's that it's so specific that you know
21:53 she's a true artist and then true
21:56 artists sometimes they hit and sometimes
22:00 they
22:03 don't and she was you know right place
22:06 right time right circumstances right
22:08 Talent all that right point of view that
22:10 she stood out in the sea of whatever
22:12 musical noise was happening and she Rose
22:14 above
22:16 it
22:18 that's her her uniqueness to me feels
22:22 like what it's going to take to rise
22:24 above the noise with AI that there are
22:26 going to be artists and and they're not
22:30 necessarily going to be artists that
22:32 started out as artists like the thing
22:34 we're about to watch with Sher D she's
22:37 got 20 30 years in
22:41 nursing and she's leaning into this AI
22:43 stuff and she's she's starting to find
22:45 an artistic voice for
22:47 herself and so the thing that I that I'm
22:52 just I can't wait to see and and we're
22:55 not quite there yet because not enough
22:57 people know about AI the tools are not
22:59 quite there but there's going to be an
23:02 entire new class of creatives that come
23:04 and and there's just going to be tons of
23:06 work that comes out and it's just going
23:08 to be all this noise and it's going to
23:10 be super slick like the the floor of
23:12 creativity is going to rise right but
23:15 the base is still going to be the base
23:18 and then out of that soup someone is
23:20 going to rise up out of it and we don't
23:23 know who it's going to be and I just I
23:25 think about Billy ish because she rose
23:27 up out of you know this sea
23:31 of kind of Pop sameness with this really
23:34 unique voice and I think I don't think
23:37 any of that changes with
23:39 AI I I still think human beings
23:42 expressing themselves with a very
23:44 particular point of view remains
23:47 critical and remains increasingly
23:52 valuable so anyway all right so so let's
23:55 watch so this is the rough cut of Sher
23:58 D's Runway ml
24:02 project which she's got 12 hours left so
24:04 we're not going to give you notes Sher
24:06 we're just going to watch it um she's
24:07 got 12 hours left so so whatever she
24:09 thinks is isn't isn't here for this um
24:12 she's gonna she's going to do up and
24:14 then she's going to submit it so here's
24:16 the other thing AI salon and Irregulars
24:19 we got to support the [ __ ] out of this I
24:21 don't know how Runway is doing this but
24:22 but let's let's tuck this thing up so
24:26 all right so here we go so let me
24:28 get my my video
24:33 [Music]
24:50 right wait sorry pause pause for one
24:53 more second Joanie Mitchell's about to
24:55 sing for the first time in the history
24:57 of the Grammys
25:00 and Brandy Carlile introduced
25:10 [Music]
25:27 her
25:30 [Music]
25:40 [Music]
25:50 [Applause]
25:56 she's still so beautiful
26:00 [Music]
26:07 yeah yeah I know I can't I can't leave
26:09 this on too long or I'll get Shadow
26:11 banned anyway
26:16 okay all right that'll be all over Tik
26:18 Tok so we'll be able to see that um okay
26:21 um so let's watch Sherry's thing sorry
26:22 Sherry for the delays but you know
26:24 Grammys and all um okay dance with a
26:32 bear a film by Sherry
26:34 D Rich lantto Don Chappelle and
26:41 Gustavo Baldo I assume that's the the
26:45 band that did the music
26:48 [Music]
26:57 maybe
27:00 uhoh ch's going to sing I think it's
27:03 only appropriate this is
27:06 great I
27:13 [Music]
27:17 know
27:19 of
27:20 [Music]
27:23 right but they
27:25 cross family
27:30 [Music]
27:32 Lord better know the r for a fight
27:37 that's a good
27:38 [Music]
27:44 song shall
27:45 [Music]
27:49 I
27:51 no
27:53 now will
27:57 Li
28:00 oh oh a night for a
28:05 night they going to
28:08 dance all the night with the bed wow
28:14 this is great going to
28:17 dance
28:19 like
28:23 on the going to dance and till no really
28:29 [Music]
28:33 care so we embrace the night just B
28:41 me
28:43 synchronize light show with
28:48 shadows
28:49 [Music]
28:53 falling
28:56 for
29:03 no arm for
29:06 [Music]
29:18 me
29:20 [Music]
29:24 wow you
29:26 know
29:28 no will
29:33 I
29:36 out oh life for a
29:41 life going to
29:44 dance all night with the
29:50 bed going to
29:53 dance like
29:55 no on day
30:00 I going to dance until no really
30:09 cares so we embrace the light just
30:14 [Music]
30:17 the
30:19 syron light
30:21 show
30:23 [Music]
30:26 Shadow damn
30:30 damn damn
30:32 damn that's your band oh that's so cool
30:36 so cool um by the way I we just I just
30:39 got a call from Champ's manager Sherry
30:42 uh Champ's going to require credit at
30:45 this point sorry he's in a live
30:47 performance it it this is not me this is
30:51 It's the management company um damn that
30:53 was so good that was so good what did
30:56 what did you use for the the images is
30:58 is it um Mid Journey you bet Champion
31:02 you got credit
31:03 [Laughter]
31:06 buddy yeah that's really really good
31:09 really really good like powerful
31:11 song powerful song and then just you
31:14 know really nice creative choices to
31:16 have everything be slow and just sort of
31:18 lean into that which I think that's
31:20 actually a really smart tactic because
31:23 the more movement you put in these
31:25 things the the more they just distort
31:27 and go [ __ ] Haywire so I think it's
31:30 actually a really good really
31:32 good tactical Choice as well as creative
31:36 Choice um also when you have messages
31:39 with typos how do you
31:41 delete or edit
31:45 them Laura made in Leonard oh okay oh a
31:48 Laura made in Leonardo okay so she used
31:51 a specific model in in Leonardo so
31:54 Leonardo is based on top of stable
31:56 diffusion
32:00 the but wait where did you make oh you
32:02 made the video in
32:03 Runway so she okay so she used mid
32:07 Journey for some images Leonardo for
32:10 others so that's stable diffusion so
32:11 it's a mix of stable diffusion and um
32:15 Mid journey to make the initial images
32:17 and then Runway to do the video and then
32:19 her band for the for the music which is
32:21 really good that's a really good song
32:23 really really good song congratulations
32:24 that's that's that's super cool so okay
32:27 so as soon as that's up and out you know
32:30 you're going to have to let us know and
32:31 then we'll boost the [ __ ] out of it and
32:33 do what we can do so so pay attention to
32:36 the uh the runway gen 48 I think it's
32:40 called or something like
32:43 that all
32:45 right what is Runway so run if you go to
32:48 Runway ml.com
32:51 it is a tool that lets you make videos
32:56 out of images
32:58 so where's my sidebar is that it
33:02 yeah and where's
33:10 [Music]
33:13 my I
33:17 lost I just go to a new one let's see we
33:21 go
33:22 here Runway
33:26 ml
33:28 [Music]
33:31 the Mad lives helped oh cool there you
33:33 go uh I don't think Danielle can join us
33:37 tonight cuz she's in Southern California
33:39 and they are getting absolutely hammered
33:41 so it doesn't look like she has much
33:42 internet access but that's super cool
33:46 super cool congratulations Sher it's
33:48 really really really good work um so
33:52 this is Runway so so if I say start with
33:54 an
33:56 image I could could run over let me just
33:58 HP over to the AI salon for a
34:04 second oh you know what I'll do no hang
34:07 on let me go back to
34:09 Runway I just saved that that jet that I
34:12 had up there earlier let's let's put
34:14 that jet
34:17 there um did I download
34:26 that
34:30 oh wait it's in the wrong file format
34:32 dang
34:45 it all right is that a preview where
34:49 where is
34:56 that okay
34:58 here's this we're going to export this
35:00 as a PNG or jpeg
35:03 whatever and then we'll go back to
35:04 Runway okay so here I am on Runway so
35:06 I'm going to grab that
35:14 image oh you're listening hey Danielle
35:16 please please be safe I'm sorry it looks
35:19 like the uh the rain in Cali is is
35:22 [ __ ] crazy for the for the next day
35:24 or two so please stay safe so so what
35:28 you can do here you can do an image plus
35:30 a description so we could say something
35:32 like that but let's we'll sort
35:35 of we'll go up a little we'll go right a
35:41 little we'll roll a
35:44 little and we'll
35:48 Zoom we'll zoom in a little okay so so
35:52 we can do that and then they've also got
35:55 this thing called motion brush which I
35:57 don't I don't know if you've seen
35:59 that but we
36:04 can brush
36:07 in like all these clouds and it will
36:11 oops can I undo that yes and it will
36:18 just make the clouds
36:26 move
36:30 I I have that Su stuck Su stuck in my
36:33 head
36:36 now so we'll have those things
36:42 move to the right and then what's cool
36:46 about Runway now what they've done is
36:49 you you can have up to five brushes so I
36:52 can brush in the clouds as one set of
36:54 movement and then I can pick brush too
36:57 and I can put the
37:00 jet as another set of
37:11 movement and I don't know I haven't
37:13 played a ton with this so this might not
37:16 be good because I don't know what I'm
37:18 doing which again like I said that
37:22 according to my earlier Declaration of
37:25 we should stop [ __ ] apolog oling for
37:28 the fact that we're not expert at stuff
37:29 I'm going to go on ahead and not
37:31 apologize for that even though I just
37:33 did um so we're going to try this here
37:37 and then we're all going to learn
37:39 together and there's nothing to [ __ ]
37:41 apologize
37:42 for right right
37:48 [Music]
37:56 people all right so this we're going to
37:58 go
38:00 proximity
38:04 closer I don't know what ambient is but
38:06 it's probably going to screw some stuff
38:08 up and then horizontal will make it go
38:11 left and we'll save Okay so we've now
38:14 got the clouds going one way the jet
38:16 hopefully going the other
38:19 way how did she slowly break the bottle
38:22 that's what I want to
38:24 know brush moves only raise by one or
38:29 two oh I
38:31 see I should have made the plane INF
38:34 fuchsia
38:35 absolutely all right so now we now we
38:37 hit generate so so this is what Runway
38:40 is Runway is
38:42 a video generation tool that will take a
38:46 still image or a text prompt or another
38:50 video and turn it into a video and
38:54 there's there's other ones there's ones
38:56 called called p
38:57 CBS there's another one called forget
39:01 what it's called Pika and Runway are the
39:04 biggies right now but you can also do
39:06 video now on
39:09 um Leonardo now does video um Deco here
39:15 now does video um I think Korea might do
39:25 video
39:34 yeah really really great job Sher D
39:36 that's that's super
39:37 cool that's super cool so we're at we're
39:40 at
39:40 68% now here's so so the video that we
39:44 just watched that Sherry put together
39:45 with all the all those cool you know
39:47 scenes and cuts and men and bears and
39:51 all that sort of stuff um Runway only
39:55 makes 4 second videos
39:57 is it frozen for anyone
40:00 else hello can you hear me can you hear
40:03 me Lori Ryan no jump out and back in all
40:08 right so here's our here's our video
40:09 let's let's see let's see how bad it
40:12 goes I I have not had good luck with
40:14 Runway but let's see what we did
40:19 here well it's actually kind of cool
40:22 except I missed a tail apparently or it
40:25 missed a tail
40:38 so here's what I want to do I want to
40:41 do I want to go back into motion brush
40:44 and kill the motion
40:49 brushes can I double click this to
40:51 delete it clear and I'm going to go to
40:55 brush two okay so we cleared everything
40:58 okay so I'm going to clear the motion
41:02 brushes and then I'm just going to have
41:04 the whatever camera I had in there
41:06 whatever camera movements I had in there
41:09 and now I'm going to generate it again
41:11 so so we'll generate this again and this
41:12 is this is really what you have to do
41:14 with this stuff so just like you
41:15 experiment with um just like you
41:18 experiment with
41:22 um you know image generation over and
41:25 over and over again until you start to
41:27 you know start to figure out some of the
41:29 tricks of it it it's kind of
41:32 Amplified you know times 100 with video
41:35 because it can just distort things into
41:37 unusable
41:39 video
41:44 [Music]
41:45 um all right and let's see what's it
41:49 doing there okay so that's doing its
41:51 thing um the other thing that I want to
41:56 we we'll do it we'll do a couple of
41:57 things tonight we can go look at um we
42:00 can go look at some chat GPT stuff I
42:03 want to talk about the Vision
42:04 Pro um cuz that's now out and then
42:08 there's some some potential news coming
42:10 this week from Google that I want to
42:13 talk about so yeah some good big stuff
42:18 there all right so hey Tobias what's
42:20 shaking good to see you all right so
42:23 here's the one without doing the motion
42:25 brush
42:27 it's a little better it's still at the
42:29 words on the side of the
42:36 plane clearly coming in for a
42:41 landing it almost maintains its
42:44 shape actually so so here's here's what
42:47 what is is sometimes interesting so
42:49 let's take
42:51 this let's go to a different site we'll
42:54 go to a site called pika
42:57 p.
43:00 art
43:03 and we're going to upload that same
43:13 image and then we'll go to our camera
43:16 controls we'll
43:18 do
43:21 rotate if we rotate left it'll make the
43:24 plane go down so let's rotate right
43:27 so we're going to rotate right we're
43:29 going to zoom
43:31 out we'll do two strength of motion and
43:35 we're going to pan right and then pan
43:38 panning right so it'll it'll make the
43:41 nose of the plane go up make it look
43:44 like it's moving left as we pan out I
43:49 think
43:51 it's we're in a weird place right now
43:54 where you have to kind of Imagine in
43:56 your mind what what you're doing with
43:57 your camera moves and then just hope you
44:00 got it
44:01 right um okay so that's going to go make
44:04 our thing here's a here was a video that
44:06 we did last night which came out pretty
44:09 good of a bench although this this
44:12 Shadow seems to be moving across the
44:15 pavement in a weird
44:17 way but it but it it left the pattern on
44:21 the bench intact which Runway did
44:25 not so it'll be interesting to see what
44:27 P does with this
44:30 Aeroplane no apologies thank you yeah
44:33 please keep holding me keep holding me
44:36 accountable for that um
44:40 because
44:42 a I've spent a life apologizing for
44:45 myself and I'd like to end that
44:48 [ __ ] and then B this AI stuff is so
44:51 janky right now that that I think we all
44:54 feel a need to apologize for it and it's
44:57 also it's it's kind of so easy that I
45:00 think we're we are all processing
45:04 through does it have to take a lot of
45:07 time to be valuable and I'm increasingly
45:11 increasingly coming to the conclusion
45:13 that [ __ ] no of course not if I can do
45:17 brilliant work in two minutes
45:21 great why
45:25 not we've we've put up with this
45:27 technology for oh that's [ __ ]
45:32 beautiful so it's still kind of eats the
45:35 letters at the very end of it and the
45:37 the Tails do a little weird movement
45:41 but the coherence there is actually
45:43 pretty good and it's it's sort of
45:46 rotating around and adding some City in
45:49 the foreground
45:52 here and this is this is kind of what we
45:54 did last night as well we we went to to
45:56 a bunch of uh images and uh image
45:59 generation tools and played with them so
46:01 so anyway that's kind of cool all right
46:03 so this one is p. art the other one is
46:05 Runway ml.com those are two of the big
46:09 uh image generation tools now how do you
46:12 how do you find all these tools Kyle
46:14 well I'll tell
46:15 you let me see that's some actual work
46:18 I'm doing for
46:20 tomorrow all right this AI
46:25 salon is a community that if you're not
46:29 a part of if you're curious about AI to
46:32 be fair Kyle the plane is a little
46:34 off-kilter with just one engine yeah I
46:37 know uh the people at row 29 aren't
46:40 feeling too
46:43 well yeah exactly exactly yeah the
46:47 plane's a little it's a it's it's
46:51 entering the time the time space
46:53 continuum in a bit of an odd way
46:56 um okay this group we call it the
47:00 AI
47:02 fell not Salon no the a Sal um the a
47:09 salon is a community of people this is a
47:13 group I started the week after chat GPT
47:15 came out so we just just over a year old
47:19 um we've got an online community so if
47:21 you go to the salon. that's going to
47:23 take you
47:25 to this site
47:28 oops hang on it's gonna take you here
47:32 the
47:34 salon.
47:40 a what's going on internet what's up
47:43 [ __ ]
47:48 cat uh did I
47:51 freeze Robert Rossi is giving me a gift
47:55 hang on thank you Robert
47:57 Rossy there's
47:59 pabs the
48:06 salon. is this because of this https
48:14 thing yes okay
48:18 um so if it if if going to the salon.
48:22 doesn't work for you remove the S in
48:24 HTTPS I I don't have the domain set up
48:27 right so um so anyway the first link
48:30 here is going to take you to the AI
48:32 Salon
48:33 community and when you go here this is a
48:36 really remarkable group of people people
48:38 we just um broke a thousand uh members
48:42 the site's only been up for about two
48:44 months um we were on Discord before and
48:47 we we kind of ripped off the Band-Aid
48:49 and said you know let's go to a
48:51 web-based um Community platform and so
48:53 we did and so so when you come here
48:57 check out about the salon and read about
49:00 who we are why we do what we do and
49:02 check out our values curiosity
49:04 generosity
49:06 exploration collaboration inclusion and
49:09 empathy and read what we're all about
49:13 like read who this group is if these
49:16 values resonate with you you're going to
49:18 really like this
49:20 group if yeah the salon. a is a link
49:23 tree that's correct um but when you
49:25 click on the first link it takes you um
49:28 to
49:29 here and and you can go here if you go
49:31 to AIS salon. mn. that takes you here
49:34 but there's there's some things on that
49:35 link tree that are that are worth uh
49:38 keeping it there um if you resonate with
49:42 these values you're going to really dig
49:43 this group if you don't that's cool but
49:46 this probably isn't the community for
49:47 you um but we want people that are like
49:50 up to it right people that are curious
49:52 people that are adventurous people that
49:54 are willing to share with with one
49:56 another and and learn together and
49:58 that's what this group is so so you know
50:01 please go do that
50:03 the um two the two tools that we just
50:07 looked at Runway and Pika can be found
50:10 on this big Master list of recommended
50:13 tools from the
50:14 salon um this is a list that I
50:16 personally maintain I've got some new
50:18 things to put on here and I think we're
50:20 going to have some new things to put on
50:21 here in particular this week uh which is
50:24 good I have a feeling things are going
50:26 to start accelerating this this
50:28 month and if you're like wait Kyle it's
50:30 going to get faster yeah it's going to
50:32 get faster and weirder so so buckle up
50:34 buttercups it's it's going to be if if
50:38 you here's the deal if you're feeling
50:40 clueless now just give it a month or
50:47 two and and what I can promise you is if
50:50 if you're new here and if you're new to
50:52 Ai and you're like oh my God these
50:54 people are so much farther ahead than
50:56 than I
50:57 am we
50:59 are everyone in this community and
51:02 everyone in generative AI right now is
51:05 feeling similarly clueless and out of
51:07 sorts and out of touch because things
51:09 are moving so fast so don't think that
51:11 you're too bar far behind and don't get
51:14 too comfortable thinking you might have
51:16 it figured out because no one does how
51:18 do you how did you purchase the domain
51:20 for the website I spent 200 [ __ ]
51:24 dollars because a domains are
51:30 expensive so with cash that's how I
51:34 bought
51:34 [Laughter]
51:38 it
51:43 um this is L but Turtle down you guys
51:46 are good you guys are
51:48 cool oh man um okay so this list this
51:53 this list the the uh recommended tool
51:55 list we we start out with chat gbt if
51:57 you're new to Ai and you're wondering
52:00 but c c hey c c c c c Cal C where do I
52:06 start start with chat
52:08 GPT if you can afford the 20 bucks a
52:11 month get chat GPT
52:14 plus if you want to get AI literate I
52:17 think it is actually irresponsible to
52:19 not really understand what chat GPT plus
52:22 chat GPT 4 can
52:24 do because it it is the current
52:29 state-of-the-art um large language model
52:32 I would not buy an annual subscription
52:34 for this so so you can save money by
52:36 paying for an annual one I wouldn't do
52:38 that because new things are
52:40 coming um maybe even this
52:43 week but I would start learning chat GPT
52:46 as soon as you can and and what
52:49 this list does is it points out here are
52:53 the things in chat GPT you should learn
52:55 learn how to use or not even learn how
52:57 to use that's not the right term the
52:59 things you should play with play with
53:02 play play play don't put too much weight
53:05 on getting good at this [ __ ] right now
53:08 just go play then I've also got um other
53:11 large language models here's the four
53:13 biggies here's what we call the kids
53:15 table which are these are not the
53:17 biggies but they're kind of specialty
53:19 ones grock grock not only belongs at the
53:21 Kids Table grock belongs like in the on
53:24 the back patio right now but I have a
53:27 feeling grock's about to get better
53:28 because there there there were a couple
53:30 of tweets about
53:32 it I have a feeling that's going to get
53:35 better pretty quick um then here's some
53:38 image generation tools Del mid Journey
53:40 Kaa Deco here Leonardo and Firefly every
53:43 one of those is a rabbit hole to run
53:45 down Leonardo is getting incredibly
53:48 sophisticated um firefli is getting
53:50 better and better from Adobe um Korea's
53:53 wild Deco here's wild these two are just
53:56 like the the sort of zany weirdos of the
53:59 block I think I need to also add
54:02 um the hell is it called the stable
54:05 stability AI one begins with a c I
54:08 forget what it is but anyway that's that
54:11 video creation Runway ML and P that's
54:13 what we just played with then there's
54:15 also hey Jen and did There's Visa
54:18 there's music and audio here's 11 Labs
54:21 here's sunno oh it's Grammy Grammy night
54:23 let's go play with
54:24 sunno if you haven't played with sunno
54:28 yet this is pretty
54:31 crazy
54:34 um so I'm going to go
54:37 create and log in with my 17 Factor
54:41 authentication I think I got my Tik Tok
54:44 or my uh Twitter Channel back okay so
54:47 we're going to
54:51 do create a song about Taylor
54:54 Swift winning the Grammy in the style of
54:57 Taylor
55:00 Swift we'll see it might not do it
55:03 because of copyright I don't know I
55:04 don't know if you're allowed to yeah no
55:06 couldn't generate that file dang it okay
55:12 um create a pop
55:16 song create a pop song about a pop
55:20 star winning a music award that's shaped
55:23 like a gramophone made of brass
55:29 in the style of that same pop
55:36 star we'll see we'll see how it
55:41 does I guess the Grammys on they don't
55:44 they don't want to make it too easy to
55:45 just copy an artist um sunno isn't quite
55:48 there but if you if you haven't played
55:50 with sun. it's it's it's really fun to
55:52 play with um Sharon mendenhal in the in
55:56 the AI
55:57 Salon has been making a ton of songs she
56:00 even she even made like a
56:03 Broadway um musical using
56:06 sunno
56:23 so I thought it was like that be a real
56:27 what happened to that app did it go away
56:31 which I don't know what that means I've
56:33 been playing with glyph doapp for art
56:35 generation in video to there's so many
56:38 tools now Source Camp I made a few songs
56:41 on Saturday they aren't half bad yeah
56:44 what's interesting about suo oh are
56:45 these done po pop Anthem and Echoes of
56:49 Victory all right so here's our Taylor
56:51 Swift songs without mentioning Taylor
56:54 Swift
56:56 the is
56:59 tonight my story to be the stage for me
57:04 I'm ready to be free I take this back
57:07 long my dreams
57:12 are
57:16 I
57:17 [Music]
57:19 every down
57:22 oh in the SP feeling
57:28 that's pretty good all right here's
57:29 golden
57:32 Anthem moment in my life when just
57:36 become reality Illusions come my life
57:40 lookas my and hands I reach for the
57:43 stars now I'm standing on the stage with
57:47 victory in my heart so that's sunno if
57:50 you don't know sunno you give it a song
57:52 idea and it writes a song with lyrics
57:54 and it sings them and some of them are
57:56 not half bad they're all very um warbly
58:00 and overcompressed and weird they have
58:02 all those kind of audio compression
58:03 artifacts that they all sound like
58:05 autotune right they all sound like te
58:07 pain um but oh thank you very much
58:10 Robert Rossy that was super cool
58:13 um yeah Sharon is the queen ofso she
58:16 really is can you edit that song you I I
58:20 don't think you can but one of the
58:23 things that Kevin
58:26 Vasquez is doing who's an irregular in
58:29 here I notice champ is silent exactly
58:31 Vicki you he knows he knows the score
58:35 for sher's video he was singing his his
58:37 lungs out then for those pop songs not
58:39 so much
58:42 um um the keev what Kevin Vasquez is
58:47 doing with with with modern da W's uh
58:51 digital audio workstations you can take
58:53 songs that are mixed and have the Daw
58:58 split them into tracks so so how Kevin
59:01 Vasquez has been using these he's a
59:02 music producer he's been using sunno To
59:05 Do song ideation he then splits them
59:08 into tracks and he's doing full-on music
59:11 production um out of out ofso bases
59:14 which is amazing um so with some
59:17 cleverness you can do that um Kyle what
59:19 are your favorite Chrome extensions I
59:22 actually just got rid of all my Chrome
59:24 extensions here except for for you know
59:28 you know a handful of things like
59:30 metamask and last pass and things I need
59:32 for daily work um for a couple of
59:35 reasons they tend to be Memory Hogs and
59:39 I had Edge set as my default browser for
59:42 a year and it had so many memory leaks I
59:44 was having to restart my Mac three or
59:46 four times a day and on Macs you
59:48 generally don't need to restart them
59:49 ever like once a month maybe um so I
59:52 wasn't used to that and I was [ __ ]
59:54 hating it so when I switched back to
59:56 Chrome as my default browser I just went
59:58 in and got rid of all my
1:00:01 extensions
1:00:07 I
1:00:09 I'm so it's a really good question it's
1:00:13 a really good
1:00:16 question there are so
1:00:20 many so if you go to on on my list of
1:00:23 recommended tools at the very very
1:00:26 bottom of it is a directory called
1:00:27 future
1:00:54 pedia.com what are sort of the base
1:00:56 tools like the the leading the leading
1:00:59 Tools in each of the major
1:01:01 categories um and when it comes to llms
1:01:04 I'm just trying to stick to like the big
1:01:06 commercial things I'm not even playing
1:01:07 with the open source stuff cuz that's a
1:01:09 whole another
1:01:10 world and so I feel like chrom's cro
1:01:14 Chrome extensions are just going to
1:01:16 distract me and I just be because I have
1:01:18 ADD I'm not good at managing managing
1:01:21 them if I put if I put them in there
1:01:23 they'll be there four years from now so
1:01:25 I'm I'm part of it's a personality thing
1:01:28 and part of it's just I'm trying not to
1:01:30 distract myself cuz there is so much
1:01:34 coming and so much to keep up with and
1:01:36 I'm just
1:01:38 like so can you edit it or does not it
1:01:41 does not give you that option it doesn't
1:01:42 give you that option Jason but I just
1:01:43 talked about you can you can split it in
1:01:45 a
1:01:47 Daw um can you you can do RI oh you can
1:01:51 do remixes oh okay I don't I don't know
1:01:54 a ton about sunno other than and I don't
1:01:56 have a paid version of sunno so I'm just
1:01:58 using the free version so I assume if
1:02:00 you have a paid version you can do more
1:02:02 if you can do remixes cool that's
1:02:04 great um champy love you champy champy
1:02:09 you're gonna be a star he's already a
1:02:11 star he's like yeah I know I
1:02:14 know champy champy was invited to the
1:02:17 Grammys we turned it down they they were
1:02:19 just like you know he was like am I
1:02:22 making as much as Taylor Swift they're
1:02:24 like well no she's Taylor Swift he's
1:02:26 like then I'm
1:02:28 out you want to sing in dog you you pay
1:02:31 me Swift
1:02:32 rates he's
1:02:34 he's he's rough he's rough he's get it
1:02:38 he's rough when it comes to the
1:02:40 [Laughter]
1:02:45 negotiations oh
1:02:51 man has anyone tried Luma yet uh which
1:02:55 one's Luma I think I tried Luma
1:02:58 ruff ruff ruff ruff
1:03:02 ruff chat GPT is first second and third
1:03:05 place I agree with that
1:03:07 although so here let me show you
1:03:09 something
1:03:10 to
1:03:15 um oh by the way if you join the AI
1:03:19 Salon don't just be a lurker okay
1:03:23 remember when I said we have values one
1:03:25 of those
1:03:26 values this one right here
1:03:31 generosity a form of
1:03:34 generosity is introduce yourself to the
1:03:37 community so one of the first things I'd
1:03:39 like you to do if you decide to join the
1:03:42 AI Salon is go to the introduce yourself
1:03:44 section in the community and introduce
1:03:47 yourself oh here's Cliff
1:03:50 Booker did it with an image that's super
1:03:52 cool that's there's a good generous
1:03:54 introduction he wrote a poem he made an
1:03:57 image here let's give that a heart let's
1:04:00 let's reward that with that's
1:04:03 sweet 37 minutes ago all right a minute
1:04:07 ago so there's some people introducing
1:04:09 themselves so go go do
1:04:12 that um you can also check out the show
1:04:15 and tell Channel which is a lot of fun
1:04:18 there's Shar men Hall making some more
1:04:20 music or pictures about making
1:04:23 music
1:04:26 waiting for the piano man oh yeah I bet
1:04:29 he's coming up isn't
1:04:31 he Billy Joel's singing his first new
1:04:34 song in 30 years
1:04:38 tonight oh very
1:04:40 cool very very
1:04:44 cool um okay so in water cooler so water
1:04:48 cooler is just a discussion where you
1:04:49 can discuss stuff
1:04:52 and Chris mcab posted this article from
1:05:13 decoder okay um accept and close Google
1:05:18 might launch its chat GPT plus
1:05:21 competitor Google or Gemini Advanced
1:05:24 this week so Gemini Advanced is I guess
1:05:28 their commercial term for Gemini Ultra
1:05:32 which is the name of the model Bard
1:05:34 answered me in Chinese the other day
1:05:36 anyone else get that oh boy Pate Pate
1:05:40 what's going on over there at Google why
1:05:42 is it answering people in
1:05:45 Chinese Pate Pate what are your tpus
1:05:48 doing did you do the math wrong did you
1:05:51 do a minus where you needed to do a plus
1:05:54 did did you close your Expressions
1:06:01 pate and Pate didn't give us any
1:06:04 forewarning
1:06:05 Pate wait are you still working in the
1:06:08 United States Pate blink
1:06:09 [Laughter]
1:06:15 twice um
1:06:18 Okay Google might release its chat GPT
1:06:22 plus competitor on February sex
1:06:25 7th this name
1:06:34 suggests what happened it went to a
1:06:37 different article here we
1:06:39 go the name suggests a change for The
1:06:42 Bard chatbot after Google announced Bard
1:06:45 Advanced at the end of last year the
1:06:47 Gemini Advanced chat but will be powered
1:06:49 by the aonomus Gemini model in the ultra
1:06:53 1.0 release
1:06:55 according to code Snippets posted on
1:06:58 Reddit hacker going to hack um Google
1:07:01 says the new chatbot is far more capable
1:07:03 at highly complex tasks like coding
1:07:06 logical reasoning and nuanced
1:07:09 instructions and creative collaboration
1:07:12 thanks to the ultra
1:07:13 model the company promises regular
1:07:16 updates the Gemini let's see Google
1:07:19 tries to catch up with chat GPT so the
1:07:21 7th is Wednesday right
1:07:25 February 5th is tomorrow 6th is Tuesday
1:07:28 7th is Wednesday so it is possible kids
1:07:33 that Wednesday Bard is not that bad yeah
1:07:36 but Bard Bard right now is based on
1:07:40 Gemini Pro which is essentially the
1:07:43 equivalent of chat GPT 3.5 so so Bard is
1:07:48 is at about
1:07:50 parody with the free version of chat GPT
1:07:53 now that
1:07:55 said if you haven't seen the um the
1:08:00 large language
1:08:02 model it's called The Arena right have
1:08:04 you guys seen this thing this is cool
1:08:06 let let's go look at this LM
1:08:10 Arena chatbot
1:08:12 Arena okay if you haven't seen this this
1:08:15 is a lot of
1:08:18 fun all right
1:08:23 so
1:08:30 [Music]
1:08:33 is this the one I think this is the
1:08:48 one is this
1:08:53 it
1:09:04 yeah this is
1:09:05 it
1:09:08 so Google bard Gemini
1:09:11 Pro
1:09:13 beat um the 613 version of chat gp4 and
1:09:18 the 314 version of chat GPT 4 so
1:09:21 everything but the turbo version so so
1:09:24 gp4 turbo is still at the top but Gemini
1:09:27 Bard made it to number three so it's so
1:09:30 so Bard in the non-ultra version right
1:09:33 now has beat some gp4 things now what
1:09:35 did it beat it in it beat it in this
1:09:37 thing called chatbot Arena which if you
1:09:40 haven't played with it why why can I not
1:09:43 find out how to play with this is it
1:09:46 here
1:09:53 introduction
1:09:59 try the voting
1:10:04 demo verifying that I'm human connection
1:10:08 is
1:10:10 secure okay this is it okay so so so
1:10:13 what this thing does is you give you
1:10:17 write a prompt all right so I'll do my
1:10:19 classic prompt
1:10:22 um explain
1:10:26 quantum
1:10:28 mechanics to me in an
1:10:32 mnm
1:10:34 WP this is one of my parlor tricks that
1:10:37 I like to do and so what this has got
1:10:39 it's got Model A and model B we don't
1:10:40 know what these models
1:10:42 are so when you say send it's sending
1:10:47 the same prompt to two different
1:10:49 models and you get to watch how they
1:10:51 respond in this case they're not respond
1:10:54 responding at
1:10:55 all come on what are we doing there we
1:11:00 go okay let's
1:11:04 see so the one on the left listen up I'm
1:11:07 going to break it down quantum mechanics
1:11:09 it's crazy town particles that exist in
1:11:11 many states like Schrodinger's
1:11:13 cat it's not just a
1:11:16 meme it's real man okay that's weird
1:11:21 quantum mechanics it's a mind blow it's
1:11:23 a whole new way of knowing it's all
1:11:26 about probability okay this is pretty
1:11:28 crappy all right the one on the right
1:11:31 all right hold tight I'll break it down
1:11:32 no pretense quantum mechanics a world
1:11:34 where things get intense forget the big
1:11:37 stuff the rules you understand cuz down
1:11:40 at the bottom it's a whole different
1:11:41 land this one at least Rhymes particles
1:11:44 or waves yeah it sounds kind of strange
1:11:47 they ain't in one spot their position
1:11:48 can change like a ghost in the night
1:11:50 they Flicker and fade they're here
1:11:52 they're there a probability parade so
1:11:55 this one's pretty good so so so now what
1:11:57 I do is I go B is so you can go a is
1:11:59 better B is better it's a tie both are
1:12:01 bad so I'll go B is better and then it
1:12:04 tells us that's
1:12:06 Bard and then the one on the left is the
1:12:08 vacuna 13B model so it kind of makes
1:12:11 sense that a 13 billion parameter model
1:12:13 would suck a little more and then
1:12:15 there's Bard winning so then this turns
1:12:17 into this
1:12:20 leaderboard that is
1:12:22 dynamically created
1:12:25 and so Andre karpathy who is the chief
1:12:28 scientist at
1:12:30 uh open Ai and then he went to I think
1:12:33 Google to work on something and then he
1:12:35 came back to open AI to work on safety
1:12:38 um this this is one of the one of the
1:12:41 sites that he says he trusts for for um
1:12:45 for ranking llms and so Bard the non
1:12:49 Ultra version of Bard is already
1:12:53 behaving pretty good in a in a you know
1:12:56 in a measured way
1:12:58 so I figured that might make you happy
1:13:01 there Pate that that your little your
1:13:03 little your little uh startup company
1:13:05 there is doing
1:13:07 okay
1:13:09 um let's see where
1:13:14 is I guess I'll just go to
1:13:17 this what site is this so that site that
1:13:20 site was Brandon was um look up chatbot
1:13:27 Arena and then you have to click on the
1:13:29 demo and then then you can do that back
1:13:31 and forth with it it's fun to do and
1:13:34 you're actually helping science you know
1:13:36 they're actually that leaderboard is
1:13:38 dynamically generated by people playing
1:13:40 with that
1:13:45 tool cuz I think that I think that
1:13:47 getting real people to do real prompts
1:13:49 is is something interesting uh versus
1:13:52 just you know scientific Ben
1:13:55 benchmarks
1:13:57 um is there anything else in
1:14:07 this according to a leaked web web
1:14:12 text so anyway we should pay attention
1:14:15 to that you
1:14:16 know do I have any faith that this is
1:14:18 actually going to launch on February 7th
1:14:20 no but you know someone found some code
1:14:23 that suggest gests it and the Gemini
1:14:27 Advance the the last thing I
1:14:29 remember uh in December was they were
1:14:32 going to call it Bard Advanced but I
1:14:35 think Gemini is a better word and I
1:14:36 think Gemini Advanced is good Gemini
1:14:39 Advanced is likely going to be 20 bucks
1:14:40 a month
1:14:43 also so you're going to have to pay for
1:14:45 it just like you have to pay for chat
1:14:47 GPT and if it's as good as they say it
1:14:49 is then that's going to be worth it okay
1:14:52 let's go now look
1:14:57 at some of the oh there's
1:15:00 Joanie oh my God what a mess I have
1:15:03 every Joanie Mitchell album It's amazing
1:15:06 is that amazing she never sang it the
1:15:08 Grammys until now there's the there's
1:15:11 the uh Luke combes Tracy
1:15:13 Chapman Fast Car Duo that was so good
1:15:22 duet
1:15:26 oh yeah oh this was one did you hear
1:15:28 about this one holy [ __ ]
1:15:31 so um a finance worker in Hong Kong paid
1:15:36 out $25
1:15:37 million he received a suspicious email
1:15:41 saying hey wire $25 million to this
1:15:46 account and then he got a FaceTime a
1:15:49 video call from his CFO saying no no
1:15:52 it's legitimate you you can send the $25
1:15:54 million and he sent the $25 million and
1:15:58 it was a deep fake of his CFO so holy
1:16:01 [ __ ] that's
1:16:03 coming good Lord
1:16:06 um the reason I wanted to go play here
1:16:10 oh it's all going to be full of Grammys
1:16:11 let me go let me go Vision Pro
1:16:15 here Vision Pro let me do Casey Casey
1:16:20 niat how do you spell Casey
1:16:25 is that is that it
1:16:31 no
1:16:39 KC nice
1:16:52 St
1:16:56 thanks got the messages again got music
1:16:59 set up here big screen TV set up on my
1:17:03 wall Gordon Ramsay showing me how to
1:17:06 cook above the
1:17:07 stove so let me go back and show you
1:17:09 this so so what what's starting to
1:17:11 happen so so Vision Pro's been out
1:17:14 for 48 hours now and there's there's an
1:17:18 interesting Casey neistat if you don't
1:17:21 know him he's he's a filmmaker from from
1:17:24 you know ages ago that got famous doing
1:17:26 YouTube videos like these really sort of
1:17:28 low low rent low production value
1:17:31 YouTube videos that were super
1:17:33 authentic um and and basically he wore
1:17:37 his vision Pros out in New York City for
1:17:40 a whole day um and and gave his his uh
1:17:46 you know his take on them and I I
1:17:47 thought it was pretty interesting this
1:17:49 guy's talking about so so what you're
1:17:52 looking at here is this is this guy's
1:17:54 house right he's walking around his
1:17:57 physical
1:17:58 house um and then when you pin these
1:18:00 things up they're persistent so what was
1:18:03 what was interesting about the people
1:18:05 that I've seen that that are wearing
1:18:06 these things out in public is if they're
1:18:09 walking and they try to bring up the
1:18:11 interface it just blows past them
1:18:13 because they're walking and so the
1:18:15 Vision Pro is about establishing these
1:18:18 interfaces in a physical space um so so
1:18:22 I guess they're just dropping these
1:18:24 virtual
1:18:25 objects um oh it looks like Miley Cyrus
1:18:28 W another one good for her she's never
1:18:30 won a Grammy which I can't believe um
1:18:32 anyway
1:18:34 okay got the messages again got music
1:18:37 set up here big screen TV setup on my
1:18:41 wall Gordon Ramsey showing me how to
1:18:43 cook above the
1:18:45 stove notes right here for some
1:18:49 groceries can walk around
1:18:51 freely here's what it looks like when
1:18:53 it's
1:18:57 on hey how's it going oh hey catch thank
1:19:02 you and as we walk around the house
1:19:05 everything stays pinned exactly where we
1:19:07 left
1:19:08 it and that's it let me know if you
1:19:11 crazy right
1:19:13 um trading meme coins from the Apple
1:19:16 Vision Pro uh here was Simpsons
1:19:18 predicting this of
1:19:22 course
1:19:31 [Music]
1:19:40 um this one I haven't
1:19:42 seen with div Vision Pro res sharing
1:19:47 okay that's that guy again let me see if
1:19:49 I can find Casey neistat how do you
1:19:52 spell Casey neistat St hang on let me go
1:19:57 find
1:19:59 Casey
1:20:07 niat c a
1:20:11 y c a y I had nice that
1:20:22 right
1:20:24 here we
1:20:29 go okay here yeah this is
1:20:37 it like what that's wild like real yeah
1:20:41 I mean your eyes are close your it's too
1:20:44 much teeth I think I don't know I want
1:20:46 to make a review come help me all right
1:20:47 sounds good all right meet you
1:20:49 downstairs all
1:20:50 right thanks amilios all right if you
1:20:54 want like a normal review of this thing
1:20:56 I can't recommend enough to go watch to
1:20:58 go watch Marquez brownley his is
1:21:00 fantastic this is a different like I
1:21:02 kind of have a different take on this
1:21:04 and different questions and that's what
1:21:06 this
1:21:07 is but Marquez is really
1:21:10 good what is this thing that I'm holding
1:21:15 literally look at him look at him
1:21:18 walking around New York
1:21:22 City
1:21:28 I had it for like an hour yeah okay can
1:21:31 you put Windows up now if you want yeah
1:21:32 so like I can put a window up now but
1:21:35 when I start walking they like go away
1:21:37 Fade
1:21:38 [Music]
1:21:39 Away care take
1:21:42 care this so cool when he gets the Time
1:21:45 Square and he's just
1:21:48 Computing get a random pile of bricks an
1:21:51 old toilet I found the and a bunch of
1:21:54 random wooden poles there's more
1:21:57 poles pil let's pile on that boat and
1:21:59 leave and as you can
1:22:00 [Music]
1:22:02 see but later in this video we're going
1:22:05 to
1:22:06 aund and
1:22:08 also super Island but first we're going
1:22:12 to head over
1:22:12 [Music]
1:22:21 to strategy interacting with people they
1:22:24 Gaslight you into thinking so if you're
1:22:28 moving a lot while you're moving this
1:22:30 thing comes up that says tracking fail
1:22:33 so it doesn't track if you're moving
1:22:35 lock there's nothing for the sensors to
1:22:37 like lock on to it doesn't know where to
1:22:39 put stuff but standing at a subway stop
1:22:43 watching a Mr let me Jump Ahead here see
1:22:46 if I can get him when he's in or VR like
1:22:49 I have the Oculus it's great but this
1:22:51 definitely feels like a toy this feels
1:22:53 like a little glimpse into the future of
1:22:56 what Computing could be
1:22:58 like down the road why is everybody
1:23:00 staring at
1:23:02 me do I look as ridiculous as those
1:23:04 people make me feel like I look you're
1:23:06 just your eyes are glowing how you doing
1:23:10 you
1:23:11 doing Jordan you're standing right in
1:23:14 front of my Mr Beast video come right
1:23:16 there okay years ago this used to be a
1:23:18 bar yep yeah you want to walk under
1:23:21 this
1:23:24 right like that so you know he's sitting
1:23:29 in Time Square and he's got this pinned
1:23:32 you know YouTube page this Mr Beast
1:23:36 video I saw another guy earlier and let
1:23:40 me see what he says at the end he he he
1:23:42 sort of summarizes this at the end that
1:23:44 I think is is
1:23:51 interesting
1:24:02 Billy Joel's
1:24:21 singing
1:24:31 oh that sounds good Billy Joel's song
1:24:33 sounds good nice okay um I don't know
1:24:38 why my this crap ain't working anymore
1:24:41 all right let me go off him Vision
1:24:44 Pro there was one guy one guy was
1:24:47 watching like 30 basketball
1:24:50 games live at the same time in his
1:24:53 living
1:24:54 room there's a guy at a ball game Vision
1:24:57 Pro Courtside is
1:25:02 crazy people are going to be such idiots
1:25:05 in these things they already
1:25:08 are Jen Z going wild bro literally
1:25:12 transformed into a
1:25:15 stormtrooper oh that's
1:25:19 funny let me see if I can find one of
1:25:21 these other ones what's this guy doing I
1:25:23 show you what my morning routine is like
1:25:25 with the Apple Vision Pro as you can see
1:25:27 I'm in bed right now I wear my Apple
1:25:29 Vision Pro to bed actually so that the
1:25:31 moment I wake up I can immediately start
1:25:33 being productive uh so the moment I open
1:25:36 my eyes I can see the the price of
1:25:38 Bitcoin I 90% of my net worth in Bitcoin
1:25:40 so I want to make sure that prices keep
1:25:42 going up over here I got
1:25:45 X comed it's
1:25:49 gavity there's a dweeb nut there's a
1:25:53 there's an app let's
1:25:56 see the like what I'm hear what what I'm
1:25:58 hearing consistently if you I mean
1:26:01 you're probably seeing the same [ __ ] is
1:26:03 that people are surprised at how
1:26:05 immersive it is here's another here's a
1:26:07 guy walking his robot dog with his
1:26:09 vision Pros on so we're going to have a
1:26:11 fair amount of that in the at least in
1:26:13 the major cities I think in smaller
1:26:15 cities people are going to get beat up
1:26:16 if they try to do
1:26:21 that
1:26:30 all
1:26:32 right um okay so why do I think that's
1:26:37 worth talking about here
1:26:41 well the vision the the the amount of
1:26:46 people driving in them is crazy though I
1:26:48 know that one dude driving this the
1:26:50 Cyber truck's he's
1:26:53 the the Cyber truck is driving itself
1:26:55 and he's in there doing like Tom Cruz in
1:26:59 that in whatever movie he did yeah it's
1:27:02 it's going to be dumb for a while um
1:27:06 that's okay so so why it's relevant to
1:27:09 this Channel and to us the amount of
1:27:14 machine learning and AI That's built
1:27:16 into that device is insane it's doing
1:27:21 realtime signal process
1:27:23 of all these different sensors and and
1:27:26 you know it's there's there's a lot of
1:27:28 there's a lot of AI and machine learning
1:27:29 in there um right now there's a lot of
1:27:33 this kind of you know interface kind of
1:27:36 stuff um when you have the ability to
1:27:38 just talk it and have it execute complex
1:27:42 operations um you know you're going to
1:27:44 see AI put in there um pretty
1:27:47 significantly I think what's kind of
1:27:49 weird about them is instead of them
1:27:51 being wait in instead of it being
1:27:54 augmented the image of the real
1:27:57 world the image of the real world is
1:28:02 created right yeah exactly that but but
1:28:06 what people are pretty consistently
1:28:08 saying is it is of a high enough frame
1:28:14 rate oh sorry Seline Dion just walked on
1:28:17 stage
1:28:19 and she's very sick and Taylor just lost
1:28:23 her mind hang on a
1:28:47 sec
1:28:50 here I need
1:28:53 my
1:29:17 heart right
1:29:20 nice this is become a React channel well
1:29:24 maybe a little
1:29:25 bit um I mean only tonight I I'm I'm
1:29:30 this isn't a React channel this is a
1:29:32 react a a react squared Channel because
1:29:35 I'm reacting to other people's reactions
1:29:37 um we can see too oh the the Grammys I
1:29:41 don't think I can show it I think I
1:29:43 think Tik Tok will shut the shut the
1:29:45 live down
1:29:48 um but we could do watch parties at some
1:29:51 point I suppose
1:29:54 um no I just so so here's the
1:29:58 thing based on the first 48 48 Hours
1:30:03 response Apple seems to I I thought they
1:30:06 did this when they announced it but now
1:30:08 that it's out and in the hands of people
1:30:11 you know people are pretty Relentless
1:30:13 and ruthless about this this kind of
1:30:15 technology and this is the first of
1:30:18 these goggle type Dei
1:30:21 devices
1:30:23 where people are like yeah I could wear
1:30:26 I could do my work in this right I could
1:30:28 I could walk around the world and it's
1:30:31 working and this is version 1.0 right so
1:30:36 as with all Apple products by the time
1:30:39 they get to version three they have they
1:30:41 have usually established a new category
1:30:44 I think they have clearly established
1:30:47 spatial Computing as the new
1:30:50 category and it it includes all sorts of
1:30:54 things like Robotics and self-driving
1:30:57 cars as well right because they're
1:30:59 driving in
1:31:01 space and this idea that you're you're
1:31:04 leaving these objects all around the
1:31:06 world it it's a new kind of thing and
1:31:09 it's here right so it's it's not like
1:31:11 it's coming we could look at Oculus all
1:31:13 we wanted and go well when it doesn't
1:31:16 look like you know Second Life from from
1:31:21 1999
1:31:22 you know maybe I'll drop 500 bucks on
1:31:24 the stupid [ __ ]
1:31:26 thing um and I know I know the Oculus
1:31:29 have gotten better and things are better
1:31:31 but but Apple seems to have have done it
1:31:33 and they've they've son sold a ton of
1:31:35 these things already at 3500 bucks so so
1:31:39 that just says to me that's coming as
1:31:42 well right so
1:31:45 so people have been talking about AR and
1:31:48 VR for a decade meta changed their name
1:31:51 from Facebook to meta to try to lay
1:31:54 claim to that what was that 5 years ago
1:31:56 6 years
1:31:57 ago they spent tons of money to not
1:32:01 really build anything that caught on in
1:32:02 any way and I think Apple has actually
1:32:05 done something here so this this starts
1:32:07 to feel
1:32:10 like when you can go to something like
1:32:13 Skybox blad labs.com if you haven't seen
1:32:17 that
1:32:21 Skybox
1:32:24 blockade labs.com
1:32:32 where you know here's a here's a scene
1:32:35 that you can twirl around in and then I
1:32:37 can
1:32:38 say um
1:32:41 Times Square New York City in
1:32:50 2075 uh
1:32:55 in
1:32:57 smog in smog let's just say that and
1:33:00 then we'll pick a style we'll
1:33:02 go
1:33:04 realistic and we'll go generate oh I
1:33:07 didn't pick a style wait oh here we go
1:33:09 realism go generate and so now this is
1:33:13 going to generate us that I've seen
1:33:15 people live streaming from
1:33:19 TV yep thank you than you everyone it's
1:33:23 weird it's still failing to
1:33:25 load as an apple stockholder I'm excited
1:33:28 for the future yeah you should be listen
1:33:32 I didn't I was like when is Apple going
1:33:34 to [ __ ] reinvent a new category and I
1:33:37 I the the last place I thought they
1:33:39 would do it is in augmented reality I
1:33:43 know they've been talking about it for
1:33:44 15
1:33:45 years um that announcement of Vision Pro
1:33:49 and what that technology is oh what did
1:33:51 Taylor just win
1:34:01 win she won another one what'd she
1:34:07 win the other people don't look very
1:34:12 happy they're all
1:34:14 [Laughter]
1:34:17 like oh for album of the Year first four
1:34:21 time album of the Year winner in Grammy
1:34:23 history Taylor Swift holy [ __ ] good for
1:34:26 her she is such a
1:34:29 beast God damn this this last two years
1:34:32 for her has just been insane okay so
1:34:36 anyway so here's here's you know New
1:34:38 York
1:34:39 city so if you think about if you think
1:34:43 about like right now we we have all the
1:34:45 little seeds of what we're going to be
1:34:47 able to do in the future where you're
1:34:50 going to be able to have all these
1:34:51 different different
1:34:53 screens you're going to be able to just
1:34:55 say hey I want X done or I want to
1:34:58 create X and little agents are going to
1:35:01 go off and do that [ __ ] for you and
1:35:03 bring it back to your screens wherever
1:35:04 you have them and it'll probably be
1:35:06 Contex sensitive where if you're in your
1:35:08 living room there's a certain set of
1:35:10 tasks you do if you're in your kitchen
1:35:12 there's a certain set of tasks you do if
1:35:13 you're if you're out in Time Square you
1:35:16 know there's probably something else you
1:35:17 do um and then tools like this where
1:35:22 you're you're going to be able to
1:35:26 speak worlds into existence that will
1:35:29 have game mechanics in them and objects
1:35:32 in them 3D objects in them and
1:35:34 activities in them and other people in
1:35:37 them it's going to get really weird
1:35:41 really
1:35:43 fast
1:35:44 2023 the last normal
1:35:48 year so I I just
1:35:52 looking at the just go if you haven't
1:35:54 been watching the reviews of The Vision
1:35:56 Pro just go watch a bunch of them
1:35:59 um and and
1:36:02 just do what you can to just try to
1:36:05 assimilate okay let's assume this gets
1:36:07 real in the next 5 years and by real I
1:36:10 mean it's not an embarrassingly big
1:36:13 heavy weird thing it's something more
1:36:15 like these glasses or you know maybe
1:36:17 slightly bigger but a stylized version
1:36:19 of
1:36:20 glasses that can do all
1:36:23 this and something like the rabbit R1
1:36:26 that's going to be like this little AI
1:36:28 walkie-talkie that we have that gets
1:36:30 integrated into everything like
1:36:31 everything becomes
1:36:34 everything and you don't need to program
1:36:37 anymore no one needs to program well
1:36:39 people need to program the systems that
1:36:41 we'll use but if we need an
1:36:44 application to calculate the tip at the
1:36:48 restaurant we're at it will just
1:36:50 dynamically be generated
1:36:52 for our bill based on how nice we were
1:36:56 to the waitress and how nice she was to
1:36:58 us right like it'll be able to take all
1:37:01 that into account like that's the world
1:37:03 we're heading into they are slow playing
1:37:06 Siri 2 though slow they're they're
1:37:08 incomprehensive they're non-existent
1:37:10 with here's the thing about Apple though
1:37:13 Apple always waits Apple always waits
1:37:16 they always just sit back and they let
1:37:18 Mark Zuckerberg rename his
1:37:20 company and launch Second Life
1:37:27 2.0 and fall on his face and they go
1:37:30 okay noted duly noted we know we we know
1:37:34 what not to do there what site is this
1:37:36 this site is blockade labs.com it's on
1:37:38 the if you go to the AI
1:37:40 Salon in the community section AI tools
1:37:44 and resources you'll see this list it's
1:37:46 down in
1:37:48 3D 3D worlds and objects it's the first
1:37:51 one blockade Labs all right and if
1:37:54 you're wondering wait what's the AI
1:37:56 Salon what's the AI Salon that's the AI
1:37:58 Salon go
1:38:00 there and click on the first link that
1:38:04 says AI Salon community and that will
1:38:07 take you to this web-based community
1:38:10 site that we've
1:38:12 got and go to welcome to the salon and
1:38:14 read about who we are and why we do what
1:38:17 we do and read about our
1:38:20 values
1:38:23 if you're curious about AI if you're
1:38:26 excited about this if you're the only
1:38:28 one in your circle of friends and family
1:38:30 that's into this and everyone's like
1:38:32 Sally could you stop talking about AI
1:38:35 now we're really kind of sick of it
1:38:36 could we just go out to dinner once and
1:38:38 have a milkshake and not talk about
1:38:40 stupid chat GTP or
1:38:43 something if you're that person that
1:38:45 hears people in your life talk like that
1:38:47 come join
1:38:50 us that's the end of that's the end of
1:38:53 snarky weight staff rip yeah
1:38:56 exactly ex it you know it's going to be
1:38:58 weird weird Silver Fox
1:39:02 is just so much is going to
1:39:07 be analyzed and responded to in real
1:39:12 time it's it it's going to absolutely
1:39:14 change behavior in fact that's one of
1:39:17 the things that that was in the the eu's
1:39:19 new new laws that the EU just came out
1:39:22 with um is has got to do with not
1:39:27 measuring emotional reality at work like
1:39:31 not measuring people's emotions at
1:39:34 work cuz cuz everything's going to be
1:39:37 tracked everything's going to be
1:39:39 analyzed in real time it's going to be a
1:39:41 trip so anyway when you come to the AI
1:39:44 Salon really look at what we're about
1:39:47 and if our values and what we're about
1:39:50 sounds good to you great if you're like
1:39:52 well that seems a little
1:39:54 weird AI Salon is a rag tag group of
1:39:57 Misfits Nero Wells and miscreant so if
1:40:00 you read that and you're like well that
1:40:01 sounds geeky and stupid what are they
1:40:03 nerds maybe I don't [ __ ] know but if
1:40:06 that makes you happy then you'll
1:40:07 probably dig this place if you like the
1:40:10 idea of being around super curious
1:40:12 people that are also really generous and
1:40:14 how they share what they're learning
1:40:17 great if you're like you just want to be
1:40:19 one of these guys you can make money
1:40:22 J it's all about the coin and figuring
1:40:25 it out and what's the angle you're
1:40:26 probably not going to dig this group now
1:40:28 does this group talk about how do we
1:40:30 make money with this [ __ ] yes we do but
1:40:32 we're not leading with the tool what's
1:40:34 your prediction for AI at the end of
1:40:36 2024 oh good lord Dr
1:40:42 bod
1:40:43 okay um my prediction for AI at the end
1:40:47 of
1:40:50 2024
1:40:53 [Laughter]
1:41:03 I think we'll we will be so Google put
1:41:06 out a
1:41:07 paper three months ago on the different
1:41:11 levels of AGI the different classes or
1:41:14 stages of AG right now we're at AGI 1.0
1:41:18 like
1:41:19 one which is like it can sort to do some
1:41:21 [ __ ] but not
1:41:23 much and
1:41:25 then two and
1:41:27 three like ai3 is I should probably go
1:41:32 look at it hang on let me go look at
1:41:37 this Google AGI stages is that what it's
1:41:45 called levels of
1:41:50 AGI
1:41:55 level one emerging equal to or somewhat
1:41:58 better than an unskilled human that's
1:42:01 where we are right now level two
1:42:04 competent at least 50th percentile of
1:42:07 skilled
1:42:09 adults meaning the AI is better than
1:42:14 half of the people skilled adults
1:42:16 skilled
1:42:17 adults level three expert at least 90th
1:42:20 per of skilled
1:42:23 adults I think by the end of 2024 we're
1:42:26 at least at level two and we might be
1:42:29 pushing level
1:42:32 three or at least have glimpses of what
1:42:35 level three looks like now to get to
1:42:37 that so if you want to get to the 90th
1:42:39 percentile of skilled
1:42:41 adults then AI needs to be able to
1:42:45 reason and it needs to be able to take
1:42:48 actions so do you know if the rabbit R1
1:42:51 will interface with chat GPT using the
1:42:53 button I don't know they just cut a deal
1:42:55 with perplexity so my guess is
1:42:58 not but it also looks like they're
1:43:01 giving you an interface where you can
1:43:03 set up your own actions on your own
1:43:05 websites so in theory you could set up
1:43:08 an action for chat GPT So in theory you
1:43:10 should be able to do
1:43:12 that
1:43:14 um so I
1:43:18 think by the end of 2024
1:43:22 I think we'll see whatever is the next
1:43:24 thing
1:43:25 from from open
1:43:28 Ai and as Sam Alman put it I wrote it
1:43:32 down because it's it's it's very
1:43:33 particular
1:43:36 language gp4 is best understood as a
1:43:40 preview so what we're playing with now
1:43:42 is best understood as here's the
1:43:45 training wheels kids okay you go play
1:43:50 you go make your gpts here's the GPT
1:43:52 store it's not really a store it doesn't
1:43:54 [ __ ] matter GP gp4 is best understood
1:43:57 as a preview and then he followed that
1:44:00 up with the sentence progress here is
1:44:03 not
1:44:05 linear GPT 5 isn't going to be like two
1:44:08 or three times better than gp4 it's
1:44:11 going to be a completely different
1:44:14 thing it looks like
1:44:17 Google maybe as soon as Wednesday is
1:44:20 going to be launching something that's
1:44:21 at least as good as
1:44:24 gp4 that
1:44:26 is multimodal from the
1:44:30 get-go
1:44:33 um
1:44:34 and so
1:44:36 so so we're going to have gp4 which is
1:44:40 already multimodal we're going to
1:44:43 have um what's it called Gemini Advanced
1:44:46 which is Gemini Ultra which is super
1:44:49 multimodal um those two are going to be
1:44:51 be battling it out together and I would
1:44:54 not think that Sam Alman and the open AI
1:44:57 crew are going to want Google to if if
1:45:00 Google takes the
1:45:02 lead I don't think open AI is going to
1:45:04 want them to have the lead for long so
1:45:06 it might accelerate them putting GPT 5
1:45:09 out there so I think what we're going to
1:45:12 see is we're going to
1:45:17 see these systems go from being kind of
1:45:20 d
1:45:21 where they make a lot of mistakes and
1:45:23 they don't know what they're
1:45:24 saying to reasoning where it'll give you
1:45:28 an answer and they will actually read
1:45:30 their own answer and not actually show
1:45:33 it to you until they've got it right so
1:45:35 I think the hallucination thing will
1:45:37 disappear or dramatically
1:45:41 lower I think that um we'll start to see
1:45:47 agents so think think like a GPT
1:45:51 but it takes action on your behalf and
1:45:54 maybe it's generated dynamically so so
1:45:57 like a good example of an agent might be
1:45:59 I want to go on a
1:46:00 vacation okay well if if you were going
1:46:02 to say to your assistant I want to go on
1:46:04 a vacation that assistant would know how
1:46:07 many people in your family they would
1:46:08 probably ask some questions like is it
1:46:10 just you or is it the whole family do
1:46:11 you want to go somewhere you've been
1:46:13 before or do you want to go somewhere
1:46:14 new do you want this to be a fancy
1:46:16 vacation or an inexpensive vacation
1:46:18 right so they need some information one
1:46:20 once they have that information what's
1:46:22 what's a good assistant do well let me
1:46:24 call up let me go to Expedia see if
1:46:26 there's any sales let me see anywhere
1:46:28 we've got frequent flyer miles do all
1:46:30 that so they're going to go do that
1:46:32 they're probably going to Google
1:46:34 different locations if you want to go
1:46:35 somewhere you've never been before
1:46:37 they're going to go find locations
1:46:38 probably check with you I was thinking
1:46:40 about the
1:46:42 malds or puket where where Cindy [ __ ] is
1:46:46 right now she's in
1:46:48 puket I read it and thought it was
1:46:51 pronounced [ __ ] it but you
1:46:53 know not as worldly as I thought I
1:46:59 was so these agents are going to go do
1:47:01 that for you I want to go on a vacation
1:47:04 It'll ask you some questions and then
1:47:06 boom they'll go off it'll start creating
1:47:08 you day trips making reservations for
1:47:11 you making reservations at restaurants
1:47:14 here's your itinerary I don't like
1:47:16 Wednesday Wednesday seems boring can we
1:47:17 go somewhere exciting sure off go some
1:47:20 more agents to your stuff so I think
1:47:22 we'll see you know incred deeply
1:47:25 increased
1:47:27 capabilities reasoning and
1:47:30 agents and then and then I think the
1:47:33 multimodal stuff I think video by the
1:47:35 end of
1:47:36 2024 gets staggeringly good there's a
1:47:40 post from Matt wolf who said he thought
1:47:43 that the ability to create a coherent 30
1:47:46 to 60 Minute video was going to be two
1:47:49 to three years out
1:47:51 and he said I've got to revise that I
1:47:53 just saw something it's going to be two
1:47:55 or three months or like one or two
1:47:57 months so probably by end of March we'll
1:48:01 have some video tool like either one of
1:48:04 the ones we know gets better or some new
1:48:06 one [ __ ] shows up that's just like
1:48:09 you want to make a 30-minute film just
1:48:10 describe it
1:48:11 [Laughter]
1:48:14 here so I so I think I think the
1:48:17 multimodal goes full multimodal video
1:48:20 audio
1:48:21 3D worlds 3D objects probably some code
1:48:25 like in integrated into there Dynamic
1:48:31 interfaces words
1:48:33 obviously um you know music all of it
1:48:38 you know pictures ability to see
1:48:40 pictures ability to do data
1:48:42 analysis all of that will be at our
1:48:44 fingertips in single
1:48:47 interfaces by the end of
1:48:49 2024
1:48:51 now do I expect that it's not going to
1:48:55 be janky pieces of [ __ ] the whole the
1:48:59 whole way through 2024 absolutely like I
1:49:01 don't I don't personally see any of this
1:49:05 [ __ ] like stabilizing from a technology
1:49:08 standpoint for probably 3 to 5
1:49:12 years and that that's why I think I
1:49:14 think the next 18 months so
1:49:17 the your ability to have patience and be
1:49:21 adventurous and just play and and not
1:49:24 get so [ __ ] precious about oh I built
1:49:27 a thing now it's not you know now it's
1:49:30 useless let all that [ __ ] go just start
1:49:33 playing just start like know that all
1:49:36 this stuff's going to change it's going
1:49:38 to get way better way
1:49:39 faster all you have to
1:49:42 do very simple not necessarily easy but
1:49:46 very simple all you have to do is be in
1:49:48 the conversation be actively using these
1:49:51 tools actively talking about what you're
1:49:54 learning actively you know collaborating
1:49:57 sharing saying hey where did you how did
1:49:59 you learn how to do that asking if
1:50:01 someone asks you a question how'd you do
1:50:03 that tell them don't be
1:50:05 stingy if you do that in the next 18
1:50:07 months the amount of opportunity for you
1:50:10 for all of us
1:50:12 is like Bonkers big brother will know
1:50:15 completely everything about everyone and
1:50:17 everything okay Nicholas dude dude dude
1:50:22 listen Big Brother already knows
1:50:24 everything about all of us Google and
1:50:29 Facebook asked us 20 years ago hey um if
1:50:35 I give you this cool free thing will you
1:50:37 tell me everything about yourself and
1:50:39 not read the terms and conditions and we
1:50:41 said absolutely let's do
1:50:44 this they already have all the data if
1:50:48 you don't think Google doesn't know
1:50:50 every [ __ ] thing about you you're
1:50:54 high here's what's
1:50:57 different it's not just big brother that
1:51:00 can do all the data analysis now right
1:51:02 now you have to be Google or Facebook to
1:51:05 be able to collect that much data and
1:51:07 process that data in any kind of
1:51:09 meaningful
1:51:11 way what AI does is it democratizes all
1:51:15 of
1:51:16 that so this isn't big brother this is
1:51:19 little brother and little
1:51:21 sister and and little pissed off genen
1:51:25 zir that you were a jerk to at work are
1:51:28 going to be able to do all sorts of
1:51:30 nasty things with your data and all
1:51:32 sorts of powerful things with your
1:51:34 data it's not it's not that they're
1:51:37 going to suddenly gain our data they've
1:51:39 already got our data it's that everyone
1:51:41 can process all the data like that
1:51:44 that's what changes I think we can sit
1:51:47 in VR all day and AI will do everything
1:51:49 for us I yeah I mean we we we can do
1:51:53 that
1:51:55 I I don't think it's going to be like
1:51:57 that I
1:51:59 just I think all the all the doomsdayers
1:52:02 right now they're like well all these
1:52:04 people are going to lose their jobs and
1:52:05 then they're just going to sit around
1:52:07 picking the dingleberries out of their
1:52:08 ass and I'm like have you met
1:52:12 people people like want to do
1:52:15 [ __ ] people want to contribute people
1:52:18 want to
1:52:19 engage so yeah they might lose their job
1:52:22 now if they lose their job and they're
1:52:25 like well AI cost me my job so I'm never
1:52:27 going to learn it [ __ ] ai ai is an
1:52:32 [ __ ] well that's the exact The
1:52:37 Identical response that a lot of people
1:52:39 said when the steam engine came
1:52:42 out it's just horses and plows for me
1:52:45 I'm not going to use one of those new
1:52:46 fangled machines you hear how loud they
1:52:48 are boom
1:52:51 and then tractors took
1:52:54 over and those people were still pissed
1:52:57 off the whole time tractors took over
1:52:59 but tractors still took
1:53:09 over
1:53:11 so so the shit's coming whether you want
1:53:13 it or not I still think Humanity has a
1:53:16 strong Line in the Sand regarding Tech
1:53:19 yeah they well they do
1:53:21 there's still a lot of fear right now
1:53:22 like again again Source Camp we're still
1:53:24 at I think we're still at single digits
1:53:28 percentage wise of people that are doing
1:53:31 anything remotely like what what the AI
1:53:35 salon and the Irregulars do
1:53:38 here I think very few people now it's
1:53:43 growing but I think very few people
1:53:45 relatively are using these
1:53:48 tools so we're so so early still still
1:53:51 so
1:53:53 early but you know how it is Source Camp
1:53:55 you've done this enough now where when
1:53:57 people have their Kevin McAllister
1:53:59 moment there I didn't know it could do
1:54:02 that when people have that moment it's
1:54:05 it's hard to come back from that right
1:54:08 like like most people try chat GPT
1:54:11 they're like oh let me type in a little
1:54:12 something I don't like what it gave me
1:54:14 and then they just
1:54:19 stop
1:54:21 or like the guy that hosted the the the
1:54:23 Thunder viiew dinners that I spoke at
1:54:25 the other night or he's tried it like 25
1:54:27 or 30 times and he's he's sort of
1:54:29 holding it to the standard of unless
1:54:31 it's as good as a senior copywriter
1:54:34 senior creative director it's a piece of
1:54:36 [ __ ] and he's just got this hard Line in
1:54:38 the Sand where it's like until it's
1:54:41 perfect he's like [ __ ] it it's it's
1:54:42 useless as opposed to looking at well
1:54:45 what can it do what can it do where's
1:54:47 the boundary and then let me play in the
1:54:50 area where it can it can do [ __ ] which
1:54:53 is what this group is which is what the
1:54:54 salon
1:54:56 is I'll tell you what it's a fuckload
1:54:59 more
1:55:00 fun focusing on what it can do than
1:55:04 focusing on what it can't
1:55:06 do cuz then you just sound like a bitter
1:55:09 cynical old
1:55:13 man you know get off my lawn you stupid
1:55:17 AI kids you with your your dumb machine
1:55:21 what next you're going to wear goggles
1:55:23 out in
1:55:25 public apparently
1:55:31 apparently oh
1:55:34 man so very happy I stumbled on your Tik
1:55:37 Tock last year it's so great to be a
1:55:40 part of a curious Community thank you
1:55:42 Debbie I yeah I agree I
1:55:48 mean I think it's important
1:55:51 I think it's
1:55:52 important I
1:55:55 mean I can't keep up with this [ __ ] get
1:55:59 off my lawn
1:56:01 exactly get get get off get off my
1:56:04 computer you stupid AI kids your stupid
1:56:09 AIS oh I suppose this picture was made
1:56:11 with AI you know how many comments I'm
1:56:13 getting I'm I'm I'm reading chapters of
1:56:16 my I just did this book AI Futures and
1:56:18 Anthology and so I'm reading my my essay
1:56:21 I broke it into little sections and I'm
1:56:23 I'm putting them on my Tik toks like I'm
1:56:25 getting a surprising number of people
1:56:27 asking how much of this was written by
1:56:31 AI uh could you give me a percentage of
1:56:34 uh how much of this is
1:56:38 AI no I
1:56:40 can't and it does it matter like read it
1:56:43 if you think it's a piece of [ __ ] it's a
1:56:45 piece of [ __ ] if you think it's good
1:56:47 it's good if you think it's mediocre
1:56:49 it's mediocre
1:56:50 does it matter do you care if I use
1:56:53 Microsoft Word versus Google Docs do you
1:56:55 want to know if I used
1:56:58 grammarly do you want to know how many
1:57:00 words were spell checked
1:57:03 no well tell me tell tell me how much
1:57:06 was used with AI
1:57:08 why what do you
1:57:10 care bought your book Thank You Teresa
1:57:13 that's
1:57:15 sweet yeah you can get it on Amazon if
1:57:17 you search for AI Futures and
1:57:21 Anthology Cindy [ __ ] who's an irregular
1:57:24 in here is in it Peter Kaminsky who runs
1:57:26 the AI 101 guild at the AI Salon has got
1:57:30 an essay in
1:57:36 it AI reviewed my story and informed me
1:57:39 about focusing on character thoughts and
1:57:40 adding scenery details I'll tell you
1:57:42 what AI if if you haven't been playing
1:57:44 with AI a lot what get what gets all the
1:57:47 attention with AI is the
1:57:51 um the generating stuff like write me a
1:57:53 poem write me a this write me of that
1:57:56 what's kind of the unsung hero of AI
1:57:58 it's really good at analyzing other
1:58:01 writing and like telling you how to
1:58:03 improve it and what's what it's doing
1:58:05 structurally
1:58:07 and it's really good at that it's really
1:58:10 good at that so yeah that's a really
1:58:12 good use case curiosity equals
1:58:15 intelligence yes yes yes bye guys I need
1:58:18 to step out steo
1:58:20 peace out
1:58:21 buddy thank you for inspiring
1:58:26 people steo sto's got some really
1:58:29 interesting uh prompting that he's doing
1:58:32 some really interesting image generation
1:58:33 that he's doing and he's inspiring other
1:58:35 people to
1:58:37 do you know do stuff Ann Murphy I read
1:58:41 it and love oh great thank you very much
1:58:43 an I appreciate that I use grammarly
1:58:46 pasta how much do you use grammarly of
1:58:49 the last paragraph that you sent how
1:58:51 many of those sentences were corrected
1:58:52 by grammarly we'd really like to know
1:58:54 that because we use anti-g grammarly
1:58:56 detectors here at this company we want
1:58:59 you we want you learning drunken whites
1:59:01 all on your own and if you can't write
1:59:04 your own sentences with grammar
1:59:06 Perfection we're not interested in what
1:59:08 you have to
1:59:11 say that's what it sounds like to me now
1:59:14 anytime I hear someone say well how much
1:59:17 is AI is in that it's like [ __ ] Nails
1:59:20 on a chalkboard to me right
1:59:25 now the actual answer of how much of
1:59:28 this is AI is none of your [ __ ]
1:59:37 business that's the
1:59:39 answer I'm not [ __ ] apologizing for
1:59:43 how quickly I got it done I'm not
1:59:45 apologizing for using a
1:59:48 tool I don't apologize for using digital
1:59:51 photography I learned how to do
1:59:54 photography developing in a dark room in
1:59:56 college I could choose to do that if I
2:00:00 wanted to do it
2:00:03 artistically but that's not a
2:00:04 conversation anymore it used to be a
2:00:07 conversation real photographs happen in
2:00:10 chemicals digital photographs those
2:00:13 aren't those aren't
2:00:16 photographs same [ __ ] same [ __ ] right
2:00:19 now
2:00:21 now the argument right now is well
2:00:23 generative AI can generate it you know
2:00:25 it can generate deep fakes and
2:00:27 disinformation and misinformation
2:00:30 well we were pretty [ __ ] good at
2:00:32 disinformation and misinformation with
2:00:34 just traditional computer tools so so we
2:00:37 we've been doing that we've been
2:00:39 successful at that for a while do you
2:00:41 know how far off we are in the AI
2:00:43 Community from in-person meetups yes our
2:00:47 first set of in-person meetups
2:00:51 are
2:00:52 next Tuesday this coming Tuesday in
2:00:56 Denver and then we put if you go to the
2:00:59 AI Salon if you're interested in hosting
2:01:02 a
2:01:02 local we're calling them AI Salon after
2:01:06 hours and there's social Gatherings
2:01:09 because we figure we figure at this
2:01:10 point there's a lot of groups like in
2:01:12 San Francisco there's a lot of groups
2:01:14 meeting and talking about AI already may
2:01:16 I quote you you can absolutely quote
2:01:18 me um
2:01:21 so what we want to do is is invite
2:01:24 people from other AI groups to just come
2:01:26 hang out together and in in a more
2:01:29 social kind of way like get to know each
2:01:30 other which is one of the things that we
2:01:31 want to do in the AI Salon which is one
2:01:33 of the reasons that we encourage members
2:01:36 to you know put put your pictures up
2:01:39 there and share your
2:01:42 um you know share who you are use your
2:01:45 real name because we want people to get
2:01:47 to know one
2:01:48 another
2:01:50 um in the AI salon on the le- hand side
2:01:54 so so in the salon info area there's now
2:01:56 a volunteer interest
2:01:59 form so if you're interested in hosting
2:02:02 a local um a local AI Salon after hours
2:02:07 or if you're interested in you know
2:02:09 supporting the salon in other ways we've
2:02:11 got a bunch it's a form that you know
2:02:14 you you tell us how much time you have
2:02:16 and and if if there's any times of the
2:02:18 day you can't do it what your top three
2:02:21 you know interest areas are and go do
2:02:25 that and then yes and then the the the
2:02:28 non
2:02:29 Denver um meetups are going to start as
2:02:32 soon as two weeks after that so so we
2:02:34 have our AI Salon meeting is every two
2:02:37 weeks and then the in-person meetups are
2:02:39 going to be the Tuesdays in between um
2:02:43 and and basically as soon as local you
2:02:46 know people who have volunteered to be
2:02:48 you know local Meetup hosts feel like
2:02:51 they can get a Meetup together that's as
2:02:53 soon as they'll happen so starting soon
2:02:55 probably within a month so I'm I'm
2:02:58 excited about that oh my I've been
2:03:00 squashed by my oh wait count on me all
2:03:03 right Mr it awesome what was the uh you
2:03:06 were squashed by
2:03:09 what that didn't sound
2:03:11 [Laughter]
2:03:18 good
2:03:23 some crazy ass weather
2:03:26 coming I've been squashed by my water
2:03:29 cooler teacher about AI should I tell
2:03:34 should I
2:03:37 tell about what
2:03:43 Photoshop oh that's
2:03:47 funny good show tonight My Le forever
2:03:51 hooked but this peasant has to till some
2:03:53 fields on tomorrow and needs rest good
2:03:56 day forever hooked good day to you sir
2:03:58 good day don't forget Kyle print shop
2:04:01 killed the newspaper industry exactly it
2:04:05 this is it's all the same
2:04:07 [ __ ] yeah but this is different it's
2:04:09 always different it's always
2:04:13 different the printing press was heresy
2:04:16 right like like some technological
2:04:18 change just people die over there are
2:04:21 riots
2:04:22 over wasn't it Looms in San
2:04:25 Francisco automated looms there was some
2:04:29 sort
2:04:30 of fabric people Riot I don't
2:04:35 know um have you made a list of
2:04:39 essential tools that we should subscribe
2:04:41 to Nicholas yeah I have so if you go to
2:04:43 the AI salon and I don't care if you
2:04:46 subscribe to them or not but you should
2:04:48 at least try them
2:04:50 so if you go to the AI Salon in the
2:04:52 community section go below showand tell
2:04:54 and news there's AI tools and
2:04:57 resources and and what's cool is there's
2:04:59 add your favorite tool so anyone in the
2:05:01 community can add a tool here here's
2:05:03 Cool Tools I think people should look at
2:05:05 and then I personally maintain this list
2:05:09 the AI salon tool
2:05:11 recommendations and in my
2:05:16 opinion if you're at least aware of what
2:05:19 all the tools on this list do you're
2:05:22 going to be ahead of
2:05:23 99.9% of people on the planet and if you
2:05:27 can dig deep in sort of at least one of
2:05:30 these per category so so I would
2:05:33 say anyone that considers themselves AI
2:05:37 literate at this point should should be
2:05:39 pretty pretty [ __ ] good with chat jpt
2:05:42 at least understand what all the
2:05:44 capabilities are and be able to do
2:05:47 parlor tricks with them um
2:05:51 you should probably be good with
2:05:52 conversational AI like Pi or chat GPT
2:05:56 voice with their app you should probably
2:06:00 be good with at least two or three image
2:06:02 generation tools you should probably
2:06:05 understand the difference between Runway
2:06:08 ML and Pika and he Jen and did who are
2:06:12 they're different kinds of tools that do
2:06:14 different things you should probably
2:06:16 understand 11 labs and voice synthesis
2:06:19 and then you should probably understand
2:06:21 tools likeo and refusion that make
2:06:24 songs and then you should probably get
2:06:27 your head a little bit around um 3D
2:06:30 worlds and objects and then just follow
2:06:32 smart people I've got four smart people
2:06:34 here there's a ton more but those are
2:06:36 four to start with so yeah and whether
2:06:41 or not you pay for any of these that's a
2:06:42 very personal choice if you want to run
2:06:45 down one of these rabbit holes
2:06:46 particularly when you get into
2:06:48 multimedia like video and audio it takes
2:06:51 it takes quite a bit of uh they're
2:06:53 they're they're not as
2:06:55 cheap
2:06:57 so but all of them have some sort of
2:07:00 free thing that you can mess around with
2:07:02 I love P oh my God I know isn't Pi
2:07:04 amazing and Pi the experience with pi
2:07:08 here's here's the
2:07:12 thing meta is on version three of their
2:07:16 VR goggles right
2:07:19 Apple comes out with these Vision Pro
2:07:22 things there is just something different
2:07:26 about the Vision Pro from a technical
2:07:28 perspective they're essentially the same
2:07:30 [ __ ] screens in front of your eyes
2:07:33 cameras everywhere sensors everywhere
2:07:35 some sort of
2:07:37 controllers they're technically
2:07:39 essentially the same [ __ ] except Apple
2:07:42 being Apple said no no this is a
2:07:45 completely different thing and you're
2:07:46 going to experience it in a completely
2:07:48 different way Pi the app from inflection
2:07:53 Pi is designed to be a conversational
2:07:56 chat
2:07:58 GPT it's a completely different
2:08:00 experience than if you type with it it's
2:08:03 very frustrating to type with pie
2:08:05 because it's designed to talk to but
2:08:08 when you talk with it it's it's a
2:08:10 completely different thing it's a
2:08:11 completely different skill I bet there
2:08:13 are going to be people that get good at
2:08:15 talking to things like pie and can turn
2:08:18 that into a career cuz it's a different
2:08:21 skill chat GPT says it was Lancashire
2:08:24 England that had oh the power loom Riot
2:08:27 there you go in
2:08:30 1826 yeah Power looms how dare you I I
2:08:35 shall I shall operate the loom with my
2:08:37 feet and nothing more occasionally I I
2:08:41 will reach my hand and turn the wheel we
2:08:45 shall never allow the steam engine to
2:08:48 turn the power
2:08:51 Loom Riot Riot take your strong feet run
2:08:56 out in the street and smash the power
2:08:58 looms which is what I think they did
2:09:00 they went and smashed all the power
2:09:06 looms and I guarantee you the arguments
2:09:09 were exactly the same well the power
2:09:11 looms aren't as good as real people you
2:09:13 need you need real Loom operators
2:09:16 because you can't make highquality
2:09:18 fabric with an automated process I
2:09:20 guarantee you the conversations were
2:09:22 identical if you replace power loom with
2:09:26 AI I bet you could take the the
2:09:30 transcripts of the of the newspapers and
2:09:33 the conversations of the day from 1826
2:09:36 in England I bet you could just swap out
2:09:39 power loom for AI and they'd be almost
2:09:41 identical to conversations we're having
2:09:44 [Laughter]
2:09:47 today I'm using AI to help me generate a
2:09:50 course using my prior GPT conversations
2:09:53 and it's rewriting all my stuff with a
2:09:55 much better what voice did it say yeah
2:09:57 much better voice that's super cool
2:09:59 really cool use
2:10:01 case very cool use case all right people
2:10:04 listen sounds like people are a bit a
2:10:06 bit tired it's a long weekend
2:10:09 um this is
2:10:12 a I was I was a bit more just relaxed
2:10:15 tonight about what I wanted to cover cuz
2:10:17 I felt like we went kind of deep last
2:10:20 night into
2:10:21 tools um that announcement about Google
2:10:26 Advance Gemini Advance may show up
2:10:28 Wednesday I have a
2:10:30 feeling um if that comes that's all
2:10:33 we're going to be talking about for a
2:10:35 while unless it
2:10:39 sucks but Bard right now doesn't suck so
2:10:43 it's not going to be worse than their
2:10:46 medium level tool so it's just going to
2:10:51 be is
2:10:54 it you know is its
2:10:57 usability interesting enough and
2:10:59 different enough that we're going to
2:11:00 have to spend some time learning how to
2:11:02 use it and if that's the
2:11:04 case then we probably have at least two
2:11:09 weeks probably the month of February
2:11:11 where we're really digging into a new
2:11:13 tool and then we don't know what else is
2:11:15 going to be released if some crazy video
2:11:18 tool comes out or some crazy ass other
2:11:20 large language model comes out or you
2:11:23 know pie can all of a sudden make
2:11:25 pictures or songs we don't we don't know
2:11:26 what's coming o top of the hour yeah
2:11:29 let's let's I'm gonna I'm going to wind
2:11:31 it up a little early it's Sunday night
2:11:33 I'll let you guys go um I'll be back
2:11:37 tomorrow trying to think if there's
2:11:38 anything going on I don't think so I
2:11:42 don't have anything workwise or I don't
2:11:44 think I have anything coming up this
2:11:45 week where I'm not going to be doing
2:11:46 these so so I'll see see you tomorrow
2:11:49 night um okay if you want to support the
2:11:51 channel follow me like like do it now
2:11:57 please please it's the only way I know
2:11:59 people love
2:12:01 me um you can subscribe to the lives
2:12:05 that'll support the channel you can buy
2:12:06 I've got a video series in one of the
2:12:08 corners here I think it's there called
2:12:11 underhyped that talks about
2:12:14 um why this AI stuff's a big deal a lot
2:12:17 of the ideas that are some of those
2:12:19 videos are also in the book the essay in
2:12:21 the book that I wrote so if you go to
2:12:23 Amazon uh the book is called AI Futures
2:12:26 and Anthology that I'm one of the
2:12:28 contributing authors
2:12:31 to
2:12:33 um and and really the way that you can
2:12:35 really support this community is like
2:12:37 keep showing up keep showing up here
2:12:40 join the AI Salon in a in
2:12:42 a generous way like show up with your
2:12:46 curious adventurous self
2:12:49 even if you're scared even if you're
2:12:51 like oh God they they know more than I
2:12:53 do it's submitted Sherry D
2:12:55 congratulations Sherry D submitted her
2:12:57 video to the runway
2:12:59 ml um 48 hour challenge or whatever the
2:13:03 hell it was called competition congrats
2:13:05 that's that's really exciting it was
2:13:06 really good hey first of all the song is
2:13:09 really good and second of all your video
2:13:10 is just beautiful it was like it's
2:13:13 storytelling here's a secret people with
2:13:16 this AI [ __ ] if you don't wanted to see
2:13:18 masturbatory and weird tell a story with
2:13:22 it and you can tell a story with a
2:13:24 single image by the way I don't know if
2:13:25 you knew that well how do you do that I
2:13:28 don't know you think about the story you
2:13:29 have a point of view and you make sure
2:13:32 that that image tells that
2:13:34 story I did it with my Kyle Shannon
2:13:37 dreams thing where they were starting to
2:13:38 tell stories so much that I thought I
2:13:41 want these characters to tell their
2:13:42 story so I created this thing called
2:13:44 Kyle Shannon dreams where I actually
2:13:46 wrote the stories of the characters tell
2:13:50 stories if it's just images they're
2:13:53 pretty but they ultimately fall flat
2:13:56 when you tell stories they connect with
2:13:59 people Clint Copperhead everyone in
2:14:02 regulars and mods present
2:14:04 Turtles all those other am aminals the
2:14:08 aminals of the AI learning
2:14:12 lab going to bed oh forever hooks got a
2:14:16 cold that sucks
2:14:22 more than a question about your watering
2:14:24 eyes tonight in the comments hope
2:14:25 everything's okay yeah I'm fine I'm fine
2:14:30 I mean I haven't been I haven't been
2:14:32 crying although although a couple of the
2:14:35 early Grammy the the early Grammy things
2:14:37 I might have been a little Misty eyed so
2:14:39 I don't think so maybe I've got some
2:14:41 allergies or something going on but not
2:14:43 nothing that I know of I started making
2:14:45 a story last week and it turned into a
2:14:47 song yeah there you you go I mean
2:14:49 stories is stories you you you express
2:14:52 them in a lot of different ways songs
2:14:54 songs are one of those
2:14:56 ways and that that's actually an
2:14:58 interesting I cried once last
2:15:01 year I I over index on the crying I I
2:15:05 cry at commercials so I I cry all the
2:15:08 time
2:15:09 so it's not a it's not a it's not a
2:15:12 thing I worry about but thank you
2:15:15 for having concern
2:15:19 all right everybody I'm getting out of
2:15:20 here so thank you all V dancer good to
2:15:23 see you please add new merchandise yeah
2:15:26 I know we got to do that I got to get my
2:15:27 [ __ ] together with that Apple user good
2:15:30 to see you thank you for subscribing
2:15:32 Tracy good to see you Source Camp as
2:15:34 always Wonderbread I used to cry all the
2:15:37 time too for sure now I've had that
2:15:39 taken care of share bear thank you so
2:15:41 much I appreciate that thank you V
2:15:43 dancer Lord digital Gods good day to you
2:15:46 sir good day good day good day jrc Mary
2:15:51 Mary t-rock thank you Clint Copperhead
2:15:54 as always thank you quick take J Kim
2:15:56 katkin from
2:16:00 Cleveland oh man Megan good to see you
2:16:02 thanks for being a subscriber YouTube
2:16:04 Brandon appreciate it Source Camp great
2:16:06 to see you Sherry D congratulations on
2:16:08 the video that's really exciting um I
2:16:11 assume you've done it Sherry but if you
2:16:13 haven't share it probably in Shameless
2:16:16 self-promotion and showand tell
2:16:18 share your video and then share if if I
2:16:21 assume we can go vote for [ __ ] let us
2:16:24 know where we should go vote we'll go
2:16:25 vote you up because uh as a community we
2:16:29 got to stick together Corey sandley
2:16:31 Sandler poter I added a story to a
2:16:33 recent post in Show and Tell oh that's
2:16:36 great Kyle we are so alike night
2:16:41 night we are we got we got the we got
2:16:45 the yeah we just go I like you I'm a fan
2:16:49 so creative roots dang it I always get
2:16:52 here late yeah it's they're up on
2:16:55 YouTube just keep showing up you'll get
2:16:57 here Bob Hamilton good to see you papa
2:16:59 JJ good to see
2:17:01 you um Marina good to see you all right
2:17:05 Mr it think I got everybody at least the
2:17:08 Irregulars Tracy all right I'm out of
2:17:11 here have a fantastic night everybody
2:17:13 and uh I will see you tomorrow uh 8:00
2:17:16 Mountain should should be on time if if
2:17:18 not I'll make a video and give you a
2:17:19 timer all right peace out thanks
2:17:22 Debbie