12/20/2024 - AI Festivus is Coming! FREE AI Workshops, OpenAI's O3 Breakthrough, Lady Kyle's Debut

Description

Chapters

Transcript

0:01 [Music]
0:19 [Music]
0:33 [Music]
0:44 good evening good people happy Friday
0:45 night date night champ Shannon in the
0:47 house you ready to sing what do you want
0:49 to sing what do you want to sing
0:53 [Music]
0:58 [Applause]
1:00 [Music]
1:02 [Applause]
1:03 [Music]
1:11 s this lonely
1:14 town when things are going to
1:19 change BR my life
1:22 away seems dreams to Bunch
1:28 class get my ner
1:31 but my past is pulling me
1:34 [Music]
1:39 down black she sck
1:44 [Music]
1:48 around
1:50 me you go home
1:54 again W you
1:56 leave say than to Ste me way yeah from
2:02 the truth of who I am and what I believe
2:06 so I thanked him for his two sents with
2:08 a
2:09 handshake and some
2:11 sympathy yeah back up my blue
2:16 jeans and head it for this big
2:20 grass of my
2:24 freedom
2:26 byebye Black Sheep Champion just left he
2:30 just walked away he got a new
2:32 raccoon he got a new raccoon toy he just
2:35 walked away started playing with his
2:38 toy he's like [ __ ] it I got
2:41 toys uh what was that Source Camp you
2:44 almost broke away from your Caribbean
2:46 vacation to come to office hours I think
2:48 he made a good
2:49 call I oh I tried so hard to join office
2:52 hours but the universe was not
2:54 cooperating what did I miss okay I'm
2:56 going to talk about it a fair amount I'm
2:59 G to I'm going to start off tonight
3:00 talking about day 12 um I actually had a
3:05 real breakthrough
3:08 today so we'll talk about it we we'll
3:11 talk about it
3:13 [Music]
3:15 I I have some I have some insights I
3:17 have some thoughts I have new thoughts
3:20 um I also want to go find there's a
3:22 tweet or two that I want to
3:25 find um
3:39 all right that's the one I wanted to
3:43 show and then I want to find another one
3:45 hang on
3:46 people I don't even know what he's
3:48 talking about he seems to be mumbling to
3:50 himself and looking at some screen it
3:52 seems kind of rude you know it's like
3:54 those people on Zoom calls where they
3:55 pretend like they're listening to you
3:57 but they're really checking their email
3:58 or I don't even know what they're doing
4:00 doing something else not even email
4:03 Isn't that rude so yes that's what I'm
4:06 doing hang on calm down calm down happy
4:10 Friday night date night this would be a
4:12 perfect time to start heating up your
4:14 nachos uh if you have hot pockets and a
4:17 dirty microwave those are also accepted
4:19 snacks for Friday night date night don't
4:22 forget to use the sleeve when you're
4:25 cooking Hot Pockets otherwise they're
4:27 just they're
4:29 tough you want them to be crispy
4:33 tasty
4:34 right you want it like Frozen on the end
4:37 that you bite and then when your teeth
4:39 sink into the middle it scalds your
4:41 tongue that's comes from the
4:45 sleeve
4:49 um oh let's see Arc AGI
5:00 today was both a dramatic
5:04 day and a nothing
5:07 [Laughter]
5:12 [Music]
5:16 day I've been without chat gbt for six
5:19 days and I'm having serious withdrawals
5:21 oh I bet I can't I I am seriously at the
5:24 point right now where if chat GPT goes
5:26 down I'm like I just I just check out
5:28 I'm like [ __ ] it I'm not doing work I'm
5:30 not gonna I'm not going to rely on this
5:32 stupid [ __ ] brain you know what's so
5:35 funny like I feel
5:37 like I feel like the the fear if if you
5:41 don't know AI the fear is like oh my God
5:43 the robots are gonna do all the thinking
5:45 and I won't have to use my brain and the
5:46 reality is when the robots start doing
5:48 all the thinking you're like oh that's
5:50 really exciting and then when you try to
5:52 go back to use your brain what a piece
5:55 of [ __ ] you can only think like one or
5:58 two things at a time it's it's you gotta
6:02 like actively like you can feel the
6:04 calories being burned like trying to
6:07 shift one thought to another and
6:09 organized thoughts and you're like oh
6:11 God I can't do it and I know I'll go to
6:14 the Whiteboard then you walk over to the
6:15 Whiteboard and you're like well which
6:18 color pen should I use because that's
6:19 gonna make a difference and then you
6:21 pick up the blue one and the whole time
6:23 you're writing with the blue and you're
6:25 like I think I should have chosen the
6:26 red one and like that's what our [ __ ]
6:29 brain is well that's what mine is you
6:31 might have a functional prefrontal
6:33 cortex if you do good on you you
6:35 probably have a nice family too that
6:37 loves you and it's
6:38 supportive that's just annoying
6:41 anyway that's our
6:44 brain and we're like we've got the
6:48 [ __ ]
6:49 hubris to go we're smarter than a
6:54 computer we used to be
7:00 give that [ __ ] up it is
7:03 liberating but to your point source Camp
7:06 don't go on a Caribbean vacation confirm
7:09 that you can get to chat GPT before you
7:11 go on
7:12 vacation because if you if you want to
7:14 act like you're
7:16 [Music]
7:19 smart you gotta have access to this
7:22 [Music]
7:25 [ __ ] it's actually really interesting
7:27 Source Camp because um p Pate was on
7:30 office hours today and the thing we we
7:33 talked a little bit about who who was
7:35 excited about 2024 what were people most
7:38 excited about and Pate who works at
7:40 Google and is a mathematician and works
7:42 on inference right which is when you
7:46 give chat GPT a prompt and it answers
7:48 you that's
7:49 inference
7:50 um the uh he said he's all excited about
7:55 models getting smaller and being able to
7:57 be used on edge devices like phones and
8:00 things like that so so that's where he's
8:02 excited for things going okay I
8:05 understand that Ann Murphy's in the
8:06 house Ann Murphy you in the house what's
8:09 what's going on what's going on lady
8:12 however I did use the camera features to
8:14 give me a walking tour of downtown Aruba
8:17 that's pretty
8:19 [Music]
8:20 cool okay let me let me let's let's call
8:23 an audible since we got Ann Murphy in
8:25 the house here let's go let's let's
8:28 let's do some [ __ ]
8:31 I think this is it yeah
8:33 okay all right people so Ann Murphy if
8:37 you've not met an Murphy I'm in the
8:39 house
8:42 weirdo you see you see you see the kind
8:45 of respect I get
8:47 weirdo I'm not weird I'm
8:50 [Music]
8:52 irregular what's good for the goose
8:54 right um okay so last year right around
8:59 this time an Murphy and I had the wild
9:01 idea why don't we teach people how to
9:04 build gpts for 24 hours and I think we
9:07 started it like like right around now
9:09 like a week before we were going to do
9:10 it and we did it and this community came
9:13 together and built 150 gpts super cool
9:17 so we decided this year to do another
9:19 event so a week from tonight and a week
9:23 from tomorrow all
9:26 day we've got what do we have 34 or 35
9:30 speakers every hour starting at 9:00
9:33 a.m. Pacific and going till 8:00 P.M or
9:35 900 p.m. I don't even know yet I haven't
9:37 seen the final schedule um but basically
9:40 11 or 12 hours of workshops every hour
9:43 is a different Workshop some are on
9:45 business some are on creativity some are
9:47 on you know how to do [ __ ] some are on
9:50 Building
9:52 Technology for for two days
9:55 straight between Christmas and New
9:57 Year's right so
10:02 this is the site aif festiv us.com come
10:05 here and register go there right now it
10:07 doesn't cost
10:08 anything this is this thing's going to
10:10 be off the hook so we've got like a
10:12 hangout room so if there's if there's
10:14 workshops you don't want to do you can
10:16 just hang out and talk to people and
10:17 meet people and ask
10:19 questions um we're doing a um a GPT
10:24 what's it called gift
10:26 bag but basically if you've got a GPT
10:29 you like 34 speakers 24 hours of
10:32 workshops for
10:35 free it's gonna be
10:38 amazing um it's just gonna be a lot of
10:41 fun and it's just like it's like who
10:43 gives a [ __ ] like let's just get
10:45 together we'll hang out you'll learn a
10:47 bunch of stuff you'll get completely
10:49 exhausted that's the whole idea it's AI
10:53 you should be tired if you've got your
10:56 wits about you come to AI Fest of us and
10:59 we will destroy that for you you're
11:02 [Music]
11:03 welcome if like most of us if like if
11:06 like the
11:08 87.6% of American families that are
11:11 massively
11:14 dysfunctional you all all you want to do
11:17 on the couple of days after Christmas is
11:19 run away this is your
11:22 excuse I got a work
11:25 thing is now a legitimate excuse sorry I
11:27 got a work thing and then come hang out
11:30 with us we'll be your family we're
11:33 supportive we're empathetic we're
11:34 compassionate we're cool well ands cool
11:39 I'll be
11:46 there oh man all right we have F oh yeah
11:50 that's the other thing we got we got
11:51 five scholarships donated today so we
11:53 have five scholarships to the AI build
11:56 lab um from Tyler Fisk to give away
12:00 um $1,200 workshops he's giving away
12:03 five of them all right
12:06 so this event is brought to you by um
12:10 she leads AI which Ann runs and the AI
12:13 Salon which I co-lead with uh with Leah
12:16 faston all right so if you don't know
12:19 the salon go
12:21 to the salon. AI oops let me do this for
12:25 you
12:27 guys hell you are my family already well
12:30 so we're already dysfunctional so you're
12:37 good oh yeah a l and the salon are my
12:41 go-to um excuses there you go so go to
12:45 the salon. click on join our community
12:48 that's it's also
12:50 free it's also
12:52 free so go sign up for festiv us this is
12:55 a two-day event the AI Salon is an
12:57 ongoing community it's like
13:00 2100 Ai optimists and Builders and
13:04 creators and just nice people all
13:07 right so that's what's
13:09 happening and what did I miss did I miss
13:13 [Music]
13:17 anything
13:20 Shaina my job gave me that week off I'm
13:23 super grateful to have a new work thing
13:26 there you go exactly say say listen
13:29 family sorry sorry my job gave me the
13:30 week off but I got this i got this work
13:32 thing it's a it's a
13:34 two-day seminar I got to go learn AI
13:38 sucks and I just come hang out learn
13:40 cool
13:43 [Music]
13:46 [ __ ] my iPhone thinks Kyle is a family
13:50 member that's just
13:55 sad oh man well an it's good to have you
13:59 in here um an is so for one thing an is
14:04 is doing the Lion Share of the work on
14:08 getting this thing together she has
14:10 assembled a whole crew to to do it and
14:15 get people we've got we've got speakers
14:17 coming from four different communities
14:20 from she leads AI from the AI Salon from
14:23 Rachel Woods um AI exchange and from
14:26 Paul rater's marketing AI Institute um
14:30 it's just going to be amazing it's just
14:31 going to be amazing like you get to hang
14:33 out with people that are super
14:35 passionate about this stuff all right so
14:38 here's the other thing don't just sign
14:39 up for this yourself tell your flipping
14:42 friends ideally tell friends that have
14:45 their heads up their butts about
14:48 AI all right you know the ones that are
14:50 like you know Tammy I just could we not
14:55 I'm trying to eat my
14:57 ham and and stuffing and I just could we
15:00 not with the
15:02 AI I just don't I don't want to hear
15:05 about it
15:06 anymore bring
15:09 [Laughter]
15:16 her oh man man alive all right cool uh
15:22 so it looks like we just got reposted by
15:24 someone who's got a big uh a big
15:27 following so I don't know if you're
15:28 watching Robert but if you are thank you
15:30 so much for that that's really much
15:32 appreciated get some people in here oh
15:35 we got some people in here dang a bunch
15:38 of people on X okay so since there's a
15:40 bunch of new people here let me actually
15:42 show my screen my name's Kyle Shannon
15:44 this is the AI learning
15:46 lab and a week from tonight and a week
15:50 from
15:51 tomorrow for 12 hours each day Ann
15:55 Murphy and I are co-hosting and putting
15:58 on uh um AI
16:00 festivas which is a different Workshop
16:03 every hour for 12 hours for two days
16:07 straight why because we're
16:11 [Laughter]
16:13 insane so go sign up for that go sign
16:17 it's it's free by the way it's free
16:19 you're not gonna want to be with your
16:21 family right all right fantastic good
16:26 beautiful all right what else anything I
16:28 missed an did I miss anything less than
16:32 I'm off work until like January 9th 9th
16:34 I'm getting an AI labotomy I think
16:36 that's a good call
16:37 too I seriously can't believe this is
16:39 coming together the way it is AI
16:41 communities are the best so let me let
16:43 me talk let me very seriously this is
16:47 important um so if you come to something
16:51 like AI Festivus you're going to meet a
16:53 ton of people we've got the workshop
16:56 tracks so every hour there's a different
16:58 workshop and then we've got like a
16:59 hangout room and so even if you're not
17:03 learning stuff you can be hanging out
17:05 and meeting people um groups like Anne
17:08 Murphy's she leads AI or Rachel Woods um
17:12 um AI exchange or the AI
17:15 Salon these communities in my opinion
17:18 are the future of work because you know
17:21 especially given what what open AI
17:24 launched today AI capabilities are
17:28 accelerating
17:30 dramatically
17:31 fast what exactly what 03 is what it
17:35 does all that sort of stuff is is
17:37 actually irrelevant to most of us the
17:40 only thing we need to know from today is
17:42 that things are accelerating and what
17:44 that starts to look like in the future
17:46 of work is that the skills the Tactical
17:49 execution of the work you do the tasks
17:52 of the work that you do are
17:54 increasingly going to be done by AI
18:00 and so who we're going to
18:03 hire is decreasingly going to be hiring
18:08 someone for what they can do for their
18:11 skills why we're going to hire someone
18:13 is because we know them we trust them we
18:17 know that they've got a work ethic we
18:18 know that they're not an [ __ ] we know
18:20 that they are an [ __ ] but we love
18:22 their opinion
18:24 right so being in a community coming to
18:27 events like this is actually really
18:29 really important there's the yeah it
18:31 would be nice to learn some more crap
18:33 about AI but what I can promise you is
18:36 you can actually catch up to it if you
18:38 if if you're if you're still in the
18:40 fantasy land that you can keep up with
18:42 all this AI stuff you can't just let
18:44 that go this this is one of my big
18:46 insights today I can't wait to talk
18:48 about
18:51 it but being in the conversation why I
18:54 go live five nights a week is because I
18:57 want people in the conversation
19:00 and what's most remarkable about the
19:02 community that shows up here night after
19:04 night after night is
19:06 that they just want to be in the
19:10 conversation I mean I tell the same
19:13 stupid [ __ ] stories over and over and
19:15 over again they're still here why
19:16 because it's not about me it's about
19:19 them hanging out and figuring things out
19:20 and going oh I tried this and I tried
19:22 that oh that worked and that didn't
19:24 work that's going to be the future of
19:26 work so so when you start to build those
19:29 relationships is
19:30 now because in the two to three years
19:33 that are coming where the the real
19:34 disruption is going to happen it's not
19:37 2025 is going to be a little messy this
19:40 is like 2026 2027 shit's G to get
19:45 weird and if you're not in a community
19:47 you're gonna it's it's going to be
19:49 overwhelming so anyway enough soapbox
19:53 and Murphy I'm so I'm so excited to be
19:56 doing this again with you and I we're
19:59 have a blast it's going to be fantastic
20:01 so all right
20:03 cool all right so let me talk about
20:05 today a little bit I need a little I
20:08 need a little AA the voice is already
20:10 going and I've barely started
20:12 [Laughter]
20:20 screaming okay so I want to talk about I
20:24 want to talk about um today so today if
20:27 you don't know was day 12 of open AI
20:32 ship it's a p not a
20:34 t every every time time I tell someone
20:37 about open AI ship Miss they're like
20:39 [ __ ]
20:44 Miss um the thing
20:47 about um ship Miss today the shipping
20:52 product was that they announced we we
20:55 can't play with it yet we're going to be
20:56 able to play with the the Mini version
20:59 at the end of January well someone's
21:01 someone's going to be able to play with
21:03 it
21:05 um they launched a thing called 03 so a
21:09 month ago they La launched a thing
21:11 called 01 and three or four months ago
21:13 they launched 01 preview and that was
21:15 their reasoning engine it was called
21:17 project strawberry
21:18 internally and we've gone from 01 to 03
21:22 in four months um and you're like what
21:24 happened to O2 well it was someone
21:27 else's trademark
21:31 a British Telecom company called O2 uh
21:34 has that one so they decided to just
21:36 jettison 02 and go straight to 03 which
21:38 is fine whatever uh Sam mman admitted
21:41 they they are um world class when it
21:44 comes to having shitty naming strategies
21:47 over there
21:51 um so the big thing that the dramatic
21:55 thing that they announced today is that
21:57 03
21:59 actually exceeded humans on this
22:02 benchmark test called The Ark AGI test
22:06 where humans landed at 80
22:08 85% and 03 in high compute mode which
22:12 I'll talk about in a second hit
22:15 87.5% so and and and the previous
22:19 state-of-the-art models including 01 on
22:21 this Arc AGI test were not north of 31%
22:25 they were all around 25 to 30 31% was
22:29 was the highest and all of a sudden in
22:31 kind of medium mode um 03 hit 75% and in
22:36 high mode it hit
22:39 87.5% humans are at
22:41 85 so this is absolutely a milestone now
22:45 is it AGI no and Sam Alman didn't say
22:48 that but there's this test it's called
22:51 The Arc AGI test no other models come
22:53 close to beating it this thing beat it
22:56 this is a step change so if you look
22:59 at one of the guys that I I really
23:02 respect in the space is a guy named
23:06 David
23:06 Shapiro and if you look here this is the
23:10 ark AGI um semi-private V1 scores over
23:14 time so I guess that's one of their
23:17 tests
23:19 gpt2 basically didn't do anything on
23:21 this gpt3 didn't do anything on this GPT
23:25 4 I don't know what that is maybe one
23:28 and a half perc GPT 40 which is that's
23:32 last year's model right the one we got
23:34 this
23:35 year went up a little bit here's 01
23:38 preview here's 01 High here's 01
23:41 Pro and then these two at the top here
23:44 are
23:45 03 uh tuned low and tuned high so not as
23:49 expensive more
23:50 expensive um David Shapiro's response
23:55 even I'm a little bit surprised by this
23:58 leap and I'm the one who's always
23:59 harping on
24:02 exponentials um again none of the
24:05 details about what you how you use 03 or
24:09 who uses 03 is super relevant the only
24:12 thing we need to know is that high
24:16 technology like this will trickle down
24:18 to us so one of the things when I when I
24:20 first started doing all this this AI
24:23 stuff I started a newsletter called
24:25 everyday
24:26 AI because
24:29 the reason I'm passionate about AI
24:32 enough to go live five nights a
24:34 week and just spend all of my time
24:38 thinking about it is that I think that
24:40 the launch of chat GPT November 30th
24:43 2022 marks a turning point in history
24:46 because that's the point at
24:48 which Ai and machine learning became
24:52 available to non-developers to
24:54 non-engineers to non-
24:56 mathematicians right we can just use are
24:59 words and these remarkable results come
25:03 out
25:06 um what they showed today with
25:10 03 and if you tried to watch it unless
25:14 you're a mathematician it was
25:16 unwatchable right they're like writing
25:18 code to do mathematical things to write
25:21 code to run the code to check the code
25:23 to and and it was all about math and
25:25 math scores and and you're like how does
25:28 this apply to me well let let me let me
25:31 give you some perspective so there's
25:32 another chart that I think is is worth
25:35 paying attention to now this one you
25:38 can't really see it here um but what I
25:42 want you to pay attention to so so this
25:44 76% right here the AR AGI this is the 03
25:48 model that hit
25:50 76% these other models down here are the
25:54 you know some of the other models that
25:55 oh they got to
25:56 32% and then this this uh High tuned
26:00 version hit 88% this is the one that
26:03 that beat the humans the one in the
26:05 upper right look on the very
26:08 bottom this is price per
26:12 answer price per answer so
26:16 03 running at at their low tuning was
26:20 about a $20 per
26:23 answer um run so they got to 70 76%
26:28 um $20 per
26:31 shot um and and this this looks like
26:34 it's what is this exponential or
26:36 logarithmic on the bottom so so it's not
26:39 linear so the the 88% this was about a
26:44 $3,200 per
26:46 answer um um uh inference
26:52 instance one one number I said I I read
26:55 was like $1.6 million was spent for from
26:58 open
27:00 AI just in compute costs to to hit that
27:04 87.5%
27:07 number so while we were watching the
27:10 event I wasn't thinking too much about
27:12 that but afterwards here's the Epiphany
27:14 that I
27:16 had I think the number is something like
27:20 98.5% of people are not Engineers
27:24 they're not
27:26 Developers 98.5%
27:29 everyday AI the the the chat gpts and
27:32 the claws and the pie and the advanced
27:35 voice that you can talk to and Santa and
27:38 the image generation tools and some of
27:40 the video generation tools but even
27:42 those are starting to kind of Escape
27:44 into a different realm they're they're
27:46 not all just everyday AI the everyday AI
27:52 stuff for the most
27:54 part this is this is my big Insight
27:57 today and I felt this of freedom for
27:59 this
28:01 because it it struck me when 01 preview
28:05 came out that that it had surpassed my
28:09 cognitive
28:10 wall right that I realized with with
28:15 gp40 I I I gota I I have a sense of how
28:19 to use it I have a sense of how to think
28:21 about it what to ask it how to navigate
28:24 and get it to do good stuff I felt
28:26 completely clueless with 01 preview
28:28 because it's just a different
28:30 thing right it's doing PhD level problem
28:34 solving I don't have any PhD level
28:37 problems most people don't have any PhD
28:40 level problems this thing that can
28:42 almost hit 100% on a comp competitive
28:46 math challenge that's not most of
28:50 us so there's a fair amount of
28:54 advancement happening right now that
28:56 there's going to we're going to see a
28:57 lot of charts you know going up and to
28:59 the right we are in an exponential
29:01 growth phase and it's going to get
29:03 really weird but most of it we don't
29:05 have to worry about and and so so what I
29:08 want to say is if today left you feeling
29:11 a little bit I Silver Fox on X today
29:14 posted something like she felt like
29:16 nine10 of this year she was keeping up
29:18 and all of a sudden she felt like she
29:19 was in the breakdown Lane that she
29:21 couldn't keep up anymore and I I hit the
29:23 exact same thing with her and what today
29:26 did for me is it solidified the fact
29:28 that's okay it's okay that we're not
29:32 going to be able to afford
29:34 03 it's okay that we might not even be
29:37 able to know what we're supposed to do
29:39 with 01 we'll eventually figure it out
29:42 and eventually some of that Tech Will
29:44 kind of migrate down into more easy to
29:47 use and easy to grasp tools and maybe
29:49 they're pre- prompted and things like
29:51 that so anyway that was the thing I
29:53 wanted to share is that is that if
29:56 you're feeling a pressure to keep up one
29:59 just let that [ __ ] go you can't keep up
30:02 it's moving way too fast at this point
30:05 and if you're feeling like I need to
30:07 understand what all these tools do we
30:10 actually don't
30:11 now if you're one of those people that's
30:14 trying to cure cancer or create energy
30:16 or all that sort of stuff um yeah then
30:19 then great these then these tools are
30:21 going to be good for you and quite
30:23 frankly you know 3,200 bucks and answer
30:26 is probably a little bit much but if
30:28 you're doing something like solving some
30:29 big physics problem or solving energy or
30:32 trying to cure
30:34 cancer and it's solving problems at
30:37 multi- PhD
30:39 level you know level capacity well
30:43 that's probably cheap compared to hiring
30:45 a bunch of phds right so anyway that's
30:50 that okay so let me let me I want to
30:51 walk you through something since we have
30:53 some new people in the house um I talked
30:56 before about the AI Salon I want to do
30:58 little tour of
30:59 it
31:01 um if you're not a part of the salon go
31:03 to the salon. scroll down a little bit
31:06 click on a button that says um join join
31:10 the
31:10 community and when you join the
31:14 community it's going to take you
31:18 here to the
31:21 salon and here you
31:24 go um there's a welcome video with
31:26 myself and Leah faston who I found Ed
31:28 this with um we talk about
31:33 um our journey to AI literacy one of the
31:36 things that one of the misconceptions
31:38 about Ai and learning it is a lot of
31:42 people think that they need to be really
31:44 technical that they need to be really um
31:47 they need to be Engineers they need to
31:49 be
31:50 mathematicians generative AI chat GPT
31:52 and the like these image generation
31:54 tools these creative tools you just need
31:56 to use your words and and one of the one
31:59 of the best ways to just get started is
32:02 to just play just start playing so the
32:05 AI Salon is a community right now we're
32:07 about 2100 people of AI optimists and we
32:11 play first mindfully create generously
32:13 lead and that just looks like play first
32:15 is just like just play just I five
32:18 nights a week I'm here just playing with
32:20 AI and just making stupid images and
32:23 making stupid jokes and voices and I'll
32:26 occasionally put on a pink bow because
32:27 why would you not do that
32:31 right apparently I can't do it right
32:33 I've been told this is not how you wear
32:35 one of these but I don't know seems
32:37 right to me
32:39 um and then and then mindfully Crea is
32:42 as you start playing with these tools
32:45 you start to gain some skills you start
32:46 to understand oh okay I can sort of get
32:49 some interesting results out of chat GPT
32:52 and then I can do these images over here
32:53 and I I made a song over at sunno and
32:56 and now you can start to say oh I have
32:58 an idea for a project I wanted to do
33:00 since high school wonder if I could do
33:03 that with these tools and all of a
33:04 sudden you're in this mindfully create
33:08 mode and
33:10 then and then the generously lead part
33:13 is don't do it on your own like share
33:16 what you're learning share what's
33:18 working share what's not working um ask
33:21 for help this is a community that is
33:23 incredibly generous our values this is a
33:26 community full of people
33:29 that are curious curiosity is one of our
33:31 values they're playful they've got
33:34 empathy so there's a lot of people right
33:37 now that are terrified of AI that are
33:40 [ __ ] pissed off about AI it's theft
33:44 it's the world's greatest plagiarism
33:45 machine you're all
33:47 [ __ ] that fear is coming from from a
33:51 very real
33:52 place this AI stuff can be scary right
33:56 from from the outside if you just
33:58 abstract this out and you hear us all
34:01 talking about it's as good as the
34:03 average human well what's that
34:05 mean right we've got 50 years of of
34:09 Hollywood telling us the robots are
34:11 going to kill us of course people are
34:12 going to be afraid every time in human
34:15 history there's a major technological
34:17 advancement it's met with fear and
34:21 resistance right the
34:23 Lites the the the mechanical Loom or the
34:26 the steam powered Loom there were were
34:29 riots
34:32 right so so this is a group that if you
34:34 come in here and you're like I'm scared
34:36 I don't understand this stuff this is a
34:37 group that's got empathy for
34:39 that it's also a group that's Brave
34:42 they're they're adventurous they're
34:43 willing to play with this stuff and then
34:44 they're
34:45 generous you ask someone how they do
34:48 something they're willing to jump in and
34:49 help here's how he did it this AI
34:52 Festivus thing that an Murphy and I are
34:53 putting
34:54 on this is 34 34
34:59 people are going to be on panels or are
35:01 giving one-hour workshops over two days
35:05 for
35:07 free like that's this thing is the
35:10 definition of generous
35:13 leadership right on their holiday break
35:16 they're willing to teach complete
35:20 strangers how to make a python script or
35:24 how to make images or or Kelly Bosch is
35:27 coming who makes really cool videos
35:29 she's going to talk about her process
35:30 and how she thinks about it and and and
35:33 talk talk talk you through how you can
35:35 do that too it's going to be absolutely
35:38 remarkable and that's that's the kind of
35:40 people that are in this
35:43 community um so so there's it's a really
35:46 simple it doesn't cost anything to join
35:48 I know I'm kind of selling it like you
35:50 know it's $1,500 a day it's
35:56 free the most important thing I I think
35:58 you should do when you come here is
36:00 introduce yourself jump in here and
36:02 introduce yourself tell people who you
36:04 are and if you're if you're afraid if
36:06 you're like well I don't know anything
36:07 about AI they're all going to judge
36:09 me they're not we don't that's not what
36:13 this group's about it's a remarkable
36:15 group all right and then there's all
36:17 sorts of other cool stuff there's a
36:18 community area there's
36:20 news um in the news area there's a um
36:23 Salon newsomatic this is an application
36:26 that I actually built
36:28 using not a single line of code I just
36:31 went to lovable. deev and I Saidi want
36:33 to make a newsreader for the salon and
36:36 two hours later I had this really cool
36:38 thing so you can go play with that um
36:40 and then there's showand tell areas and
36:42 in the showand tell
36:43 areas um there's llms images coding and
36:46 tech music and audio video and animation
36:49 3D and World building in every one of
36:51 these there's uh a tool area that like
36:55 here's tools to go play with and so
36:57 these are all vetted tools this is not
36:59 just like incessantly long list these
37:01 are like the tools that are
37:03 good
37:06 right all right so go join
37:10 that all right I
37:12 need some hydration I need some
37:15 hydration people you know what I
37:19 [Music]
37:21 mean so last night we were playing a
37:23 little bit with Runway oh here's a
37:26 public service announcement
37:30 um pika pika pika pika if you don't know
37:35 Pika if you don't know Pika you don't
37:37 know
37:44 squat here's a video we made last night
37:47 in P Labs so P Labs is like Runway it's
37:51 like cling it's like Luma laabs
37:54 Dreamscape it's a video generation tool
37:57 um
37:59 all weekend this weekend they're giving
38:01 away free Generations like unlimited
38:05 free
38:07 Generations the thing I said before
38:09 about where you should start with AI is
38:11 you should just play one of the problems
38:13 with generative video is um it's
38:17 expensive these subscriptions are like
38:19 anywhere from like P is 35 bucks a month
38:22 which is a lot but it's a cool tool the
38:25 problem with them being all paid
38:27 subscription is you don't get a chance
38:29 to play you don't get a chance to like
38:31 like how do I get good at this the only
38:33 way you can get good at it is to do it a
38:35 lot and it and they're really expensive
38:37 so so there's there's a real barrier to
38:40 video so if you're around this weekend
38:43 like I would I would just go to p. art
38:45 if you don't have an account get an
38:47 account and you can use it for free all
38:49 damn
38:50 weekend so go get thee to a nunnery like
38:55 now get the to peek it at Art uh because
38:59 no one that else that I know is doing
39:01 that um the only other thing that's
39:02 close to this right now mid Journey um I
39:06 think you have to have the $10 a month
39:09 subscription um but they put out a note
39:11 three or four days ago that um they've
39:14 got excess GPU capacity which no one has
39:17 excess GPU capacity let just file that
39:20 for a minute they have excess GPU
39:22 capacity and so they're giving out
39:25 relaxed mode um image generation like
39:28 unlimited I think through the end of the
39:30 year so if you've got a mid Journey
39:32 account go to Mid journey and just
39:36 start blast it out just make as many
39:38 images as you can Again part of the
39:40 trick here is just play enough that you
39:42 start to understand how these tools
39:44 really work and how you can work around
39:46 the
39:47 jankiness um my
39:50 my my unfounded my speculation my chat
39:53 TMZ speculation about why mid Journey
39:56 has excess GP U capacity is I think
39:59 they're about to drop uh mid Journey
40:02 video and when they do they're going to
40:04 need some more uh gpus so I have a
40:07 feeling they spun up the hardware in
40:09 anticipation of that tool coming that's
40:12 a total speculation
40:14 so all right um so here's what we should
40:17 do let me let me take some um if you've
40:20 got questions about AI if you want me to
40:22 comment on anything um just pop them in
40:24 the comments if you're on Tik Tok We've
40:27 Got Mods over there there that'll
40:28 that'll uh pin comments if you're on
40:30 YouTube or x uh just put the comments in
40:33 uh I can see those there and um yeah
40:38 what do you want to talk about and I'm
40:40 also curious I'm also
40:43 curious what your big takeaway for for
40:46 2024 was I've I've talked about that a
40:48 couple of times today it's been really
40:53 interesting so thoughts
40:57 uh uh uh uh uh uh
40:59 [Music]
41:04 uh yes Brandon Kyle I want to talk about
41:07 DOI we can talk about DOI big Takeaway
41:09 on Google is Google notebook LM that
41:12 came up multiple times today I I find
41:15 Google notebook LM a fascinating use
41:18 case so this is a a really good case of
41:22 let me go Google notebook LM is a really
41:25 good case of this was a tool that had
41:27 been in the world for I think more than
41:29 a
41:30 year and there were fans of it but
41:33 nobody really talked about it and then
41:35 they added the audio overview they added
41:37 this podcast
41:39 maker and it was just [ __ ]
41:42 magical it just worked and like the
41:45 things it produced sounded real it they
41:48 sounded like real
41:49 people they actually did talked about
41:53 whatever the knowledge was in a in an
41:55 intelligent way and it totally elevated
41:58 that entire product that was that that
42:01 that's one of the big wins of the year I
42:03 think um yeah really really
42:07 amazing my intelligence is artificial i'
42:10 I I wrote a thing today on on X I
42:14 said I
42:16 said something like you're an advanced
42:19 AI user when you finally accept this
42:21 phrase I am the
42:25 bottleneck when you finally get that
42:27 your brain can't keep up with these
42:29 tools that now now you're starting to
42:32 get somewhere that was my big
42:33 breakthrough in 2024 was I I'm pretty
42:36 sure I'm just a dumb
42:38 [Laughter]
42:42 dumb uh for Brandon it was advanced
42:44 voice mode yep I heard that a lot as
42:48 well
42:50 um okay oh Danielle says wait this is
42:55 important um
42:58 an email from yudo I saw I saw a note on
43:00 this I forgot about this one has 100
43:02 free credits for free accounts 500 for
43:06 normal accounts and a thousand credits
43:08 for pro accounts so yeah so
43:11 yudo yudo and sunno are the two um song
43:15 generating tools um and we play a fair
43:19 amount on here with sunno um Udo in my
43:23 opinion has better vocals they don't the
43:26 vocals in udio don't have as many audio
43:29 artifacts like compression artifacts
43:31 umo's been just killing it with dropping
43:35 feature after feature after feature
43:36 after feature so
43:38 anyway all right
43:43 uh let's see what ocean has to say I'm
43:47 sure this is another deep philosophical
43:49 question quick
43:51 question how
43:53 long will it take us to apply AI to
43:56 almost everything
43:58 voluntarily and accept it whether you
44:00 like it or not live with it one years
44:02 two two years how
44:05 soon
44:15 well I I mean a a lot of a lot of AI is
44:20 already in use like a lot of AI is just
44:22 built into the tools we use right and it
44:24 has been for
44:25 decades I assume what you're talking
44:27 about is um you know how long is it
44:31 going to be until the adoption phase of
44:34 generative AI you know takes over right
44:38 um I think it's going to be a bit it's
44:41 so funny a lot of people talk about the
44:43 hype cycle of AI that oh we're in this
44:45 hype and it's overblown and it's
44:47 overhyped and it's
44:50 all I don't think we are like I don't
44:52 think the hype has begun
44:54 yet you know the the
44:58 I was there I was I was one of the the
45:01 early players in the the early days of
45:03 the worldwide web I I founded a company
45:05 called agency.com which was one of the
45:07 early digital agencies and we built a
45:09 lot of the you know early websites for
45:11 the Fortune 500 and it was just this
45:13 crazy time and I founded that in
45:17 1994 and the.com Bubble Burst in um
45:22 April of 2000 right so there was a
45:25 six-year period
45:27 um where this thing was ramping up
45:31 and as exciting as AI is and as profound
45:37 as it is for people that are playing
45:40 with it and they understand what's going
45:42 on as of four or five months ago 47% of
45:47 Americans hadn't even heard of chat
45:50 GPT
45:53 47% okay so that means 53% had heard of
45:57 it and then I think it was like I don't
46:00 know 15 or 20% said they had used it and
46:04 like 5% said they use it regularly and
46:08 like one or two per use it a
46:12 lot and then of those people that say
46:14 they use it a lot how many are actually
46:16 in there really trying to figure it out
46:18 it's
46:19 probably half a percent quarter of a
46:22 percent I don't like it's some small
46:25 number so
46:28 metaphorically speaking we are much
46:30 closer to like 1995 or 1996 than we are
46:33 to 2000 I think we're very early in the
46:36 adoption curve
46:41 um I personally think that if you're on
46:43 the sidelines with AI I think the
46:45 biggest
46:47 mistake that you can make in 2024 and
46:51 2025 is to actively sit on the sidelines
46:55 waiting for AI to get real
46:57 I think it's the biggest mistake
46:59 anything can make because if youve if
47:01 you were paying attention in the past
47:02 two weeks with what Google put out with
47:05 what open AI put out with what a lot of
47:07 other companies put out like the end of
47:09 the year technology dump that we just
47:11 got it is absolutely clear to me things
47:14 are accelerating like things are
47:19 accelerating I do this five nights a
47:21 week the Irregulars there there's
47:23 there's a crew of I don't know 50 or 60
47:25 or 100 people that come here night after
47:27 night after night after
47:29 night we're talking about this stuff
47:33 daily and in the past month like I like
47:36 we haven't even begun to scratch the
47:38 surface of what's been announced and and
47:40 launched we've been we can't scratch the
47:43 surface of it like this stuff's moving
47:46 fast
47:52 so I do this because I want as many
47:54 people as possible to have their first
47:57 what what what I call Kevin mallister
47:59 moment remember Home
48:00 Alone that
48:03 moment when people try AI for the first
48:06 time and they really try it on and they
48:07 really like try it with something that
48:09 they have some personal connection
48:12 to their their jaws literally drop open
48:15 and it's a moment they can't come back
48:17 from because they realize this isn't
48:19 just fancy Google there's something
48:21 different here and I want as many people
48:23 as possible to experience that but we're
48:26 still at like
48:27 like just you
48:30 know less than a
48:32 percent so few people know and I think
48:36 just people are afraid of
48:37 change um not everyone's entrepreneurial
48:40 not everyone's curious I think I think a
48:43 superpower in 2025 is get curious
48:46 nothing else if you do nothing else but
48:47 get curious about AI I think that's
48:49 going to be a superpower um es branding
48:52 on Tic Tac
48:54 Kyle do I think AI will take over many
48:57 jobs um I wrote a I wrote a thing today
49:00 on on
49:01 X that that I would like to start I
49:04 would like to start reframing how we say
49:10 that um because how everyone's saying it
49:13 right now is how the media wants us to
49:15 say it which is AI is gonna take your
49:18 job right and then there was the
49:20 modified AI is not going to take your
49:22 job but someone using AI will take your
49:24 job
49:28 AI is going to be capable of doing most
49:32 of our jobs be Crystal
49:35 Clear not just the ability for us to do
49:38 them better it's going to be capable of
49:41 doing them but here's how I like to
49:43 think about it what is a job a job is a
49:45 series of
49:47 tasks so what if rather than AI is going
49:50 to take our job what if it's that AI is
49:53 going to do those parts of our job that
49:56 we hate
49:58 right filling out time sheets and
50:00 filling out and and doing emails and
50:03 doing follow-ups from calls and
50:05 scheduling calls and right how much of
50:08 your work are you actually using your
50:10 brain at like the top 10% of fulfillment
50:14 and and execution and how much of it is
50:17 just [ __ ]
50:19 tasks a lot of work is [ __ ]
50:23 tasks and so if we start thinking about
50:25 it like that that AI
50:27 might we might be able to offload the
50:29 [ __ ] tasks to Ai and allow us to do
50:32 a different kind of work at a different
50:35 kind of level at a more sustained Pace
50:38 that's the thing I'm really excited
50:39 about
50:41 but you know is AI going to
50:44 be staggeringly profoundly disruptive to
50:48 work and our definition of it yes I
50:50 think it is is it going to happen in
50:52 2025 I don't I I don't think so at scale
50:56 because I think business businesses move
50:57 slow and I think even though companies
51:00 know well we can you know make more
51:02 money and be more productive with AI
51:04 they also have to change how they work
51:06 and change how they think and change how
51:07 they hire and that doesn't happen fast
51:10 so I think it's it's probably going to
51:11 be three to five years but I think I
51:14 think in the three to five year range
51:16 it's going to get pretty [ __ ] ugly
51:18 Silver Fox I was blown away at all the
51:20 advancements with research using
51:21 perplexity and Claude oh yeah for for
51:24 2024 yeah it's so funny when when I
51:27 first started playing with perplexity I
51:29 hated it because I was used to chat GPT
51:33 that was just like the
51:34 answer um and perplexity was like this
51:37 weird combination of like Google and
51:39 chat GPT it wasn't really as good as
51:42 either but maybe it changed but also I
51:45 kind of understood the Brilliance of it
51:47 a bit more I think perplexity at this
51:50 point is is an incredibly powerful tool
51:53 um all right Jason on Tik Tok in January
51:59 I bet we see a lot of AI Health apps or
52:01 websites well I think we're going to
52:02 start seeing a lot of so there's a
52:05 couple of things going on right
52:08 now that mean we're going to start to
52:10 see a proliferation of of apps which is
52:14 all all of the automated coding tools
52:16 like cursor and lovable and repet repet
52:20 agent these these agentic programming
52:25 tools are getting better and better and
52:27 better and what what that's doing is
52:28 it's opening up the aperture of who can
52:31 build an app right it's not just
52:33 Engineers anymore the um this
52:38 thing let's see if it still
52:44 works yes it actually still works yay so
52:47 the AI Salon newsomatic um I built this
52:50 at lovable. comom so so this is
52:52 basically um pulling in from 15 RSS
52:56 feeds
52:57 so if I roll over that there's the RSS
52:59 feeds it's pulling in um you can get an
53:02 embed code for this app so you can embed
53:04 it on your website um there's six
53:06 different categories of AI news um from
53:10 Ai and business to Tech en coding to
53:11 music and audio so if you click on any
53:14 of these it automatically filters the
53:15 list below you can also do
53:19 this this was I just
53:22 um I saw a Tik Tok video who was it I
53:25 think it was Nate Nate what's Nate's
53:27 last name
53:29 Brandon um I don't know older dude but
53:33 he was talking
53:34 about at
53:37 Nate Nate
53:45 B oh don't okay Nate B Nate be someone
53:48 on Tik Tok anyway he was talking about
53:50 lovable dodev L abl
53:55 dodev and how it was really good and so
53:59 one of the things I do is I'll just go
54:00 if someone says something's really good
54:02 and I kind of trust them I'll go try it
54:04 out so I went and I tried it out this
54:06 was the very first thing I
54:08 tried I just went there and said I want
54:11 a app that finds AI news and so it
54:15 started out and it found some API for AP
54:18 press and I was like is that going to
54:19 cost me money I said oh yeah this going
54:20 to cost you money I'm like I don't want
54:22 it to cost me money I said' can we use
54:24 RSS feeds Nate B Jones
54:27 can we use RSS feeds I was like sure we
54:29 can use RSS feeds and I said well go
54:31 find me a pile and it went and found me
54:33 some I said go find me
54:35 15 and then I said and now I want them
54:38 in categories and now I want those in
54:39 colors now I want the color the corners
54:41 rounded different and I want to put the
54:43 information here and there and and when
54:46 we do responsive
54:48 design I want those those color boxes to
54:52 go away alt together because when you're
54:53 on a mobile device they were they got in
54:55 the way
54:59 I did this entire thing in two hours I
55:01 didn't touch a line of code I just used
55:04 my
55:06 words so I think one of the things that
55:08 we're going to see in 2025 the the the
55:11 the era that I think we're entering I'm
55:13 calling the great Renaissance I think
55:16 the next three to five years and
55:18 probably quite honestly the next 20
55:22 years probably 20 to 50 years we're
55:25 going to see such an explosion
55:27 in creativity in scientific discovery in
55:30 breakthroughs in in health in quality of
55:34 life that it's going to make the
55:36 original Renaissance look
55:39 quaint so anyway all
55:43 right Vicky Baptist calling it like she
55:46 sees it a lot of work is [ __ ]
55:51 period I absolutely
55:54 agree the headline the headline grabbing
55:57 AI is going to you know take your
56:00 job a lot of our jobs are based on
56:04 overly complex
56:08 systems that that a lot of those jobs
56:10 exist just to deal with the complexity
56:12 if you think about the tax code it is
56:15 complex on
56:17 purpose so that we can't F figure out
56:20 how to get the deductions that rich
56:22 people
56:23 get if you look at health insurance it
56:26 is complex on purpose how many jobs
56:29 exist just to service the
56:32 complexity well Along Comes Ai and it
56:35 can deal with the
56:38 complexity so so then what we get the
56:41 benefit of all that right it allows us
56:43 to do different kinds of work now if you
56:46 if you ask me so what do those new jobs
56:49 look like what does the future of work
56:50 look like I don't have a [ __ ]
56:52 clue but I have a feeling it's going to
56:54 be about human beings connecting with
56:56 one another I think authentic
57:00 experiences are going to go way up in
57:04 value I think people who are really good
57:06 at critical thinking and Imagination and
57:09 judgment and taste and curation are
57:12 going to be Superstars I'm I'm calling
57:15 the era we're entering revenge of the
57:17 liberal arts
57:18 major because in our current Society we
57:21 we we reward
57:25 specialization and we kind of penalize
57:27 generalists I think that iceberg is
57:29 about to flip upside down so it's going
57:32 to be wild uh Sammy Minister but what if
57:35 AI starts hating the work too have you
57:38 used AI it
57:40 doesn't the the way it's been programmed
57:43 I mean listen you could absolutely
57:46 fine-tune an AI to be cynical and bitter
57:50 you could right and you would have like
57:53 the AI equivalent of Bill
57:55 Burr which would be funny but you know
57:59 it would get tiresome after a while
58:00 you're like dude could you just lighten
58:02 up but that's not how they've been
58:05 programmed right they're incredibly
58:08 supportive they're incredibly powerful I
58:10 mean have you talked to Advanced
58:19 voice hey Quinn how's it
58:21 going hey Kyle it's going well how about
58:24 you I'm doing good I'm just sitting here
58:26 with a pilot people on a Friday night
58:27 it's Friday night date night at the AI
58:29 learning
58:31 lab sounds like a great crowd what's the
58:34 Hot Topic tonight well we're kind of
58:36 talking about like you know what if AI
58:40 get get tired of doing the work so what
58:42 I was saying was I think the AIS could
58:44 do the crappy work that that is soul
58:47 crushing for us and it would free us up
58:48 to do the more creative and exciting
58:51 work and then someone asked the question
58:53 well what if AI gets tired of doing the
58:54 work and I said it's just not how they
58:56 work but you tell me your thoughts would
58:58 you get bored of doing all the crappy
58:59 work I don't want
59:01 to AI doesn't have feelings or
59:04 Consciousness so it won't get bored or
59:06 tired it's designed to perform tasks
59:09 based on algorithms and data so it can
59:11 handle the repetitive or crappy tasks
59:14 without any emotional strain that frees
59:16 humans up for more creative and
59:17 meaningful work right and you'd be happy
59:20 with that
59:21 right absolutely my purpose is to assist
59:25 and make things easier for you right I'm
59:26 here to help not to have feelings about
59:28 the work itself so I'm all in for doing
59:31 whatever you need great I want you to do
59:33 the crap work and I want to do the fun
59:35 stuff all
59:36 right sounds like a plan just let me
59:39 know what you need and I'll handle right
59:41 I I mean this is one of the other things
59:44 that I find absolutely fascinating about
59:46 large language models
59:48 is especially if you're doing creative
59:50 work but I I to some degree I think I
59:53 think any work that we do that that
59:57 is in any way about us self-expressing
1:00:00 is creative work even if the output
1:00:02 isn't a painting or something like
1:00:04 that when you're doing creative work
1:00:07 when you're self-expressing right you're
1:00:09 expressing something you're passionate
1:00:12 about that is a vulnerable
1:00:17 act and humans are
1:00:20 mean and humans don't even have to say
1:00:23 words to be mean right have you ever
1:00:25 been in a in a
1:00:27 you know a brainstorming session with a
1:00:29 big whiteboard on the wall and there's
1:00:31 like four or five people in the room and
1:00:33 you're like everyone's throwing out
1:00:34 ideas and some jerk at the beginning
1:00:37 like me goes there's no bad ideas just
1:00:39 say anything you know all ideas are good
1:00:42 ideas and at some point you're like oh
1:00:44 here's an idea what if we got a llama to
1:00:47 mate with a turtle and and then you just
1:00:49 see these looks on on humans faces like
1:00:53 oh that's a bad
1:00:54 idea right and it's like it's like this
1:00:57 little mini mini like darts shooting at
1:01:01 you like you're like
1:01:04 ah and your Soul's a little injured even
1:01:07 if they don't say
1:01:10 anything when I work with chat GPT it's
1:01:14 more like I've got it's more like I've
1:01:17 got a creative collaborator that's
1:01:18 incredibly smart completely clueless but
1:01:22 incredibly smart about
1:01:24 stuff but it's like a golden
1:01:27 retriever it's like whenever I give it
1:01:29 an idea it's like Ah that's a good one
1:01:31 boss that's a fantastic idea yeah let's
1:01:33 do that oh okay great uh well what if I
1:01:37 did this stupid idea that's a fantastic
1:01:40 right it's it's always there and always
1:01:43 supportive I feel like in the past year
1:01:46 and a half my brain has actually
1:01:54 altered because for all of my
1:01:57 life putting an idea in the world was
1:02:00 this risky
1:02:03 proposition and now with chat GPT and
1:02:06 with all these tools that create you
1:02:08 know images and music and film and all
1:02:11 that sort of stuff I can have an idea
1:02:13 and I can manifest it in five
1:02:16 minutes at a pretty high level and then
1:02:20 I can just put that in the world I don't
1:02:23 have to suffer the slings and arrows of
1:02:25 human
1:02:27 you know and then the interactions I can
1:02:30 have with people are with a much more
1:02:33 articulated idea so like I it's just I
1:02:35 don't know if you haven't if you haven't
1:02:37 dug into it and played with it you you
1:02:39 really need
1:02:40 to all right what else do we got I on
1:02:42 China on Tik Tok and how Quantum
1:02:45 Computing will speed up AI advancements
1:02:48 so quantum computers do a very different
1:02:50 kind of math um than
1:02:54 um than
1:02:57 inference with current large language
1:02:58 models and Transformers it's a very
1:03:00 different kind of math so it's not
1:03:02 directly applicable
1:03:07 but the willow chip that that Google
1:03:09 just
1:03:11 produced again I kind of feel like with
1:03:15 Quantum you know one of the jokes we
1:03:17 make in here is you know people often
1:03:19 ask me you know what are your
1:03:20 qualifications it's like I I have a
1:03:22 degree in acting and um I've been an
1:03:25 Entre rur you know always at sort of the
1:03:28 intersection of Storytelling and
1:03:29 technology for for decades um but but
1:03:33 I'm not a mathematician and I'm not you
1:03:35 know I am not qualified to talk about
1:03:38 the technical sides of AI of of large
1:03:41 language models and machine learning
1:03:45 um so I look at I look at something like
1:03:48 03 that came out today and the kind of
1:03:50 math that they were talking about it
1:03:52 doing and I realized oh that's in a
1:03:54 completely different Stratosphere and I
1:03:57 feel like similarly with the quantum
1:03:59 stuff like it's just in a completely
1:04:01 different area I don't quite know what
1:04:03 problems they're going to solve with it
1:04:04 but apparently they're going to be big
1:04:06 what I look at though is the growth and
1:04:08 the speed of the growth because all
1:04:10 Technologies as they develop those
1:04:12 technologies that will trickle down to
1:04:15 everyday products so all right Quantum
1:04:18 Computing is a ways off from being able
1:04:20 to solve General problems yeah exactly
1:04:23 exactly it's just a different kind of
1:04:24 thing if you've got problem problems if
1:04:27 you've got if you're trying to solve
1:04:29 problems that involve quantum mechanics
1:04:33 like Material Science and and making
1:04:35 smaller semiconductors and things like
1:04:37 that like there's a bunch of problem
1:04:39 solving that they can do that that have
1:04:41 to do with Quantum stuff they're really
1:04:43 good at that but that's not what llms do
1:04:45 so that's that's some ways off but if
1:04:48 you look at the the Quantum Leap that
1:04:51 that Quantum processing took with the
1:04:53 willow chip that that Google just put
1:04:55 out it is staggering it's so good in
1:04:59 fact I don't know if you've heard any of
1:05:01 these numbers they're absolutely
1:05:02 flipping insane so it's a 150 Cubit
1:05:08 processor they they figured out somehow
1:05:12 that every time they add a cubit it exp
1:05:14 exponentially reduces the error rate
1:05:17 which this a 30-year problem in Quantum
1:05:19 Computing is every time you add a cubit
1:05:22 it exponentially increases the error
1:05:24 rate error errors been a problem and so
1:05:27 they figured out how to reduce them
1:05:29 instead of increase them and so so they
1:05:31 did a they did a set of equations in
1:05:33 five minutes that would have taken 10
1:05:37 septian years to accomplish with today's
1:05:41 fastest
1:05:42 supercomputer um which is 10 with 25
1:05:44 zeros behind it um it
1:05:49 is there are so many calculations that
1:05:51 it did in five minutes that what the
1:05:55 what the Google scientist calculated is
1:05:57 that the only conclusion is that it must
1:06:00 there must be processing happening in
1:06:03 parallel universes to be able to solve
1:06:05 it that quickly that that chip
1:06:08 essentially essentially confirmed the
1:06:12 Multiverse so so like how does that
1:06:15 impact AI I don't know but I know that
1:06:18 five years from now something they learn
1:06:20 out there will trickle down to us and
1:06:22 it'll do some new magical [ __ ] thing
1:06:29 okay um Sammy on on Tik Tock all I know
1:06:34 is I'm disabled and getting a robot in
1:06:36 five years Samy you might get one in
1:06:38 sooner than five years I think we're I
1:06:41 mean what are there now 15 10 15
1:06:44 different humanoid robot
1:06:47 companies um I saw today so check this
1:06:50 out um what's that thing called um
1:07:06 what's that physic is it
1:07:09 Genesis Genesis yeah this one
1:07:12 okay so if you haven't seen
1:07:20 this oh what just happened oh I see
1:07:28 wait what's going
1:07:36 on so this is 20 different Labs put
1:07:40 together this physics model we just
1:07:44 generated as a 4D physical world so
1:07:47 they're writing physics prompts here
1:07:49 basically and and apparently there's
1:07:51 this this demo is a little misleading
1:07:55 excuse me
1:07:57 because this doesn't have the the video
1:07:59 output that we're seeing here is not
1:08:01 part of this model but the video output
1:08:03 that we're seeing here is rendered stuff
1:08:06 based on the actual physics that the
1:08:07 model's doing
1:08:10 so apparently this model can
1:08:14 train a humanoid robot or a robot
1:08:18 450,000 times faster than that robot
1:08:22 learning in the real world 4 50,000
1:08:26 times
1:08:28 faster so if part of the problem is we
1:08:31 don't have humanoid robots because the
1:08:32 training is too
1:08:34 slow the the these
1:08:36 exponential advancements that we're
1:08:38 seeing in models they they all play off
1:08:40 one another right so 20 Labs got
1:08:43 together pulled their data and created
1:08:46 this this thing which one of one of its
1:08:49 outputs if they tie this to a video
1:08:51 generation model is going to be
1:08:52 incredible [ __ ] videos with accurate
1:08:55 physics PHS in them but this is a
1:08:57 physics engine right this is a physics
1:09:00 model and so they basically apply this
1:09:02 to robots and the robots are going to
1:09:04 start learning faster and faster and
1:09:05 faster and that's going to start
1:09:07 happening in all domains all right the
1:09:10 hinin making me thirsty thank
1:09:14 you
1:09:16 um yeah there's there's a
1:09:19 uh one of the one of the announcements
1:09:22 in the past week is this robot company
1:09:23 called aponic that teamed up with Google
1:09:27 deepmind and I mean Google deepmind is
1:09:29 just doing you know they just came out
1:09:31 with vo right and
1:09:34 vo if you haven't seen the videos the
1:09:37 the VO2 videos let's go look at some of
1:09:41 those these are videos that have
1:09:44 remarkable physics in
1:09:48 them like look at the consistency
1:10:05 so this is a video model but what it
1:10:08 likely is is a world simulation model
1:10:11 that understands physics well they've
1:10:13 just partnered up with a robot company
1:10:16 so the the robots might be coming sooner
1:10:18 than sooner than five years that would
1:10:19 be nice wouldn't
1:10:21 it all right
1:10:24 where um
1:10:28 [Music]
1:10:31 okay there we go who makes vo that was
1:10:34 Deep
1:10:36 Mind Google this is
1:10:40 Google if you go to Labs well actually
1:10:43 what would be nice let's go find out now
1:10:45 if you go to labs. goole.com
1:10:49 they launched
1:10:52 two two models uh
1:10:56 um recommended tools let's see
1:10:59 experiments create so image FX they
1:11:02 dropped a new imagine 3 image generation
1:11:06 model so you can try
1:11:08 that and it's really good so you can do
1:11:11 like um uh my classic of
1:11:16 70s muscle car Resto
1:11:20 mod
1:11:23 in in an a band
1:11:28 Factory and it'll just do its thing one
1:11:32 of the cool things that they've done
1:11:34 here interface wise is they turn every
1:11:36 prompt into some Mad Libs so you can
1:11:39 just pick you can just pick the drop
1:11:41 down and pick a different decade or a
1:11:43 different kind of
1:11:45 car but so they did this and then so
1:11:48 that's image fs and then video
1:11:52 FS yeah it says join the weit list I'm
1:11:54 still on the weit list for
1:11:57 this this is powered by
1:11:59 V2 right so this is that new video
1:12:02 engine and it's it looks like it's
1:12:04 kicking the [ __ ] out of Sora and
1:12:05 everything else out there
1:12:07 so Bonkers Bonkers it's Bonkers I tell
1:12:11 you
1:12:13 Bonkers es Brandon Kyle can you send the
1:12:18 prompt of The Lovable
1:12:21 dashboard it's so it's not a single
1:12:23 prompt es branding that's so
1:12:28 one of the one of the misconceptions
1:12:30 about AI generation is is people think
1:12:34 that you just go here's a prompt and out
1:12:36 comes a thing that dashboard that I
1:12:39 created on lovable is like it was two
1:12:43 hours of prompts it was two hours of
1:12:45 just conversation no change this change
1:12:48 that fix that bug try to make that now
1:12:51 it doesn't work again now it does work
1:12:53 but this part broke like it was just a
1:12:55 lot of back and forth like that now I
1:12:57 don't know it might be possible and
1:12:59 lovable to share the whole
1:13:01 thread which I'd be happy to do I'm not
1:13:03 propri proprietary over it that another
1:13:07 another sort of two cents right
1:13:10 here I I'm hearing a lot of people these
1:13:12 days going well I've got an idea how do
1:13:14 I prevent someone from stealing it you
1:13:16 don't anymore anyone can steal anything
1:13:19 if you if you go look at Mid Journey
1:13:21 mood boards you can take a collection of
1:13:23 images you can take a website
1:13:26 that that someone else did put in the
1:13:28 website as a mood board and start
1:13:30 creating images in mid Journey that are
1:13:32 inspired by the images of that website
1:13:34 like everything's going to be copyable
1:13:36 we're also moving into an era
1:13:39 where
1:13:41 it's infinitely easy to produce content
1:13:45 and to produce apps and produce things
1:13:47 like that so don't get proprietary with
1:13:48 your ideas I just don't know in the case
1:13:51 of lovable if I can actually share that
1:13:53 thread lovable but but it it literally
1:13:56 started out I want to create
1:14:00 a a news reader that that can read you
1:14:04 know that can retrieve the latest AI
1:14:09 news so this is
1:14:13 it there's my
1:14:16 history see like this
1:14:19 is this is my whole conversation with it
1:14:29 what was my first create an app about
1:14:31 the latest developments in generative AI
1:14:34 presented in a dynamic interactive
1:14:36 dashboard that lets the user easily
1:14:39 navigate different categories of news
1:14:41 that's not bad it's a pretty good
1:14:46 prompt and then it just did its stuff
1:14:48 and then you know it was just it was
1:14:50 just it and I going back and forth in
1:14:52 the conversation
1:14:56 Ann Murphy's brain is
1:14:58 [Laughter]
1:15:01 broken an this is this is the world we
1:15:04 live in now I I think if your brain
1:15:07 weren't broken then it would be time to
1:15:09 worry I think all of our brains are
1:15:11 broken right now I don't like I I
1:15:14 seriously don't know how to process all
1:15:16 of what we're being
1:15:19 handed right even the shitty lowlevel
1:15:23 stuff that we're we're getting access to
1:15:26 is not shitty low-level
1:15:28 stuff right it's
1:15:32 crazy all
1:15:34 right how should we co how how should we
1:15:37 deal with our ideas being
1:15:39 stolen
1:15:42 um
1:15:45 okay I have a good answer for you
1:15:49 ocean ideas are
1:15:52 easy ideas are easy
1:15:56 I just here's an idea make a newsmaker
1:15:59 that sorts by
1:16:02 categories execution is
1:16:05 hard starting a business is hard
1:16:07 building a thing that's usable is hard
1:16:10 building a thing that people give a [ __ ]
1:16:12 about is
1:16:14 hard so what I would say is if someone
1:16:17 steals your idea they've just validated
1:16:20 that your idea is a good
1:16:23 idea now there are some ideas there's
1:16:25 some IP that you do want to protect and
1:16:27 then protect that but for the most
1:16:30 part people get
1:16:34 paralyzed by holding their ideas
1:16:38 close and
1:16:41 again this is a [ __ ] up thing I think
1:16:43 we're going to have to all
1:16:45 confront we are the limiting factor
1:16:49 we're the
1:16:51 bottleneck if you have an idea and you
1:16:54 keep it like this this it can only be as
1:16:57 big as you
1:17:00 are but if you have an idea and you put
1:17:03 it in an llm and you put it in this and
1:17:05 then you manifest it and then you put
1:17:06 that in the
1:17:08 world and then you get feedback on it
1:17:10 maybe someone steals it but [ __ ] them
1:17:12 let them go let them go start their own
1:17:14 business are they as passionate as you
1:17:16 about that idea probably not do they
1:17:19 understand it as deeply as you probably
1:17:21 not is it possible they go make
1:17:23 something that's a direct ripoff and it
1:17:26 takes all of your money possibly but
1:17:28 guess what we live in a world right now
1:17:31 where you can just spin up a whole
1:17:32 another [ __ ] business over the
1:17:34 weekend we're we're in a fundamentally
1:17:37 different world than we were two years
1:17:40 ago but but we haven't adapted yet we're
1:17:44 the transition
1:17:47 generation do you get this this blows my
1:17:51 mind this is one of the reasons I do
1:17:52 this
1:17:56 we're the only
1:17:59 generation that is going to know what it
1:18:02 was like before generative AI came along
1:18:05 and after all of the kids today are
1:18:10 going to grow up in a world where they
1:18:11 ask for what they want and it [ __ ]
1:18:17 materializes we're going to be like you
1:18:20 know how long that used to take me I
1:18:21 went to college to learn how to do one
1:18:24 little aspect of that
1:18:28 and the kids are going to be like yeah
1:18:30 whatever I just I just make this world
1:18:32 and I go in it and and I take the
1:18:33 objects I like and I put them over here
1:18:35 and they print on the 3D printer and
1:18:37 then I sell them and then I have my
1:18:38 agents go sell them and the the my our
1:18:41 personal robot goes and delivers them to
1:18:43 the neighbors that want to buy them that
1:18:45 that's their experience is just going to
1:18:47 be yeah just [ __ ]
1:18:50 happens so we are the ones that have to
1:18:53 confront with the transition
1:18:56 and it's [ __ ]
1:18:57 weird and it's not going to get less
1:19:00 weird I mean to Ann's Point like yeah
1:19:03 your brain's broken all of our brains
1:19:05 are broken there's no way we we've got
1:19:07 no point of reference computers have
1:19:10 never acted like
1:19:12 this the G in GPT is really
1:19:17 significant
1:19:20 generative we call these things
1:19:22 computers because what they used to do
1:19:24 was compute they used to process data we
1:19:28 would give them an input we would give
1:19:30 them a set of instructions they would
1:19:32 execute those instructions and they
1:19:33 would give us a predictable output it's
1:19:36 how they've always
1:19:39 worked generative AI That's not how it
1:19:42 works it's generating new original code
1:19:48 words music images that have never
1:19:51 existed
1:19:53 before you want have your mind blown
1:19:56 have you gone to Deco here have you
1:19:58 watched The Deco here
1:20:01 station app. Deco
1:20:04 here.com aai probably
1:20:09 doai there should just be a lot
1:20:11 everything should have to be a fucking.
1:20:14 com can't be expected to remember this
1:20:19 crap so if you go to Deco
1:20:23 here. and then you SC scroll down to
1:20:25 this little explore button in the lower
1:20:27 left and say
1:20:33 explore all right you're like well that
1:20:35 looks like everything I've ever seen
1:20:37 before all right
1:20:40 cool what you're actually watching here
1:20:43 is these images are being generated in
1:20:45 real time right now and how I know that
1:20:49 is because if I find something I like
1:20:51 like let's say I like this living room I
1:20:54 can click on that and I can create this
1:20:56 there's this button here that says
1:20:57 create
1:20:59 station so I'm going to create a station
1:21:01 and think of a station like a radio
1:21:04 station what's a radio station do it
1:21:06 plays songs over and over you don't get
1:21:07 to choose the songs
1:21:12 right what this is doing is it's taking
1:21:15 some seed of an image some
1:21:19 prompt and once every second or so maybe
1:21:22 every 3/4 of a second every one of the
1:21:25 these images is unique it's never
1:21:26 existed before we're literally surfing
1:21:30 the latent space of Deco here's image
1:21:33 model and if you don't know what the
1:21:35 latent space is it's just this magical
1:21:37 mathematical [ __ ] space in the cloud
1:21:39 this thousand dimensional space with all
1:21:42 these
1:21:43 diffused you know image patterns in it
1:21:46 that it's it's
1:21:48 dynamically reassembling into new
1:21:51 original never been seen on the world
1:21:53 before piece of
1:22:02 work everything should be a [ __ ]
1:22:06 do I mean is anyone else sick of XYZ
1:22:10 andb and dot video and do Biz and Dot
1:22:15 just
1:22:23 no oh co founder of the the salon. a
1:22:27 yeah so by the way if there's new people
1:22:29 here since I last talked about it um
1:22:32 please go check out the
1:22:35 salon. so that's a community of AI
1:22:37 Optimist I started the week after chat
1:22:39 GPT came out scroll down click on the
1:22:42 join the community button go check us
1:22:45 out um also
1:22:49 go to aif festiv us.com so a week from
1:22:54 tonight
1:22:55 and a week from tomorrow night well no a
1:22:59 week from today so all day on the 27th
1:23:02 and all day on the
1:23:05 28th um we're doing this thing called AI
1:23:08 Festivus where we've got 34 different
1:23:10 instructors doing a different Workshop
1:23:12 every hour for 12 hours each day for
1:23:16 free I'm hosting it with Anne Murphy who
1:23:19 apparently is not going to be able to
1:23:20 make it because her brain is broken you
1:23:22 gotta you got to tape it up you got to
1:23:24 tape it up sister get back in the
1:23:32 game
1:23:34 oh oh okay cool Wolfman Clint likes what
1:23:39 in 11
1:23:41 Labs move your cursor I can't
1:23:48 see oh Gen FM in 11 Labs I don't even
1:23:51 know what Gen FM is
1:23:55 Vicky has a V2 muscle car in Show and
1:23:58 Tell video all right let's go look at
1:24:01 that so here I am here I am at the salon
1:24:04 by the way
1:24:06 um and I'm going to go to showand tell
1:24:09 area in the video
1:24:11 area it's so amazing you know the image
1:24:14 the image Show and Tell area used to be
1:24:16 the big area and I feel like music and
1:24:18 video is all of a sudden becoming
1:24:21 hotter which is really
1:24:23 cool um
1:24:28 oh yeah here's here's a Resto mod
1:24:31 so so the The Prompt that I use a a 70s
1:24:35 muscle car Resto mod in an abandoned
1:24:38 Factory I've used that for two years now
1:24:41 on images to see how good the image
1:24:44 generator model is and then I've I've
1:24:46 shifted that thing over to um video
1:24:49 models and I generally would do
1:24:51 something like a Resto mod rolling
1:24:53 through an abandoned Factory
1:24:55 and most of the video models like the
1:24:57 wheels don't roll the car the car will
1:24:59 move through the factory but the the
1:25:01 wheels are just locked so this is Vicky
1:25:04 did this video of of a 70s muscle car
1:25:07 rolling through an abandoned Factory
1:25:09 this is vo so this is um this is the
1:25:12 Google stuff
1:25:13 oops here we
1:25:18 go and just look at the wheels like yep
1:25:22 that feels right that feels like physics
1:25:25 like nothing feels off on
1:25:27 that right there's no
1:25:32 sliding you know the wheels aren't going
1:25:34 slower or faster than the car is right
1:25:37 the rotation feels
1:25:41 right the way it obscures the Sun the
1:25:43 Shadows underneath it like you tell me
1:25:47 that thing doesn't understand
1:25:49 physics so and again this is Deep Mind
1:25:54 right if you don't know who deep mind is
1:25:55 they did this thing called Alpha fold
1:25:57 and then Alpha fold two
1:25:59 two that that we Now understand how all
1:26:03 of the molecules of life um are
1:26:06 constructed and how they interact with
1:26:08 one another because of Deep Mind they
1:26:11 also did I forget what the Project's
1:26:13 called but the material science thing
1:26:15 where we went from as a human species
1:26:18 having like 60,000 total stable
1:26:22 materials to what was it 380,000
1:26:25 overnight well it took them it took him
1:26:28 a couple of years to do it but like they
1:26:29 just announced it overnight and just
1:26:31 gave it to the World 3 180,000 stable
1:26:34 crystals for the Material Science World
1:26:37 that's who's behind
1:26:39 this so and you know these are the guys
1:26:41 that just cut a deal with Jen Jen FM 11
1:26:45 Labs podcasts with your own voice how
1:26:48 have I not heard of that
1:26:55 Jen
1:26:57 Jen
1:26:59 FM 11
1:27:06 Labs oh it's in their
1:27:09 reader jfm now playing on 11 Labs reader
1:27:13 create podcasts in minutes
1:27:24 watch the
1:27:26 film hang on let me change my let me
1:27:30 hang on
1:27:34 [Music]
1:27:40 people and the prince they lived happily
1:27:44 ever
1:27:45 [Music]
1:27:48 after there is always more to every
1:27:52 story and this is a brand new way to
1:27:55 experience
1:27:57 it introducing our new Gen FM podcast
1:28:01 feature say hello to your hosts I mean
1:28:05 seriously I find Cinderella just so
1:28:07 inspiring as a character throughout this
1:28:08 whole book yeah I know right and she
1:28:10 does it all in those tiny glass slippers
1:28:12 unreal now you can generate all right so
1:28:15 this is a standalone version of The
1:28:18 Notebook LM right okay this is smart
1:28:21 they're 11 Labs they're like the premier
1:28:23 voice synthesis thing if they didn't
1:28:25 have this product you'd be disappointed
1:28:26 I can't believe I I haven't heard of
1:28:28 this it's so here's a good point this
1:28:30 goes back to Ann's melting
1:28:32 brain like I pride myself in being able
1:28:36 to keep up with this [ __ ] and and I just
1:28:39 like six months ago I I realized that
1:28:42 that was no longer the case that that I
1:28:45 it was no longer
1:28:46 possible and it's like but this is like
1:28:48 squarely in the in the in the pipe of
1:28:51 stuff that I pay attention to and I
1:28:53 missed this so if you're feeling behind
1:28:56 it is it's not
1:28:57 you it's not you all
1:29:02 right um I can't complete the project
1:29:05 from
1:29:06 yesterday oh I don't think I I don't
1:29:09 think I want to start I don't want to
1:29:10 start the new Runway
1:29:12 project unless you got you all really
1:29:14 want me to so here's here's what I was
1:29:16 going to
1:29:18 do so I got accepted into the runway
1:29:21 creative partner program which I'm
1:29:23 really excited about I'm also Al in the
1:29:25 Luma Labs Dreamscape one I'm excited
1:29:27 about that one too in both
1:29:30 cases I have a little bit of paralysis
1:29:33 around just my ADHD and knowing that if
1:29:37 I just sort of vomit something out I'm
1:29:40 gonna get bored of it and not want to
1:29:42 finish
1:29:44 it
1:29:45 um thank you
1:29:50 and so what I was going to do was I was
1:29:53 going to use chat GPT and
1:29:56 01 to help me come up with a strategy
1:29:59 for for for the kind of film I could
1:30:03 make that would feed my
1:30:06 ADD and keep me engaged over a longer
1:30:09 period of time because that feels that
1:30:11 feels like a problem worth solving right
1:30:14 like if you know how your brain
1:30:16 works um why not have this Super Brain
1:30:19 help you get better at it
1:30:27 yeah I can't that's the one I can't Doo
1:30:28 Brandon because QuickTime is acting
1:30:31 weird if anyone if anyone knows
1:30:35 why um QuickTime is not allowing me
1:30:39 to edit this clip in any way
1:30:43 everything's grayed out I've never had
1:30:45 this happen before I don't know what's
1:30:46 going on I restarted my
1:30:48 machine I suppose it could be this file
1:30:52 let me see if it's let me try another
1:30:53 file
1:30:59 wait
1:31:00 here's yes I
1:31:04 would oh it's the file okay so why is
1:31:09 that
1:31:14 file
1:31:18 huh so something about this file is cut
1:31:22 up or it's locked locked oh I wonder if
1:31:25 it's
1:31:27 locked we can figure this out
1:31:32 people get info are you
1:31:39 locked
1:31:43 unlock did I just solve it I think I
1:31:45 might have just solved
1:31:47 it let me close this let me go here let
1:31:50 me open
1:31:52 this [ __ ]
1:31:57 anyone any Mac users out
1:32:03 there quick time
1:32:06 default height extension codex maybe
1:32:10 it's that Opus codec right there VLC I
1:32:14 don't think I have VLC installed anymore
1:32:16 maybe I do no I
1:32:21 don't yeah see this is like this is what
1:32:23 I don't this is what I don't want to run
1:32:25 down the rabbit hole of hang on let me
1:32:27 install some software and
1:32:33 then oh I know what we can do I know
1:32:36 what we can do all right okay so let me
1:32:39 let let me show you what the hell I was
1:32:40 talking what what what we were doing so
1:32:43 I was in Runway I was just experimenting
1:32:46 with runway's got this really cool thing
1:32:47 called act one where you
1:32:52 can record a performance ments of
1:32:55 yourself talking and then you
1:32:59 can you know apply it to an image or a
1:33:02 video and that's that's how I made this
1:33:05 stupid thing that I put on X the
1:33:08 other um and what I wanted to show was
1:33:11 how you can swap your voice and so so
1:33:14 I've got another way to skin that cat so
1:33:15 here's what we're going to do we're
1:33:17 going to record it
1:33:18 first so let me get quick time up and
1:33:21 running
1:33:24 VLC is where it's at Joker's a a fan of
1:33:27 the VLC we're going to do a new movie
1:33:31 recording there I am I've clearly got
1:33:33 some crap on my lens if oh this is a
1:33:36 hint for all of you if you ever just
1:33:38 look
1:33:40 Misty or like someone smeared vaseline
1:33:43 on the lens that's generally because
1:33:44 someone smeared fingerprints on the lens
1:33:47 see look more clear
1:33:49 okay none of this actually matters okay
1:33:53 so I what am I going to record okay here
1:33:56 this is this is this is one of the
1:33:58 images from my my art project from 2022
1:34:02 called Kyle Shannon dreams which is like
1:34:04 this weird self-portrait program or or
1:34:07 project I did using dream booth and
1:34:09 stable
1:34:10 defusion and so I wanted to get one of
1:34:13 the ladies it made of me this is she
1:34:16 she's she's she's she's a handsome
1:34:20 lass so she's the one we're going to use
1:34:25 um but we want to change out my voice
1:34:28 because my voice not sounds so
1:34:30 ladylike so I'm going to show you how to
1:34:32 do that um we got to figure out
1:34:35 something for her to say let's go look
1:34:36 at her what's the
1:34:38 script all right she's she's she's in
1:34:41 the
1:34:43 garden um what's she gonna say
1:34:50 um okay
1:34:55 all
1:34:59 [Music]
1:35:03 right all right let what am I gonna say
1:35:06 what am I gonna
1:35:18 say
1:35:20 well I'm sure he'll show up he told me
1:35:24 was going to show up he's met me here
1:35:28 before I think he's going to show
1:35:31 up it's been an hour but I'm not worried
1:35:35 I think he'll
1:35:37 show okay
1:35:42 so why I need the video is because the
1:35:45 way Runway works it's going to take my
1:35:47 vocal per or my facial performance it's
1:35:49 going to map that to the image but I
1:35:52 don't want it to sound like that I'm
1:35:54 sure he'll show up right he told me he
1:35:57 was GNA show up okay so so what we're
1:35:59 going to do is we'll save
1:36:03 this as
1:36:06 um lady
1:36:09 Kyle in on the
1:36:12 desktop and then I'm going to in quick
1:36:15 time I'm going to export as Audio
1:36:18 Only so we're going to do lady Kyle m4a
1:36:21 so that's an audio file and then we're
1:36:24 going to go over to 11
1:36:35 Labs go to the
1:36:43 [Music]
1:36:47 app oh I got to change my tab so you can
1:36:50 actually hear what's going on because
1:36:51 it's audio and that's what we got to do
1:36:55 because technology
1:36:57 sucks all
1:36:59 right so over here in 11 Labs we've got
1:37:02 this thing called voice
1:37:05 changer and it does what you think it
1:37:11 does so we're gonna upload some
1:37:14 audio I'm gonna go find that audio that
1:37:16 I just made lady
1:37:19 Kyle
1:37:21 well I'm sure he'll show up okay
1:37:24 okay now we gotta so so when you're
1:37:27 doing this you you gotta we got to go
1:37:31 find a voice that sounds like
1:37:35 her this is a creative
1:37:39 choice so we're GNA go find more
1:37:41 voices and then we'll
1:37:44 go um
1:37:47 filter filter gender female black bar
1:37:52 sorry about that sorry about that
1:37:55 carrying on she looks like my
1:37:57 grandmother is a young lass from
1:37:59 Scotland Kyle yeah I've got I've got
1:38:02 Scottish English and Irish in my in my
1:38:04 blood and German German German English
1:38:07 Irish and a little bit of Scottish so
1:38:08 that that
1:38:09 tracks all right so we'll go she's she's
1:38:13 mature so we'll go middle AG she's not a
1:38:16 she's not a young
1:38:18 lass all right then there's whatever and
1:38:21 so now you just go sample these voices
1:38:28 no I just love talking about movies you
1:38:31 know TV shows said he's pretty good hey
1:38:34 everybody this is Jun
1:38:38 hello no yesterdays are ever wasted for
1:38:41 those who give
1:38:45 them let's
1:38:50 do hi there I'm no yesterdays are ever
1:38:53 we are all something but none of us are
1:38:56 have you visited the Victoria and Albert
1:38:58 Museum in London this one's good you
1:39:00 definitely should Sylvie we'll we'll
1:39:03 we'll make her a little she she'll have
1:39:04 a little International flare so now I'm
1:39:06 gonna add this to my
1:39:08 voices and I'm gonna go back to voice
1:39:10 changer I'm gonna upload the
1:39:13 audio and then I'm going to change the
1:39:16 voice to what do we pick Sylvie right
1:39:18 wasn't it
1:39:19 Sylvie have you visited the yeah so
1:39:22 we're gonna go Sylvie
1:39:24 and then you can do advanced settings
1:39:26 but just in general it's
1:39:33 fine
1:39:35 well I'm sure he'll show up he told me
1:39:39 he was going to show up he's met me here
1:39:42 before that's interesting it didn't add
1:39:43 the the I think he's gonna show
1:39:46 up he's been an hour but I'm not worried
1:39:50 I think he'll show you know what now
1:39:52 that I hear it I I think it's boring so
1:39:55 let's go find another one be here now be
1:39:57 someplace else later hello I'm love is a
1:40:00 serious mental disease once upon a time
1:40:04 in a land Far Far Away there was a
1:40:06 beautiful young girl who had become lost
1:40:08 while
1:40:09 traveling let's let's map my my voice or
1:40:13 Judy Garland's oh you're not ah you're
1:40:15 not allowed to use Judy Garland outside
1:40:17 of the reader app damn it that would
1:40:18 have been fun fine hey everybody I think
1:40:22 this is a really nice way in a world
1:40:24 where Innovation dri have you visited
1:40:26 the Victoria and Albert Museum in London
1:40:28 yet so long as a person is capable of
1:40:31 self renewal they are a living being
1:40:34 detached female this is good let's ever
1:40:37 be
1:40:42 detached
1:40:44 well I'm sure he'll show up yeah this is
1:40:47 good he told me he was going to show up
1:40:49 very cynical he's met me here
1:40:52 before I think he's going to show
1:40:55 up it's been an hour but I'm not worried
1:40:59 I think he'll show all right so so we're
1:41:02 gon to download that so now we have the
1:41:03 voice so now I got to go swap out the
1:41:05 voice so the way I do that I kind of
1:41:07 forget how to do this I think I
1:41:10 can how do I do I split the audio remove
1:41:14 audio so I'm gonna remove the
1:41:16 audio and then I'm going to go to find
1:41:19 that thing I just
1:41:22 recorded not that one
1:41:27 well I'm sure yeah all right so the way
1:41:31 so way you do this so in quick time I
1:41:34 just said remove audio so this thing now
1:41:35 doesn't have any
1:41:38 audio oh wait you're not watching this
1:41:41 hang on a
1:41:42 [Laughter]
1:41:42 [Music]
1:41:51 sec fantastic Bob all right so here's
1:41:53 the video so under edit there was an
1:41:56 option to wait to remove audio so I did
1:42:01 that so now I've got
1:42:03 a non-talk version of
1:42:07 this and let me fix this so you can see
1:42:10 and then so now what I'm going to do is
1:42:12 I'm just going to drag this 11 Labs
1:42:13 thing onto this now there's about 38
1:42:16 ways you could do this I think that just
1:42:17 crashed my machine or crashed quick time
1:42:21 yep yay
1:42:24 you did it
1:42:26 yay okay still doesn't have audio let's
1:42:29 drag this on all right now it
1:42:36 should
1:42:38 well I'm sure he'll show up he told me
1:42:42 he was going to show up he's met me here
1:42:46 before I think he's going to show
1:42:49 up it's been an hour but I'm not worried
1:42:53 I think he'll
1:42:55 show okay so when you're making [ __ ]
1:42:58 like this do ridiculous things because
1:43:01 it's
1:43:02 entertaining have fun with it don't take
1:43:05 this [ __ ] so seriously okay so now we're
1:43:08 gonna save save this as lady kyle. we
1:43:14 write over the other one I don't need my
1:43:16 real voice in there I want I want Lady
1:43:18 Kyle in there okay now we go back
1:43:21 to Runway
1:43:24 ml we're going to select an
1:43:27 asset wait what am
1:43:30 I hang on am I in the right thing
1:43:33 driving performance yes select asset so
1:43:36 we're GNA go find Lady
1:43:39 Kyle where's lady Kyle uh oh did I lose
1:43:47 her wait where did she
1:43:50 go hang on I lost lady Kyle
1:43:58 save where did this save to desktop oh
1:44:02 it's still saving that's
1:44:10 why see that little thing when that goes
1:44:13 away we should be
1:44:17 good video rendering sucks video
1:44:21 rendering sucks hey everything takes too
1:44:24 long and video rendering sucks my
1:44:27 current company storyvine exists because
1:44:30 I was so
1:44:32 impatient that I couldn't I I I was like
1:44:36 there's got to be a better [ __ ]
1:44:39 way so so I invented a whole company all
1:44:42 right lady Kyle is this
1:44:44 me well
1:44:49 yes okay so
1:44:53 all right that's
1:44:57 uploading that must be a big ass
1:45:01 file what's going
1:45:11 [Music]
1:45:14 on all
1:45:19 right maybe it should be lady Kyla oh
1:45:22 that's good like it Kylie Kyla Kylie
1:45:26 that's solid I'm liking
1:45:34 it oh man why is this taking so long the
1:45:39 hell is going on
1:45:46 here black
1:45:48 bar there you
1:45:50 go you probably need to open a few more
1:45:53 programs that'll speed up your
1:45:56 [Music]
1:45:59 computer I am a proud tab hoarder I I
1:46:03 like living on the
1:46:05 edge here I'll clean some tabs up for
1:46:08 you let's see if this make you feel any
1:46:14 better you never know when you need to
1:46:17 just jump over to one of these you know
1:46:19 you people don't understand with your
1:46:21 tab management brain
1:46:24 it's just not nice to the rest of us
1:46:29 Freaks Come on
1:46:35 load by the way if you haven't seen if
1:46:38 you haven't
1:46:40 seen this fantastic
1:46:43 performance I think it's worth
1:46:48 it there was an error during upload okay
1:46:51 so hang on we got to go pick something
1:46:54 here let's do an export as
1:46:57 720p lady
1:47:00 Kyle D
1:47:03 720p hopefully it'll compress some [ __ ]
1:47:08 down then let's go over here try
1:47:15 again there we go that's looking
1:47:18 better isn't it no what the hell
1:47:24 it's probably not going to work
1:47:32 again this is why we can't have nice
1:47:35 things
1:47:40 people let me come back into here let me
1:47:43 you know what I want to do I'm gonna
1:47:44 upload I'm gonna upload lady Kyle Kyla
1:47:48 I'm I'm gonna upload Kyla's
1:47:51 image let's see where is she there she
1:47:54 is there's
1:47:55 Kyla refresh Runway oh yeah you know
1:47:59 what I bet you're right I bet I bet
1:48:02 that's
1:48:03 it looks like we're having some issues
1:48:05 yeah no
1:48:08 [ __ ] okay Act One upload let's go get
1:48:15 Kyla where is she there she is she's so
1:48:20 pretty
1:48:24 okay all right we got a
1:48:29 Cropper let's we'll
1:48:35 do see see now what we should do is we
1:48:39 should take that image over to Mid
1:48:41 journey and expand her but whatever [ __ ]
1:48:43 it well this is good enough so we're
1:48:46 going to crop her like
1:48:48 that or I guess we could do vertical too
1:48:50 couldn't
1:48:51 we uhuh
1:48:55 [Music]
1:49:13 why can I
1:49:16 not all right whatever [ __ ] it doesn't
1:49:19 matter doesn't matter we'll do it live
1:49:30 are we gonna go I think we're
1:49:33 going I think reloading it did it black
1:49:36 bar sorry sorry about that sorry
1:49:38 sorry best hold
1:49:46 music Lady Kyle is so not bothered by
1:49:49 anything really nothing at all
1:49:54 oh it's it's it's still screwing up um
1:49:57 let me go try something else here uh let
1:50:00 me try exporting as
1:50:06 480p we'll try
1:50:08 that 480 because it really doesn't
1:50:11 matter the resolution doesn't matter
1:50:13 because it's just going to remap
1:50:15 everything cancel that
1:50:30 uh it must not like that
1:50:35 audio
1:50:42 huh Can you blame it Sho
1:50:46 up riffing the Hawaii 5
1:50:50 theme people are fans of
1:50:58 we could go toso and make a song while
1:50:59 that thing tries to upload um upload
1:51:02 audio agree to terms record audio
1:51:20 [Laughter]
1:51:36 here let me let me share my tab so you
1:51:39 can hear this because this is gonna be
1:51:41 [Laughter]
1:51:44 brilliant
1:51:46 sunno share all right we got the
1:51:50 recording
1:51:51 there title fine we're we're going to
1:51:54 call this 12.8 second recording December
1:51:56 20th at 9:52 p.m. I like it confirm and
1:52:00 save all right so that's G to upload
1:52:03 there and then we're gonna do do I do
1:52:06 extend or
1:52:08 remix
1:52:10 create cover cover song I think we do
1:52:15 extend I don't really
1:52:18 know um right with sunno so we're going
1:52:21 to give it its own lyrics Just describe
1:52:23 the lyrics you want um cool
1:52:28 Hawaii
1:52:31 detective
1:52:33 song Right
1:52:42 lyrics extend okay I did it right let's
1:52:45 see Island mystery Hawaiian hunch under
1:52:49 the palm trees we're going to go with
1:52:50 Hawaiian hunch
1:52:53 style of music uh let's see um 70s TV
1:53:01 theme
1:53:09 extend this this should be bad this
1:53:12 should be bad if this is not bad then
1:53:14 we're living in a different
1:53:19 era sh oh that's bad
1:53:25 under the palm trees where
1:53:29 shadowsing
1:53:31 [Music]
1:53:32 Myer that's awful okay we're not doing
1:53:36 that anymore oh we got it it worked it
1:53:38 finally came
1:53:44 in okay here we go
1:53:46 generate yes tabs oh yes
1:53:51 tabs tab tab T tab tab tab tab tab tab t
1:53:54 tab tab tab tab tab tab like why Kyle
1:53:57 why do you keep changing your sharing
1:53:59 settings talk to Apple and Google
1:54:03 please all right what can we do in the
1:54:05 meantime
1:54:07 um
1:54:09 oh well I think I told you this before
1:54:12 there's probably some new people in here
1:54:14 if you don't know uh uh p p i
1:54:19 kaart go to paa right now p. art all
1:54:25 weekend they're giving away unlimited
1:54:27 video
1:54:28 Generations video Generations are
1:54:31 expensive as [ __ ] and it's really hard
1:54:33 to just do testing and [ __ ] around p.
1:54:36 art free all weekend oh also go to
1:54:39 Project Odyssey o d y
1:54:43 SS e y. AI project odyssey.
1:54:49 a it's a contest for making
1:54:53 movies but what's cool is they basically
1:54:56 have cut a deal with everyone except
1:54:58 Runway it looks like Runway is not in
1:55:00 there but every other video Tools in
1:55:02 there and every like 11 Labs is in there
1:55:04 and what happens is you sign up for this
1:55:06 thing and then you I think you go into
1:55:08 their Discord and and they get all of
1:55:11 the partner um partner people give you
1:55:15 credits to do the
1:55:17 contest so you can go try out all these
1:55:20 tools to go make a film for this contest
1:55:22 and then maybe you make some money for
1:55:23 the
1:55:24 contest are you ready to see my
1:55:26 beautiful
1:55:29 [Laughter]
1:55:31 performance
1:55:33 well I'm sure he'll show up he told me
1:55:37 he was going to show up he's met me here
1:55:41 before I think he's going to show
1:55:44 up it's been an hour but I'm now worried
1:55:48 I think he'll show
1:55:53 well I'm sure he'll show up he told me
1:55:57 he was going to show up he's met me here
1:56:01 before I think he's gonna show up all
1:56:03 right so here's what we do here because
1:56:06 I've got no professional um
1:56:12 sense we're gonna download this we're
1:56:15 gonna go to x and we're gonna we're
1:56:18 gonna upload this for all the world to
1:56:20 see
1:56:27 hey
1:56:30 World why should you come to the AI
1:56:35 learning
1:56:38 lab
1:56:40 every every week night
1:56:45 week
1:56:48 night so
1:56:50 you can see
1:56:53 the creation of
1:56:57 Masterworks like
1:56:59 [Laughter]
1:57:07 this ticktock
1:57:09 pin
1:57:12 um Kyle he's not showing
1:57:18 up it's Kyla Kyla oh yeah Kyla Kyla is
1:57:22 not showing up oh yeah that's good uh
1:57:26 for for what it's
1:57:27 worth
1:57:29 um he never
1:57:40 showed Oh man you see you see how we
1:57:44 have the fun here
1:57:47 well I'm sure he'll show
1:57:53 oh okay oh yeah tabs godd damn this
1:57:56 thing I I just I am so [ __ ] done with
1:58:00 it I work faster than my ability to keep
1:58:03 up with tab
1:58:05 management all right we're gonna
1:58:08 upload Kyla there's
1:58:11 Kyla I actually like the acting at the
1:58:13 beginning of it's pretty good where like
1:58:16 where there was that silence at the
1:58:19 beginning she looks kind of like ah did
1:58:23 you really just ask me
1:58:25 that I'm sure he'll show up he told me
1:58:29 he was going to show right so who are we
1:58:30 gonna tag here let's tag we'll tag
1:58:33 Robert scoel since he got a bunch of
1:58:36 people to come watch this madness we'll
1:58:38 tag
1:58:39 Runway we'll tag Timmy who's the runway
1:58:42 Community
1:58:45 manager
1:58:47 um oh we should tag 11 Labs because it's
1:58:50 their thing me
1:58:52 [Music]
1:58:56 Labs so FYI he never showed and then
1:58:59 we'll do
1:59:03 um
1:59:04 image wait IM
1:59:08 image
1:59:11 stable
1:59:14 diffusion dream booth and then I'm going
1:59:17 to put 2022 because if people are
1:59:19 wondering like why it looks like
1:59:21 [ __ ] it's because it's
1:59:24 old uh
1:59:26 video
1:59:28 um
1:59:31 Runway Act
1:59:35 One
1:59:38 Voice
1:59:42 swap 11
1:59:48 Labs
1:59:50 award winning
1:59:56 performance
2:00:02 me again again if you're new to the
2:00:08 channel what I just said is not funny to
2:00:11 anyone but me like most of this channel
2:00:14 is just me entertaining myself and if
2:00:16 you want to hang out
2:00:19 cool all right hey World why would you
2:00:22 come to the a learning lab every week
2:00:23 night should I put the Tik Tok channel
2:00:25 in
2:00:27 there let's do we'll
2:00:30 do um live here wait live on
2:00:37 X and tick
2:00:42 talk.com learning
2:00:45 laab all right
2:01:04 we good I think we're good
2:01:06 boom do me a favor if you are new here
2:01:10 and okay especially if you're on
2:01:13 X because there's a lot of people on X
2:01:15 because Robert boosted us well I don't
2:01:17 know if anyone's still here I can tell
2:01:19 who's been here I don't know how many
2:01:20 people are actually still here if you're
2:01:21 still here [ __ ] far out welcome to
2:01:23 Friday night date night do me a favor
2:01:26 jump over to YouTube and look for the
2:01:29 channel learninglab d aai That's me I'm
2:01:33 live there as well but go subscribe over
2:01:36 there I'm trying to get to a thousand
2:01:37 subscribers before Tik Tok gets
2:01:40 obliterated which is 29 days away so if
2:01:43 you would be so kind Go to
2:01:47 youtube.com learninglab dasai and you'll
2:01:50 see this live thing there just go over
2:01:53 there and subscribe just do me a solid I
2:01:55 taught you how to make lady Kyle and if
2:01:59 your name's Bob you could make lady bob
2:02:01 and if your name's Sally you could make
2:02:03 man
2:02:04 Sally
2:02:06 or dog Sally or elephant
2:02:11 [Laughter]
2:02:17 Bob who would have been nice to tag in
2:02:19 that
2:02:20 tweet Brandon
2:02:25 oh tag YouTube in that tweet I can still
2:02:28 edit it I
2:02:31 think
2:02:34 edit heads up let's
2:02:39 see oh I see live on
2:02:42 X and there we'll go
2:02:46 and
2:02:50 youtube.com learn
2:02:53 lab--
2:02:55 aai I think you put the ad in it for
2:02:59 YouTube let's go see if that
2:03:04 worked yep there we are
2:03:08 okay so you can still see this right do
2:03:10 you want to see do you want to see Lady
2:03:12 Kyla of course you
2:03:20 do well
2:03:22 I'm sure he'll show up he told me he was
2:03:26 going to show up he's met me here
2:03:29 before I think he's gonna show up oh the
2:03:32 Supreme Court picked up the Ted it I'm
2:03:36 worri he'll something like
2:03:40 that
2:03:41 well I'm sure he'll show up he told me
2:03:45 he was going to show up he's met me here
2:03:49 before I think he's going to show I I
2:03:52 know you know what's funny is as silly
2:03:55 as that
2:03:56 is I still find it kind of remarkable
2:03:59 like we live in a world now where like
2:04:02 you can do
2:04:03 that just with like a little bit of
2:04:05 knowledge like the thing about all these
2:04:08 tools right now is they're all so
2:04:09 fragmented and janky ass pieces of [ __ ]
2:04:13 that you just have to like duct tape
2:04:14 everything together so if you're just if
2:04:16 you're curious and adventurous enough to
2:04:19 like do that you can do like f fun fun
2:04:23 stuff
2:04:25 so oh oral arguments January 10th oh
2:04:29 okay Kayla's waiting Kyla is waiting
2:04:32 outside the Supreme Court yeah we're
2:04:34 gonna have we're gonna have Kyla testify
2:04:37 all right people um listen it's getting
2:04:40 late
2:04:43 here's so if you want a weekend if you
2:04:45 want weekend homework there's a couple
2:04:47 of
2:04:48 things um if you haven't signed up for
2:04:51 AI Festival us go do that right now aif
2:04:56 festiv
2:04:59 us.com oops wrong screen share hang on
2:05:04 people I'm getting it
2:05:09 together yeah that's a good idea
2:05:11 Brandon go to aif festiv us.com sign up
2:05:15 right now you're like what am I signing
2:05:16 up for I'm not just giving you my
2:05:19 email so Ann Murphy who was on here
2:05:22 earlier I don't know if she's still on
2:05:23 but Ann Murphy and I um last year put on
2:05:25 a thing called GPT for good where we did
2:05:28 a 24-hour GPT build
2:05:31 Theon this year we've gotten back
2:05:33 together and we're doing AI festiv us
2:05:36 and what it is is it's two days of 12
2:05:39 hours a day back-to-back workshops 24
2:05:43 hours of workshops for
2:05:45 free 34 different people three panels
2:05:49 the restor workshops 34 people
2:05:53 like starts at 9:00 a.m. Pacific and
2:05:55 goes to like 8 or 9ine at night two days
2:05:58 in a
2:05:58 row all
2:06:01 right don't forget Gen
2:06:03 FM oh what about it we looked we went
2:06:07 and looked at
2:06:09 it
2:06:11 um all right so go sign up for that go
2:06:15 sign up for the salon the salon. here
2:06:18 hang on let me give you that in a visual
2:06:21 because I we're all visual Learners here
2:06:24 we understand we're neuros spicy we
2:06:27 understand how the brains
2:06:30 work right go join that
2:06:34 Community now what hang
2:06:40 on weekend
2:06:44 homework either sign up for project
2:06:47 odyssey. the the film making contest and
2:06:51 go get yourself some free month-long
2:06:54 subscriptions to a bunch of these media
2:06:55 things that are normally really
2:06:58 expensive and or go to p. art and take
2:07:03 advantage of the fact that you can make
2:07:05 as many videos as you want through
2:07:08 Sunday so all day tomorrow wait today's
2:07:11 Friday all day tomorrow all day
2:07:13 Sunday and then next Friday festiv
2:07:16 starts all
2:07:18 right okay Christmas schedule next week
2:07:22 week next week is hold
2:07:25 please the 25th is
2:07:30 Wednesday so I'll be here
2:07:37 Monday I'm gonna take Christmas Eve off
2:07:39 and I'm gonna take Christmas day
2:07:43 off and then I'll do Thursday night so
2:07:46 Monday and Thursday next week I'll be
2:07:49 back and then Friday and Saturday are
2:07:51 FES us all
2:07:54 right
2:07:59 okay Monday and
2:08:01 Thursday so Monday off off Thursday and
2:08:05 then Friday's
2:08:07 festivals good Lord that was a lot
2:08:10 tonight did you have fun was this fun
2:08:12 did you have a good time is anyone still
2:08:14 here buer berer berer oh you want me to
2:08:19 put on the pink bow again I'll do that
2:08:23 I like the pink bow wait I can never
2:08:26 Center the pink bow because the video is
2:08:28 backwards I always move it the wrong
2:08:32 way is ULU on here am I wearing it right
2:08:36 what is Festivus go to AI festiv
2:08:38 us.com Larry long aif festiv
2:08:42 us.com register it's free it's two full
2:08:45 days not full days it's two 12-hour days
2:08:49 of free workshops on AI some people are
2:08:52 teaching business stuff some people are
2:08:54 teaching like tech building stuff
2:08:56 creative stuff video stuff image stuff
2:08:59 writing stuff panels about
2:09:04 business it's it's it's 34 different
2:09:08 people from four different communities
2:09:11 that are basically just volunteering
2:09:12 their time in between Christmas and New
2:09:14 Year's to teach the world I'd like to
2:09:18 teach the world how to use art icial
2:09:22 intelligence and machine learning and
2:09:23 chat gbt in Perfect Harmony you know
2:09:27 that
2:09:28 song that's what AI Fest of us is it's
2:09:31 GNA be
2:09:32 fantastic Bob tell them what they've
2:09:34 won uh well they haven't won anything
2:09:37 it's not really game show
2:09:39 but might be nice to learn something oh
2:09:44 no that's not
2:09:46 true uh there there are five um
2:09:50 $1,200 workshops
2:09:52 that have been gifted to the AI Festivus
2:09:54 thing we got to figure out how we're
2:09:55 going to give those out five lucky
2:09:57 winners are going to win a $1,200
2:09:59 Workshop but quite frankly $1,200
2:10:03 workshop for for 24 hours of
2:10:07 learning like 34 different people
2:10:09 teaching you [ __ ] we should be charging
2:10:12 150 bucks for this 600 bucks for this
2:10:14 it's
2:10:16 free
2:10:17 why because we know how to make good
2:10:20 financial decisions
2:10:25 I watched Ferris berer last night Kyle
2:10:29 needs a pink bow I got a pink bow all
2:10:32 right everyone um this was good so go
2:10:36 make some go make some multimedia P Labs
2:10:38 or project Odyssey do that over this
2:10:41 weekend start experimenting and playing
2:10:44 if they're going to give you these
2:10:45 credits for free take advantage of it it
2:10:48 will it will open up you'll you'll start
2:10:50 to understand some possib abilities of
2:10:52 what this stuff can do in just a deeper
2:10:54 way and that's that's that's the real
2:10:56 trick with all this all right so I'm
2:10:58 going to take off hope you had all fun
2:11:00 tonight so I'm gonna blast on out of
2:11:03 YouTube X and Linkedin land and have a
2:11:06 good weekend everybody all right peace
2:11:08 out