Aug 29, 2024 - Navigating the New Era of AI Tools and Techniques

Description

Chapters

Transcript

0:01 [Music]
0:18 girl you're
0:21 looking find a
0:25 Night de fell got you in his s
0:33 what you doing with a clown like
0:39 me surely one of life's
0:42 little
0:45 Mysteries so
0:47 tonight I'll
0:50 ask stars above
0:53 [Music]
1:03 what did I
1:06 do what did I
1:10 say to turn your
1:13 Ranger as as my
1:19 [Music]
1:21 way I'm the guy that never learn to
1:28 dance never even got one second chance
1:34 like last night he doesn't sing at all
1:36 now he can't shut him
1:38 [Music]
1:57 up standing between
2:01 you had a hard plac
2:04 [Music]
2:06 isane standing to
2:10 near you and a fire makes it
2:16 clear your trouble to
2:22 me real trouble can't you
2:28 say leave me close
2:30 [Music]
2:32 smell of your perfume scares me
2:37 most leaning away
2:40 hey it feels stronger every
2:47 day your trouble to
2:50 [Music]
2:52 me R rebble can't you say
2:59 [Music]
3:22 every time I see you now get that look
3:25 in my
3:27 eyes every time I see your mouth I hear
3:32 that
3:34 smile in the early Misty morning L I
3:39 heard the
3:41 engine Old Fall
3:46 [Music]
3:49 outside Mel if you want to come up you
3:51 can come up is there something to talk
3:53 about is it my hair
3:56 [Music]
4:16 see it now get that look in
4:20 M every time I see your
4:24 mouth hear that
4:27 smile in the early Misty morning Lou I
4:31 heard the engine
4:34 thing the old for
4:39 [Music]
4:45 outside again
4:48 today you convince
4:51 me again
4:55 today you leaving this hotel looking for
4:59 so else is Golden
5:04 Ring should say so long
5:10 [Music]
5:13 suzan now don't you
5:16 [Music]
5:19 cry so long
5:25 suan don't you dve for me
5:30 [Music]
5:37 chenging cigarettes and keeping warm out
5:40 on the
5:42 road chasing down a lifestyle out on
5:46 Highway
5:48 24 New York state was a rolling Breeze
5:51 in the sunshine with a blue sky
5:55 falling till of all September creeping
5:59 in
6:03 you believe in
6:06 me again
6:08 today you will convince
6:11 me again
6:14 today you're leaving this hard Town
6:18 looking for someone else's cold and
6:24 Rain should I say so long suzan
6:31 [Music]
6:33 H now don't you
6:36 [Music]
6:40 cry so long
6:43 [Music]
6:47 suzan don't you cry for me
6:53 [Music]
7:02 all
7:03 right good evening good people of The
7:05 Tick Tock what is happening what's
7:09 happening what oh I got rid of my little
7:12 window dang it there it is I got it
7:15 back I know how to do two things on Tik
7:23 Tock and not one of them is entertaining
7:25 anyone
7:30 get
7:32 it all right
7:34 champy we got to turn on the air
7:36 conditioner it's
7:38 hot it's hot that's a hot dog excuse me
7:43 that's a hot dog that dog is hot it's a
7:46 hot dog I like hot dogs hot dogs is
7:48 tasty I know it's there's some
7:50 percentage of pig lips and mouse hairs
7:53 that are allowable within a hot dog I'll
7:55 tell you what I can't taste the
7:57 difference if if I had 3% pig lips or 2%
8:01 pig lips could you tell I don't think so
8:04 I don't think so and I don't mind a
8:06 little rodent hair you know a little
8:08 rodent hair and your hot dog's never
8:09 going to hurt you just pass right
8:11 through you know you know how it is
8:14 [Music]
8:31 oh let's
8:33 see good evening good people if you'd be
8:36 so kind rip
8:38 [Music]
8:44 Frank speaking of someone who just
8:47 caught a rodent that hits a little close
8:49 to home sorry about that
8:52 Vicky but I do have a lovely dragon
8:58 egg h
9:03 bet you all you wish you had one of
9:04 these didn't you reach out to Vicki
9:06 she's got all sorts of stuff she got
9:09 fidget Spinners like or fidget
9:14 toys you got to figure it out you just
9:17 got to it's it'll eventually you'll be
9:19 able to get it
9:24 done that one stresses me out that one
9:28 grinds makes me grind my teeth when I
9:30 when I was in New York City in the 80s I
9:33 bartended for a summer and did a little
9:35 bit of
9:37 cocaine that fidget spinner makes makes
9:39 me feel like when I did cocaine and I be
9:42 grinding my fingernails into my Palms
9:44 grinding my
9:46 teeth oh I probably can't say that now
9:49 that I said that this is going to get
9:51 age restricted oh well Tick Tock this is
9:54 an
9:55 adult session this evening
9:59 [Music]
10:07 the boys were singing shingling the
10:11 summer that we
10:15 met you were 10 and
10:18 17 how could I
10:23 forget all the stars from near and far
10:26 were watching from above
10:31 watching two
10:33 teenagers fall in
10:35 [Music]
10:37 love the way we danced was not a dance
10:41 but more a long
10:44 Embrace we held on to each other and
10:48 floated there in
10:51 space I was shy to kiss you while the
10:55 whole world world could see so Shing
11:00 said everything for
11:04 me and all the poor old old folks they
11:09 thought we lost our
11:11 minds they could not make heads or tals
11:15 of the Young Folks funny
11:18 [Music]
11:19 rhymes I bet that they still knew the
11:22 words and I'll bet they sang
11:25 along to old shamine
11:29 [Music]
11:30 Ding
11:32 chamling Dang d
11:39 [Music]
11:46 [Music]
11:53 [Music]
12:10 [Music]
12:12 it's not simple to
12:14 say that most days I don't recognize me
12:20 that these shoes and this apron that
12:24 place and its
12:26 patrons have taken more than I gave them
12:32 it's not easy to know I'm not
12:36 anything like I used to be although it's
12:39 true I was never attention sweet
12:45 Center
12:47 uh I still remember that
12:50 girl she's
12:53 imperfect but she
12:55 tries she is good but she lies
13:00 favorite song for GG Zen
13:04 mode sorry I'm butchering it she's hot
13:07 on
13:08 herself she is broken but won't ask for
13:13 hell she is messy but she's
13:17 kind she is
13:20 lonely most of the time she is all of
13:23 this mixed up and B some beautiful
13:28 pie she's called
13:30 but she used to be
13:35 mine and it's not what I ask oh
13:39 sometimes life just slips in through the
13:43 back door and carves out a person makes
13:46 you believe it's all
13:48 true and now I've got
13:51 you you're not what I asked for if I'm
13:56 honest I know I would give it all back
14:00 for a chance to start over rewrite an
14:04 ending or
14:05 two for that girl that I know be special
14:11 once again nothing once again I don't
14:16 know I haven't sung that enough recently
14:19 I I ain't know the words the words don't
14:22 escape me the chords weren't too good
14:24 neither
14:26 [Music]
14:31 I I never did get that school learning
14:33 on guitar I didn't go to
14:36 [Music]
14:40 Berkeley it is true I did
14:44 not I have a modum of
14:47 [Laughter]
14:49 talent and and and less
14:54 discipline oh good lord all right so
14:58 we're going to talk about I what are we
14:59 going to talk about tonight lovely image
15:02 here this is from Corey this is uh in
15:05 honor of it's really hot so this is not
15:08 going to be this not going to go well in
15:11 honor of my my pink bow which apparently
15:15 I don't know how to wear I've been told
15:18 by the the lady folk here in the AI
15:21 learning lab that this is not how you
15:22 wear a headband this is how I grew up
15:24 wearing headbands though this is the
15:27 first time in my life I've won one with
15:29 a fuzzy pink bow on it
15:33 [Music]
16:12 really breaks my
16:14 heart see a dear old
16:17 friend go down to that one out place
16:20 [Music]
16:26 again of broken door
16:30 have you heard that sound somewhere
16:35 before do you wonder
16:47 [Music]
16:49 what wait something about publishing a
16:52 song what was that Call Me
16:56 Al I wish you would publish your song
16:59 list I really like these songs most are
17:02 sad but really nice that is
17:04 my my song Library everyone's like oh
17:08 you should play guitar at my party I'm
17:10 like maybe at the end of the party where
17:12 you want people to leave cuz my songs
17:14 are all sad and depressing but [ __ ]
17:16 beautiful um I could I could put
17:19 together uh a little uh song list little
17:23 playlist it's
17:25 mostly they're kind of all over the
17:27 place Martin ston John
17:30 hayatt Brandy
17:33 Carlile Sarah
17:39 burelis some Kyle Shannon Originals in
17:42 there too
17:48 [Music]
18:05 I'm selling my rabbit as memorabilia I
18:08 am keeping my rabbit I'm keeping my
18:10 rabbit that is a piece of History I like
18:13 my
18:14 rabbit I don't use it but it still makes
18:17 me
18:19 happy I had the original Palm Pilot I
18:22 had the original Newton the Apple Newton
18:25 remember the
18:27 Newton the Newton was going to change
18:29 everything with their graffiti
18:31 handwriting artificial intelligence
18:33 handwriting system so long as you
18:34 learned their stupid shitty ass 26
18:39 [Laughter]
18:45 letters um
18:49 [Music]
18:57 yeah what else do I sing Simon and gar
19:02 [Music]
19:07 funcle let us be lovers will marry our
19:11 fortunes
19:15 together I've got some real estate here
19:18 in my
19:19 [Music]
19:23 bag so we bought a pack of
19:27 cigarettes and Mrs swag
19:30 P we all come to look for
19:36 [Music]
19:50 America oh all right
19:56 so Sam Alman posted today which was
20:01 interesting they've come to some sort of
20:04 agreement with a national us AI Safety
20:08 Council or some [ __ ] like
20:11 that that uh they've agreed how they're
20:14 going to do their testing when they
20:16 released their big ass
20:20 models and I'm assuming that's something
20:24 saying
20:25 they've the
20:28 the hopeless Optimist in me is saying
20:31 they announced it today because they've
20:33 already done the testing but the
20:35 announcement said they've reached an
20:36 agreement to be able to do the testing
20:43 so
20:48 but I'm fairly confident we're
20:53 within 6 weeks of something new I've
20:57 been trying to explore critical thinking
20:59 with chat gbt it's like a circular
21:01 firing
21:04 squad well the fact that you the fact
21:07 that you find it like a circular firing
21:09 squad two block Tom is the critical
21:12 thinking part like your critical
21:13 thinking is okay it's not ready for
21:15 prime time right and you might be able
21:18 to prompt it in a way that you can get
21:21 better at it right maybe
21:23 you I don't know maybe you create a
21:26 series of gpts that call one another or
21:29 something like that right but the the
21:31 fact that it's not quite there means
21:33 you're doing critical thinking so I
21:35 would argue that you are
21:37 successfully living
21:39 [Music]
21:57 there you can make money Jud
22:02 time welcome to chat
22:04 add here's a gold button let's push it
22:07 let's see what it
22:12 does we have AGI not yet we don't unless
22:17 you're giving me a big compliment well
22:20 AGI I don't mind if I do listen I'm no
22:23 ASI you know what I mean this is not
22:26 super intelligence general intelligence
22:31 sure I have an arts
22:35 degree it's
22:38 fantastic oh you you weren't referring
22:41 to me as the AGI okay I I I probably
22:47 um I I made that mean more than
22:50 it you meant the the computer the the
22:53 chat GPT yeah yeah no I I no no offense
22:57 taken no no offense taken
22:59 I uh I am I am I am perfectly
23:02 comfortable with my my self-esteem no
23:06 I'm I'm not going to cry I am not going
23:08 to cry I'm not I'm not going to cry this
23:11 time
23:13 okay all right all right
23:19 people I didn't go to the Ivy Leagues
23:21 okay it was a state
23:25 school can I just say one thing though
23:29 when people are like where'd you go to
23:32 school especially when I'm talking to
23:34 people of a a certain
23:37 caliber I say I went to Penn
23:40 State and they go oh you of
23:46 pen in
23:48 Philadelphia and I just smile uh-huh
23:55 [Music]
24:04 oh that's very impressive wow I don't
24:07 like to brag so I
24:10 won't AI will make smart assistants
24:12 actually conversational totally totally
24:16 [Music]
24:18 will listen I can't help it if rich
24:21 people are stupid enough to think that
24:24 everyone went to an Ivy League school
24:26 it's not my it's not my fault the fact
24:29 that I don't correct them that's just
24:32 being opportunistic that means I got a
24:34 good
24:36 [Music]
24:41 education AI playing musical instruments
24:43 via figure botal robots with AGI that's
24:46 actually you know what
24:50 Peter it I I
24:53 [Music]
24:56 mean it was interesting there was an
24:58 interview with the with the CEO of
25:00 figure the figure robot
25:03 company and what he was talking about
25:05 was there's like 9 million jobs that no
25:08 one
25:10 wants it's like it's like really bad
25:12 unemployment in those 9 million jobs
25:14 nobody wants them they're like all the
25:15 shitty jobs and that's what they're
25:17 going to Target first which I thought is
25:19 interesting
25:20 but at some point someone's going to
25:23 take some open- Source robot and make
25:25 a AGI robot band that's gonna [ __ ]
25:30 [Laughter]
25:39 shred you know and like trust fund trust
25:42 fund babies that want to front a band
25:44 but like they're such douches that no
25:46 one will play with them they'll just buy
25:48 some robots to to back them and then
25:52 they'll Sing Sing poorly up front
25:59 oh
26:01 man I didn't go to school I'm
26:05 degre I 100% GRE G GPT Kiss band
26:14 exactly oh man all right I guess we
26:17 should get started shouldn't
26:19 we yeah is this ever going to start what
26:23 what is this channel
26:25 I this this the AI learning lab is is he
26:29 the professor is this a is he's a he's a
26:32 scientist what what are his
26:35 qualification oh he he doesn't he
26:39 doesn't have
26:40 qualifications why why is everybody
26:43 here cuz his dog sings his he's got a
26:47 singing
26:49 dog what what's that got to do with AI
26:52 is is it an AI dog it's an AI musical
26:55 dog it's a real dog I'm so conf
27:02 confused what do you learn here you you
27:06 don't
27:06 know why do you
27:09 come you don't
27:11 know you're
27:14 weird you're what you're not weird
27:17 you're irregular I don't understand this
27:20 channel so if you don't understand this
27:22 channel either me me neither I don't
27:25 either I don't get it just hang out hang
27:28 out out you're going to love it we're
27:31 going to do such good
27:33 fun we're going to do such good fun hey
27:36 you want to learn about chat JP we could
27:38 do chat JP we could do claw artifacts we
27:42 could do perplexity Pages we could do
27:44 mid Journey images we could do videos
27:47 videos videos I got some credits to
27:51 [Music]
27:52 spend why am I talking like this I don't
27:54 know why
27:56 not this Channel's like SNL when SNL
28:01 wasn't
28:02 funny get it cuz it's never funny
28:06 anymore no
28:08 anyway this is like that but with
28:11 shittier characters and worse
28:15 scripts fantastic Bob tell everybody's
28:18 won we can't do surveys anymore
28:25 why why can't we do surveys do I need to
28:27 enable something is that on
28:35 me poll and gift
28:38 vote
28:40 poll quick
28:43 poll I see polls no button for
28:47 surveys maybe they changed the name to
28:50 polls cuz it's an election year get
28:53 it it's a survey in non-election years
29:01 okay all right so here's what we do my
29:03 name's Kyle Shannon by the way this is
29:05 AI learning lab if you have questions
29:07 about AI pop them in the comments below
29:08 we got mods in here that'll pick those
29:10 questions up and they'll pin them so I
29:12 can see them most
29:14 times my ADD is so bad I'm looking at
29:19 like [ __ ] crawling up my walls rather
29:21 than the comments so I may miss them
29:23 doesn't mean I'm not going to get to
29:24 them except I can't see because I have a
29:27 janky eye oh no
29:29 janky ey is no
29:31 good it's okay if your AI is janky but
29:34 it's not good if your eye is
29:37 janky and can I just tell you this don't
29:39 get old your eyes just stop working just
29:43 hairs grow out of parts that you just
29:45 never thought a hair
29:50 should what's the coolest AI thing
29:53 that's happened this
29:56 week well
29:59 I think the most significant thing that
30:01 happened this week is that the
30:03 information essentially
30:05 confirmed project strawberry gp5 project
30:10 Orion that open AI indeed
30:15 has and has had for a year essentially
30:18 they they had it since last November so
30:20 they've had it for 10 months they've had
30:22 they
30:25 have a next level
30:30 model and I'm not even going to call it
30:31 a large language model because I think
30:33 it's going to be
30:35 multimodal and I think it's going to
30:37 have reasoning on it and and reasoning
30:39 is a big [ __ ] deal I don't think
30:43 people understand how big a deal it is I
30:45 don't think I understand how big a deal
30:47 it
30:48 is but
30:50 imagine if your chat GPT could actually
30:53 analyze its answer it gave you and keep
30:58 whacking at it till it got it
31:00 right that's that's kind of what's
31:04 coming um it it looks like it's coming
31:07 this fall the definition of
31:10 fall is from this weekend
31:13 until I don't know
31:20 Thanksgiving but that means in the next
31:22 three months we're probably going to
31:24 have some some crazy new model we're
31:27 also going to have Claude 3.5 Opus which
31:30 is their big model Claude 3.5 sonnet's
31:34 what we've got right now that everyone's
31:35 losing their mind over so when the big
31:37 the big model shows up that should be
31:39 even
31:40 better um December let's call it January
31:44 or February you're going to have grock
31:46 the new version of grock that's being
31:47 trained on 100,000 h100 gpus right now
31:52 so that's going to be coming so we've
31:54 got we got some [ __ ] coming so for me
31:58 that's that's the that's the most
32:00 significant thing this that happened
32:01 this week but it doesn't mean anything
32:05 because we've
32:07 been pontificating and predicting
32:09 something would happen with this [ __ ]
32:11 project strawberry for a month now and
32:13 we got nothing we get nothing got
32:16 nothing we got pink
32:18 bows this isn't even [ __ ] related to
32:21 project
32:22 strawberry this is ULU RX sent me some
32:26 gifts including this lovely
32:32 that look
32:36 pretty all right how does someone who is
32:39 not a traditional artist break
32:41 into the art community real question uh
32:45 see if you had asked a fake question I
32:48 could answer it now that I know that's a
32:50 real question I'm not qualified
32:56 [Laughter]
32:59 okay here's the secret to being an
33:02 artist make
33:05 [ __ ] and the secret to breaking into the
33:08 AR Community is make [ __ ] and share [ __ ]
33:11 now if the [ __ ] you're going to make is
33:14 going to be AI
33:16 generated understand that the art
33:19 community RIT large is going to hate
33:22 [Laughter]
33:25 you it's just where it is right now
33:31 that doesn't mean you shouldn't do it
33:33 when the digital camera came
33:35 out photographers lost their [ __ ]
33:38 mind and they said oh no only real
33:42 photographs in this show
33:44 M take your digital photographs over
33:47 there to the digital ghetto you can show
33:50 your little pretty pixel art on the
33:53 screens over there this is where the
33:55 real photography
33:57 is so we're in this we're in the middle
33:59 of the same [ __ ] thing still very
34:02 much a rookie what's the best free GPT
34:05 to use right now chat GPT chat GPT gave
34:08 you access to gp4 Omni um it's a little
34:11 bit hobbled um it's a little bit
34:13 restricted but it's the best uh second
34:16 best would be Claude
34:20 claw. third best probably
34:25 perplexity um I think perplexity
34:28 has a free version I'm pretty sure
34:30 Claude does I know chat GPT does those
34:33 are the biggies just start honest to God
34:35 start with chat
34:37 GPT it's got Vision it's got I mean we
34:40 could go play with chat
34:45 GPT if you want
34:48 it here here's the deal okay this is
34:51 this is actually
34:55 important we're at a stage now
34:58 with all of these AI tools that for
35:01 every category there are
35:11 multiple trying to think I'm trying to
35:13 think of the right adjective here there
35:15 are multiple I was going to say world
35:17 class but that's not quite right it's
35:19 not quite
35:20 state-of-the-art there are multiple
35:22 Tools in every
35:24 category that are essentially as good as
35:27 one another
35:30 so there's some people that like Claude
35:31 3.5 more than they like chat gbt other
35:35 people still swear by chat gbt um you've
35:39 got Google Gemini some of their models
35:41 are really good they're really hard to
35:42 find they're all [ __ ] all over the
35:44 place they're hard to find but they're
35:45 really good if you are into open source
35:48 and want to Tinker the Llama models are
35:51 really good perplexity is really good
35:53 for research cla's really good for
35:56 coding and for writing
35:59 um chat GPT is really good for that then
36:00 you can go down the image route and
36:02 there's probably 20 different tools
36:05 right now that are incredible at image
36:08 generation there's
36:11 probably three or four video tools that
36:15 are remarkable at video generation
36:18 there's three that I know of song
36:21 generation tools suo Udo and Loud
36:25 me loud. me or loud me
36:28 AI or some [ __ ] like that that's a new
36:31 one but you can just put in a prompt and
36:35 it generates a [ __ ] song with lyrics
36:38 and words and singers sing singers
36:44 singing you know there's programming
36:46 tools that are staggeringly good there's
36:48 all sorts of really specific tools like
36:51 presentation Builder so the tool scene
36:54 right now is massive so it's it's almost
36:57 like you can't go
36:59 wrong what you where you can go wrong is
37:02 to not do
37:04 anything
37:06 so what's the easiest principle to ask
37:09 it to do okay Brian wise so here's the
37:11 here's the big here's the big unlock
37:14 with all the large language
37:17 models it's not
37:20 Google Google for 20 years trained us
37:23 put as little as possible in that stupid
37:26 [ __ ] box right
37:28 you know uh automated sprinkler systems
37:33 lawn and then you know it would go find
37:35 you all the lawn sprinkler
37:37 systems chat
37:39 GPT needs a lot more context and so the
37:43 big unlock for chat GPT you can go read
37:46 about prompt engineering and there's all
37:48 sorts of methodologies with give it a
37:50 roll and there's acronyms it doesn't
37:53 [ __ ] matter as far as I'm concerned
37:55 here's the big unlock
37:58 talk to chat GPT as
38:01 if it were the best educated most
38:05 clueless intern you've ever worked
38:10 with seriously think of it like just an
38:14 absolute adult that knows nothing about
38:16 the world but happens to have memorized
38:19 everything if you had an employee like
38:22 that how would you talk to them well
38:23 you'd have to give them a lot of context
38:25 okay okay Jimmy now you're going to act
38:28 like a college professor and you're
38:31 going to teach me about marketing you
38:34 know you're a marketing Professor that
38:36 works at the Wharton business school and
38:38 you're going to you're going to talk to
38:40 me about marketing strategies and then
38:43 it'll spit out some things and you're
38:44 like oh before you give me those
38:46 strategies let me tell you a little bit
38:47 about my business and what I'm trying to
38:48 do and then it'll give you some more
38:50 things you're like okay that's still not
38:52 bad but but uh I want to go after you
38:55 know this particular audience and so the
38:58 more you talk to it like in a back and
39:01 forth way just imagine that you're on a
39:05 that that you have a remote
39:08 employee and just start talking to it
39:10 like a remote employee and if it gives
39:12 you shitty work say ah
39:16 Jimmy what that's pretty shitty
39:19 work and tell it why it's shitty and I
39:22 I'm sorry about that and it'll and the
39:25 more context you give it the better and
39:27 better and better it gets
39:29 Kyle what you just said would be
39:31 perfect a perfect 60-second Tik Tok
39:34 that's not a bad idea okay that's
39:37 good that's good I like
39:39 it I know I'm a little stuck on on what
39:42 Tik tok's to make right now so I'll get
39:45 there here's the here's the thing the
39:47 month of August for me is historically
39:51 has been a productivity [ __ ] hole just a
39:53 just a [ __ ] black hole of nothingness
39:55 of getting nothing done so I am living
39:59 up to my my my annual take August off
40:03 but don't go on vacation copy even
40:05 created by a human can be detected as Ai
40:09 and if we provide
40:11 it to
40:13 clients that state of AI wait that that
40:17 state
40:18 AI di I don't quite know what that
40:23 means yeah so here's the thing about AI
40:27 detectors
40:28 it doesn't [ __ ] matter it doesn't ma
40:34 okay when you put something out in the
40:37 world you're responsible for
40:41 it
40:43 okay so if you go to chat GPT and you
40:45 say write me an article it's going to
40:48 write you a shitty article and if you
40:50 don't read that and you copy it and
40:53 paste it and post that to LinkedIn
40:55 you're going to look like a [ __ ]
40:56 idiot cuz every is going to go that's
40:59 [ __ ] AI That's
41:01 horrible but if you use your critical
41:03 thinking if you use your thinky
41:07 brain there's an infinite number of ways
41:10 you can write an article for LinkedIn
41:11 and one of them involves using AI if
41:14 you're smart if you learn how to
41:16 interact with it in ways that it gives
41:18 you structure and it gives you ideas and
41:20 it gives you context and it helps you
41:21 frame up arguments and then you actually
41:23 look at what it wrote and you're like oh
41:25 this is redundant let me actually go in
41:26 and edit this and fix
41:28 this and then you take that and you put
41:31 it on the internet and it doesn't suck
41:34 then it doesn't
41:36 suck it's not anyone's business whether
41:39 or not it was written with
41:41 AI what is their business is if you put
41:44 shitty content out there they're going
41:46 to they're going to um what did I do to
41:49 my arm I just noticed that was that was
41:52 the uh I got in a I got in a tussle I
41:55 got in a tussle with a professional
41:57 wrestler
41:58 and uh you know he he scratched me he's
42:01 like I was like that doesn't seem very
42:04 wrestler like and he yelled at me a lot
42:07 and I'm like okay I'll I'll go
42:10 away that was this afternoon I forgot
42:12 about that
42:16 um all right the Council of Europe is
42:20 hosting a conference on AI interesting
42:22 don't know anything about that I'll look
42:23 into it why does it keep saying I've
42:26 just joined
42:28 why do you keep just joining
42:38 steo beautiful Danielle thank you for
42:41 that reminder Brandon let's see what we
42:43 got going here Oprah Prime Time
42:46 interview details and news Okay prime
42:49 time oh Oprah doing a uh who she
42:51 interviewing some AI big wig right Brian
42:55 wise there are some good guilds in our
42:57 community the salon. a okay so let's
43:00 let's do a couple of
43:01 things
43:08 [Music]
43:11 um okay if you're new
43:16 here this
43:18 group this
43:21 community you need to be a part of it
43:24 especially if if you're curious about AI
43:28 if
43:31 you if you're really into Ai and no one
43:33 else in your life is this is your group
43:37 so this is a community that I started
43:40 the week after chat GPT came out with
43:43 Leah fasten who's a photographer out of
43:47 Boston
43:48 and we've got an online community I
43:51 think we're up to like 1,800 people at
43:52 this point and it doesn't cost anything
43:56 and the people in this community are
43:58 quite remarkable and so if you go to the
44:04 salon. scroll down a little bit you'll
44:06 see a button that says join our
44:07 community and then if you join our
44:09 community if you if you click on that
44:11 button it pops you
44:15 into this site which
44:18 is based on the mighty Network or yeah
44:22 Mighty yeah Mighty networks
44:24 platform so it's kind of like you know
44:26 Facebook or Discord or something like
44:29 that and there's sort of a little
44:32 six-step getting
44:34 started
44:35 process so there's a welcome page where
44:38 we talk about the journey to AI literacy
44:41 play first mindfully create generously
44:43 lead we talk about what that means the
44:46 next step is to introduce yourself to
44:48 the community you're like why would I
44:50 introduce myself to the community I
44:51 don't even
44:53 know because you should
44:58 because it'll be good for you just like
45:00 eating your
45:02 spinach some for some people introducing
45:04 themselves to a group of strangers is
45:07 you know you'd rather jump off a cliff
45:10 but this group of strangers is
45:12 incredibly generous what what you'll
45:14 learn is if you go to our values which
45:16 is Step number three our values are
45:19 based on who this community already is
45:21 and what we want to
45:24 retain so this is a group of people that
45:26 are curious they're constantly trying to
45:28 figure out what's the next thing in AI
45:30 how do we do this they're playful
45:32 they're willing to play with
45:34 it not be married to
45:38 outcomes this is a group that's
45:39 remarkably
45:42 empathetic remarkably
45:45 empathetic sometimes people come in here
45:47 they'll come into the the Tik toks here
45:49 they'll come into the community they'll
45:51 either be really scared or they'll be
45:52 really pissed off and this is a
45:55 community that doesn't
45:57 get on soap boxes and and platitudes and
46:00 judge people for being afraid of AI this
46:03 is a group that says what are you afraid
46:05 of tell me about that why are you afraid
46:07 oh I I understand that you're afraid
46:09 you're going to lose your job yeah
46:10 that's [ __ ] terrifying you want to
46:13 see something
46:15 cool so so it's a group of AI optimists
46:18 empathy is a a big part of what this
46:21 group's about
46:22 bravery being willing here here's the
46:25 thing
46:28 AI right now is very very intimidating
46:33 if you don't know anything about AI
46:35 especially if you're not
46:37 technical what you know about AI is what
46:40 you've heard for the past two or three
46:42 decades oh you've got to be a
46:44 mathematician you have to have a degree
46:46 from Stamford or
46:48 MIT historically that was
46:51 true but starting November 30th
46:55 2022 with the launch of chat
46:58 GPT all of a sudden machine learning and
47:01 artificial intelligence has been
47:05 revealed handed to the rest of
47:09 us so it takes a certain amount of
47:12 Bravery to make the leap to say even
47:15 though I don't have a degree from MIT
47:18 I'm going to go for it and what you'll
47:20 learn and what this group will inspire
47:24 you um and if you just hang out on these
47:26 Tik Tok night after night and you you'll
47:29 learn it here as well is that there is a
47:32 tremendous amount of
47:34 Joy when you start to unlock what
47:38 generative AI
47:40 is it's it's incredibly joyful it's
47:43 incredibly inspiring it's incredibly
47:47 creative it's incredibly transformative
47:50 on a personal
47:52 level but you got to take that leap and
47:55 this is a group of people who've taken
47:57 that leap and then the last value that
48:00 we we really honor and encourage is
48:03 generosity and generosity looks like hey
48:06 I just learned something you want to
48:07 learn it too let me start a guild so I
48:10 can teach you what I know I like I I
48:12 like I like learning this stuff and I'm
48:14 going to I'm going to tell you what I
48:15 learn along the
48:19 way or oh you're new here and you don't
48:21 know anything here's three places to
48:23 start here's a here's a good tool to go
48:26 play with
48:28 if you run into Source Camp she'll show
48:30 you parlor
48:32 [Laughter]
48:35 tricks so go check out the
48:38 salon go introduce yourself you can
48:41 watch past meetings you can RSVP for our
48:44 next events you can join jump into the
48:46 water cooler chat where we can talk
48:48 about stuff and
48:55 nonsense move over mid Journey free pick
48:58 Mystic is the most realistic AI
49:00 generator I've tried this is the thing
49:03 where we're now at a stage where things
49:06 you've never heard of are now as good as
49:09 mid Journey right and you know uh flux
49:12 just came out and it's as good as mid
49:14 journey and everything's as good as mid
49:15 Journey which generally means if they're
49:17 saying it's as good as mid Journey mid
49:19 journey is still the king of the hill
49:20 but you now
49:22 have multiple things that are that are
49:25 getting there Salon is awesome
49:28 from Ellis Bruce
49:30 Paul is that one of you or three of you
49:34 hello I'm Ellis I'm Bruce and I'm
49:41 Paul they're a trio out of
49:46 England just
49:51 me I've got two I've got two first names
49:54 too Kyle Shannon although mine are girls
49:56 name
49:58 so oh man I tried both of those on free
50:01 pick the problem with free pick is you
50:03 can't use this Mystic model unless
50:05 you're a paid subscriber so it ain't
50:07 free don't name your product free pick
50:10 if you don't have a free version you run
50:13 around telling people oh you should go
50:14 try this thing you go to try it and it's
50:16 like well you need to you need to
50:17 upgrade there you big dumb
50:19 dumb it's
50:23 exhausting well just 20 bucks a month
50:25 yeah but there's like 42 20 Buck a month
50:31 subscriptions all right let's
50:37 see Peter wants to play with Runway
50:40 ml Peter who is Brandon are you talking
50:42 about Peter
50:45 Kaminsky we can go we can do go do some
50:49 uh he wants to play with oh with Runway
50:52 yeah we can go play with some
50:54 Runway All Right fantastic Bob tell him
50:57 what he's
50:58 won uh yes he uh he hasn't won anything
51:03 because this isn't a game
51:04 show yeah well that's
51:07 disappointing tell me about it I don't
51:10 know we
51:14 have I can't even I can't even make
51:17 stupid jokes
51:21 tonight
51:23 um no new Peter okay um
51:29 we could do cling dream machine actually
51:32 that this might be an
51:34 interesting thing so on the on the uh on
51:38 the live last night we had we had two
51:40 other Tik Tok content creators joined us
51:43 AJ and and Ralph Ralph right Ralph
51:49 Ralph was it Ralph I don't know Ralph
51:53 Miguel someone
51:59 um and what AJ was talking about was
52:02 he's doing these
52:03 automations let's play with Runway ml
52:06 okay yeah we can do
52:08 that oh you can use the flux fast model
52:12 I know but you can't use Mystic that new
52:14 the new thing I've got on screen right
52:15 here you can't use but anyway what AJ is
52:18 doing is he's making automations that
52:22 make 60-second videos that he then puts
52:25 up on Tik Tok and those channels are are
52:27 getting a lot of views and how he's
52:29 doing it is he's doing an API call for
52:36 Mid Journey images and then he sends
52:39 that mid Journey image to three
52:40 different video generators so what might
52:43 be kind of fun is we go do something
52:48 on Mid Journey make some
52:51 images and then we'll go to those three
52:53 we'll go to Runway ml cling
52:57 which I think I have some credits on
53:01 and um Luma Luma lab's dream machine so
53:05 we'll go to three different image video
53:07 generation models we could probably
53:09 throw another one in there and just like
53:10 compare them let's just see what they do
53:13 Ralph miguel. tools that's right dream
53:17 machine Blends images together so well
53:19 and it's free oh cool okay great
53:24 awesome Okay so
53:32 what was I playing on that had flux that
53:35 was pretty
53:36 good free pick had
53:40 it oh it was Korea wasn't it Korea
53:43 doesn't Korea have free pick now let me
53:47 see Korea I mean flux
53:50 now yeah we're going to leave that
54:00 this is korea. k a.
54:08 a yeah flux generate images
54:14 Okay so let's just roll the dice we'll
54:17 do a
54:19 16x9 wide image and we'll
54:22 say roll the dice
54:30 generate that's pretty [ __ ]
54:35 quick all right I think we need you know
54:37 what we
54:39 need all right let me let me show you
54:43 before we get into Tool playing Salon
54:45 has great values and I have Workforce of
54:49 about 60 employees I used to
54:53 get them interested in an AI have them
54:56 join the [ __ ] Salon the listen so let
54:59 me show you another thing if if you're
55:00 if you're interested in the salon you
55:02 should
55:04 also what what we're doing here on this
55:08 channel and I'm I'm in the middle of
55:10 figuring out what this even
55:14 means but but we've hit we've hit an
55:18 inflection point in my opinion with AI
55:21 tools where the tools are all really
55:24 [ __ ] good like there was there was a
55:26 year after chat GPT came out where you
55:30 had to pay attention to all the tools
55:32 you didn't have to but if you wanted to
55:33 be in the AI conversation you kind of
55:36 had to know what sucked and what didn't
55:38 and what's happened over the summer is
55:41 that AI uh open AI has stopped releasing
55:45 [ __ ] like they haven't released any
55:46 significant upgrade to their stuff in um
55:52 six months they there's been a lot of
55:54 announcements there's been a lot of
55:55 delays it's you know and some big shit's
55:58 coming blah blah
56:00 blah in those 6 months lots and lots of
56:04 people have caught up in the largest
56:06 language model space lots of people have
56:08 caught up in the image generation space
56:09 lots of people have caught up in the
56:11 video generation
56:14 space and
56:16 so what I'm what I'm in the middle of
56:19 exploring and and figuring out how we
56:21 how we talk about more in this
56:24 channel is not the skills of how you do
56:29 stuff but it's more the aptitude of who
56:32 you're
56:34 being to be able to be successful at Ai
56:37 and and the how how I'm phrasing it or
56:41 how I'm categorizing it is Rise of the
56:44 agile
56:45 human um and and what that looks like is
56:49 you want to find people in your
56:51 organization that that are really high
56:53 in these in these attributes curiosity
56:57 drive or bravery like willing to take a
56:59 risk creativity people that are good at
57:01 multitasking people that are good at
57:03 horizontal thinking right you might have
57:05 people that are super good at vertical
57:08 like a specific
57:09 skill well if AI gets good at doing that
57:13 skill then that person's going to be
57:15 really limited but if they can also
57:17 think horizontally across disciplines
57:20 that's going to be more valuable are
57:22 they good at critical thinking are they
57:24 good at looking at hey is chat P PT is
57:27 what it's doing actually any
57:30 good and if it's not good how can I use
57:33 my critical thinking to learn how to
57:35 talk to it so what I get from it is
57:38 good because if you
57:41 start really exploring these tools with
57:44 these kind of mindsets you can you can
57:47 get better at this
57:49 [ __ ] um discernment taste generative AI
57:53 generates
57:54 [ __ ] monumental mountains of content
57:59 that you then have to sift through and
58:01 figure out what's good what's
58:02 not hang on I got a puzzle piece I got
58:05 to put in place thanks Tick Tock you're
58:08 an [ __ ] um you know what's your
58:10 judgment you know is this is this stuff
58:12 good is it good for your
58:14 business um do you have adaptability are
58:17 you willing to change things on the
58:22 Fly the next 3 years
58:26 are going to be
58:28 chaos not because the
58:31 okay I'm I'm starting to have some
58:33 clarity around
58:35 work what we hear a lot right now from
58:40 people that are afraid of AI and people
58:42 who not are not playing with AI is AI is
58:45 going to take your
58:48 job and I think it's much more Nuance
58:51 than
58:52 that I think what AI is increasingly
58:54 going to do is it's going to automate
58:57 the tasks that currently make up your
59:01 job and if it automates all of the tasks
59:05 that make up your job and you've got no
59:08 other skills and you don't have these
59:10 aptitudes then yeah it's going to take
59:11 your
59:12 job
59:14 okay but what we don't know as
59:17 individuals and what businesses don't
59:19 know is what tasks is it going to
59:21 automate first and so I think what's
59:24 going to happen is it's not going to do
59:26 it all at once so it's not just going to
59:28 come in and just be good at copywriting
59:30 now copywriters are gone it's going to
59:32 come in and be good at these six aspects
59:35 of the 23 tasks that you do as a
59:38 copywriter and if if you don't recognize
59:40 that okay those six tasks I can offload
59:43 those to AI but these other things I'm
59:45 still required for and I think that mix
59:48 of what AI automates in the tasks of
59:50 each of the jobs that's where the chaos
59:53 is going to
59:54 live and so the reason you need to find
59:57 people in your organization that have
1:00:00 this agility
1:00:02 gene or they can develop
1:00:05 this these are going to be the people
1:00:08 who are really good in an environment
1:00:11 where the the rules keep
1:00:14 changing the rigid people the people
1:00:16 that just want to do their one thing
1:00:18 that they're going to be in the world of
1:00:21 hurt and so what's weird about this is
1:00:24 the people that have these skills
1:00:27 that I think are going to make them the
1:00:29 most valuable people as we move into
1:00:32 this chaotic
1:00:34 World they may not be in positions of
1:00:36 power you might have an administrative
1:00:38 assistant who's [ __ ] brilliant at
1:00:41 this [ __ ] I would get that
1:00:43 Administrative Assistant learning AI
1:00:48 immediately corporate has taught us to
1:00:50 find busy work where AI does the busy
1:00:53 work we get to do the essential work
1:00:55 yeah exactly this this is these
1:00:58 attributes are where that essential work
1:01:03 happens
1:01:05 so I just I wanted to share this
1:01:09 because what we're about to do with
1:01:11 tasks is I'm going to get a number of AI
1:01:16 tools to vomit out some [ __ ] and then
1:01:19 we're going to take that content vomit
1:01:21 and we're going to curate interesting
1:01:23 stuff out of it and then we're going to
1:01:25 refine that and go make videos and maybe
1:01:27 go make a song Whatever we're going to
1:01:28 do we're going to navigate between I
1:01:31 don't know six or seven different
1:01:33 tools generate some
1:01:36 content and I just want you to pay
1:01:39 attention within yourself as I'm going
1:01:41 through this stuff what is your
1:01:43 relationship with with all this are you
1:01:45 being oh thank you very much that's
1:01:47 really generous thank you so
1:01:50 much you know are you curious are are
1:01:53 you like on Twitter going hey what's
1:01:55 coming what's next what's next if
1:01:58 not start working on
1:02:00 that I gave an exercise at the beginning
1:02:03 of the week on my on my Tik Tok channel
1:02:06 for people to go take a
1:02:08 walk with chat GPT and put chat GPT in
1:02:12 conversation
1:02:14 mode and have a conversation about
1:02:17 something you know nothing about and let
1:02:19 chat educate you just let me do the one
1:02:23 thing I like and know I just joined and
1:02:26 I already feel like I'm being
1:02:31 attacked I know I know I listen I don't
1:02:34 disagree with you pate and it doesn't
1:02:38 mean that all Specialists are going to
1:02:40 be penalized but some Specialists are
1:02:42 going to be
1:02:44 penalized um especially those
1:02:46 Specialists that are doing something
1:02:48 where the bulk of the tasks for their
1:02:52 work um get automated out um I think
1:02:54 you're probably fine for a week or two
1:02:56 Pate I I think you know what I think
1:02:59 probably through the end of September
1:03:00 you're good you're probably you're
1:03:02 probably still solidly employed so
1:03:04 you're
1:03:06 fine oh man okay so let's go play any
1:03:11 questions on that Kyle I need your
1:03:14 advice hit me what's up Jolo is it Jolo
1:03:17 Joo oh gillo gillo what advice do you
1:03:20 need I'm happy to share I I okay here's
1:03:24 the thing that was pretty what was
1:03:29 pretty did my pink
1:03:31 bow question I replacing humans examples
1:03:34 self checkout at supermarket have a lot
1:03:37 of issues
1:03:39 scanning yeah see that's
1:03:42 just that's not AI replacing humans
1:03:46 that's corporate greed saying we don't
1:03:48 have to pay instead of paying employees
1:03:51 let's get our customers to do the work
1:03:53 for us I would argue that that's not AI
1:03:56 replacing humans that's shitty corporate
1:04:01 Behavior you know if you ever hear
1:04:03 billur talk about self checkout and you
1:04:07 know and going into the sandwich shop
1:04:09 and they're like you know I want I want
1:04:12 mayonnaise and mustard on it and it's
1:04:13 like the mayonnaise is um over there and
1:04:15 he's like um you know you're behind the
1:04:18 counter and I'm um paying you money and
1:04:21 maybe you could um put the mayonnaise on
1:04:24 my [ __ ] sandwich
1:04:28 um J silow I'm starting a boot camp next
1:04:30 week for AI machine learning okay cool
1:04:34 you're starting it yourself meaning
1:04:35 you're going to teach it or you're
1:04:36 starting someone else's boot
1:04:39 camp and what advice do you
1:04:43 need uh uh uh uh AI is knowing when you
1:04:48 pick up an item and if you put it back
1:04:50 or not that's actually really good J
1:04:52 Silo that was a beautiful Galaxy gift it
1:04:55 really was it was really amazing thank
1:04:56 you and what's your question I'm happy
1:04:58 to answer
1:05:03 it J Hamilton AI is knowing when to when
1:05:06 you pick up an item and if you put it
1:05:09 back or not that's brilliant that's
1:05:10 really good that's very much what it is
1:05:12 that's what we'll do here in a second
1:05:14 you're always pretty Kyle thank you
1:05:16 thank you I tell myself that every
1:05:20 day got a screenshot of the pink bow
1:05:23 thank goodness cuz there's not enough of
1:05:25 them t-rock before before my ass hit the
1:05:29 couch the first night I wore I wore the
1:05:31 pink bow t-rock had taken a whole
1:05:34 section of my video and reposted it out
1:05:37 on Tik Tok of just me and a pink [ __ ]
1:05:39 bow my wife was like what are you doing
1:05:41 you've jumped the
1:05:43 [Laughter]
1:05:47 shark all right let's see J silow uh
1:05:50 someone else someone else's boot camp
1:05:54 intellectual AI ml intellectual point
1:05:58 you can find them on Google oh okay cool
1:06:00 so what's your
1:06:04 question I've T talked to Jet chat GPT
1:06:07 longer this week than any human yeah
1:06:11 when you start when you really start
1:06:13 interacting with these things like you
1:06:15 have a partner a therapist an intern an
1:06:20 employee um it it really does start to
1:06:23 shift how you think about these large
1:06:24 language models and you'll get much
1:06:26 better at
1:06:29 it I feel pretty oh so pretty can you
1:06:34 get a job with the
1:06:35 certifications
1:06:39 um okay
1:06:41 so this is so what you're asking jilo is
1:06:44 a really
1:06:45 important question and
1:06:48 it's it's got a really nuanced
1:06:53 answer you absolutely do not need any
1:06:56 sort of
1:06:59 certification um
1:07:03 or
1:07:05 degree to be able to get a
1:07:08 job that involves being good at
1:07:11 generative AI now here's the massive
1:07:16 exception if the certification that
1:07:18 you're getting is in developing machine
1:07:21 learning algorithms and and building
1:07:24 models and doing fin tuning and creating
1:07:27 embeddings and doing things like Lang
1:07:28 chain and building
1:07:31 applications then you're working in a
1:07:34 part of AI and machine learning that's
1:07:36 the part that's been around for 30
1:07:39 years
1:07:41 so that area if you're going to go into
1:07:44 the engineering side of machine learning
1:07:46 and and
1:07:47 AI then yeah having a certification from
1:07:52 a trusted Source or you know having you
1:07:55 know a uh a certificate from Stamford or
1:07:58 MIT or Google or Microsoft one of these
1:08:01 things that that's in the if if you're
1:08:04 working in kind of kind of the
1:08:05 underpinnings of m machine learning then
1:08:07 it makes
1:08:08 sense if you're talking about generative
1:08:11 AI if you're talking about how to use
1:08:14 chat gbt and how to use Claud for
1:08:16 programming and and how to make images
1:08:20 like we're about to do and and make
1:08:22 videos and make songs and Stitch all
1:08:24 this stuff together and understand the
1:08:27 implications on all of the work that's
1:08:30 about to change not only do you not need
1:08:33 a certificate for that you can't get one
1:08:36 right now I mean you can get one but
1:08:38 they're going to be absolutely useless
1:08:39 because everything is changing so fast
1:08:42 so there's there's very much two
1:08:44 different worlds what this channel is
1:08:46 focusing on is the post chat GPT kind of
1:08:50 generative AI use the tools rather than
1:08:54 build the tools so if you're in the
1:08:55 build build the tools side of the game
1:08:58 those certificates make a difference if
1:09:00 you're in the how is AI going to impact
1:09:02 the workforce moving forward that's
1:09:06 essentially the wild west right now and
1:09:08 that's where in my opinion tons of
1:09:10 opportunity is because nobody knows what
1:09:12 they're doing there because everything's
1:09:14 changing so fast so I hope that answers
1:09:16 your
1:09:17 question no one's paying people to
1:09:19 generate anything in AI uh that's
1:09:22 absolutely not true that's absolutely
1:09:24 not true there a lot of jobs right now
1:09:27 where people are using Ai and people are
1:09:29 getting promoted within companies
1:09:32 because they know how to generate stuff
1:09:33 with AI they're figuring things out
1:09:35 they're figuring out how to automate
1:09:39 um tasks within within jobs so I I
1:09:43 completely disagree with that there's
1:09:45 there's probably not a job on LinkedIn I
1:09:48 want you to come in and do generative AI
1:09:51 for my company but people that are doing
1:09:53 generative AI within companies are
1:09:57 accelerating their careers in dramatic
1:10:00 ways this channel is full of those
1:10:02 people I see it every [ __ ] day so
1:10:05 absolutely disagree with that one where
1:10:07 is the
1:10:09 Prophet The Prophet so okay you can make
1:10:13 money don't get trapped in this gold
1:10:16 [ __ ] button you can make
1:10:21 money the the myth that there is a magic
1:10:25 [ __ ]
1:10:28 just add water and you'll be
1:10:31 rich thing with AI is just not
1:10:37 true the way you make money with AI is
1:10:41 no different than the way you make money
1:10:43 with Photoshop or the may you the way
1:10:45 you make money with Microsoft Word or
1:10:47 the way you make money with a
1:10:52 shovel people are talking about you can
1:10:54 make money with
1:10:57 like it's actually a thing oh I just
1:11:00 learned chat jpt and I'll make all this
1:11:01 money
1:11:03 no you learn chat GPT and you understand
1:11:07 what it makes
1:11:09 possible and then you say oh I
1:11:13 understand insurance
1:11:15 refactoring and I could use chat GPT to
1:11:18 do insurance refactoring faster cheaper
1:11:21 or better than the competition and so
1:11:23 I'm going to use chat GPD to come up
1:11:25 with a whole new kind of product and I'm
1:11:28 going to sell it into the insurance
1:11:29 refactoring Market I don't even know if
1:11:31 insurance refactoring is a [ __ ] thing
1:11:35 but how you make money with any tool is
1:11:37 you figure out a problem that needs
1:11:39 solving and you solve it better than the
1:11:41 other
1:11:42 people you either do it faster cheaper
1:11:44 or better than the competition what AI
1:11:47 gives you the possibility of is you can
1:11:49 potentially do it faster better and
1:11:52 cheaper than the
1:11:54 competition that's where the [ __ ]
1:11:57 massive gains are is you come in and
1:12:00 someone's charging half a million
1:12:02 dollars a year to do whatever they do
1:12:04 and you walk through that door and say
1:12:05 for a 100 Grand I can turn it around
1:12:09 twice as fast give you twice as many
1:12:12 options that are as good or better for a
1:12:15 fifth of the c or uh for for yeah a
1:12:18 fifth of the cost not a fifth 20% of the
1:12:21 cost is that a fifth yeah
1:12:26 you'll get
1:12:28 work and then if you can get really
1:12:30 efficient at doing that
1:12:32 work you'll your 100 Grand will be way
1:12:35 more profitable than that other
1:12:37 company's 500
1:12:38 Grand anyone know how to have a markdown
1:12:41 table retain its form when pasted into
1:12:45 Google doc yes Hang on we're going to
1:12:49 we're going to take a detour
1:12:55 so let's
1:13:00 see make a markdown table
1:13:05 comparing
1:13:06 frankfurters wieners and hot
1:13:10 [Laughter]
1:13:17 dogs you're
1:13:20 welcome okay so here's how how I do it
1:13:23 if
1:13:24 you oops
1:13:30 if you just copy it
1:13:32 from if you just highlight it like
1:13:36 this and copy that and then go into
1:13:40 Google
1:13:43 Docs you should be
1:13:48 able I haven't done this in a while so
1:13:51 that's what I'm saying should
1:13:57 I should be able to just paste it in and
1:13:59 it'll retain its tabless okay it
1:14:02 didn't
1:14:04 fine
1:14:06 um does it do it like that
1:14:09 no
1:14:12 huh that's behaving differently than it
1:14:14 used to
1:14:18 Serena but if it's a
1:14:21 table you could certainly say insert a
1:14:24 table
1:14:27 and then paste [ __ ] in
1:14:29 there right
1:14:42 um
1:14:45 H I don't know anymore I used to know I
1:14:49 bet Peter Peter Kaminsky knows I bet
1:14:51 Peter Kaminsky knows
1:14:53 too I don't know anymore that that's
1:14:56 behaving differently than it used to all
1:14:57 right so let's go back to image
1:14:59 generation where's
1:15:01 Korea there it is
1:15:04 okay all right is there a way to have
1:15:06 Google mini Nest
1:15:08 use there's a special import
1:15:11 button under insert
1:15:20 oh insert
1:15:27 image
1:15:31 table Smart Clips
1:15:34 no put now building
1:15:38 blocks code
1:15:47 block I don't see it
1:15:50 P is there a way to have my Google mini
1:15:52 Nest use chat GPT for answer instead of
1:15:55 Google Assistant um probably if you hack
1:15:58 it but I I honestly don't know that's
1:16:01 something that I would go to Reddit and
1:16:03 ask on there I'm sure if it can be
1:16:04 hacked it's on
1:16:06 Reddit table to Google Sheets table to
1:16:10 Google Sheets then export it yeah you
1:16:12 could do that certainly yeah so if I do
1:16:15 if I do
1:16:18 sheets oops
1:16:28 oh come
1:16:36 on so now I'm in sheets and I should be
1:16:38 able to just paste it yeah so so if you
1:16:41 just paste it into Google Sheets now you
1:16:43 can probably I can probably copy this
1:16:46 from Google Sheets let's see I'll do
1:16:49 this we'll do
1:16:50 this we'll do this we'll do
1:16:54 this and then we'll do select all and
1:16:57 we'll do format
1:16:59 wrapping
1:17:00 wrap so there's our table and so now I
1:17:03 can grab this table copy I can go
1:17:08 to oh I was there here and I can go
1:17:13 paste paste
1:17:15 UNL boom there you go all right you can
1:17:18 do it that way sorry I didn't mean to
1:17:20 stop the class I thought someone I I
1:17:23 thought someone might know
1:17:25 way to go way to go
1:17:28 Serena she here's the deal she shows up
1:17:32 late she's like please don't Mark me
1:17:35 tardy but she's
1:17:37 tardy and then you know we're getting
1:17:39 ready to do pretty pictures and make
1:17:41 videos and she's like could you show me
1:17:43 how to do spreadsheet
1:17:46 thing it's all about me it's all about
1:17:48 you it's all about you it's always about
1:17:52 you um Okay so Ena is actually one of
1:17:56 the most selfless people in here she has
1:17:58 been in here every night since I started
1:18:00 doing these lives she is she is like uh
1:18:03 she's like the Cal Ripken of the AI
1:18:06 learning lab that's a compliment by the
1:18:09 way if you don't know who Cal Ripken
1:18:11 is Kyle can you talk about olama and
1:18:15 gpus um I don't know a ton about the um
1:18:20 open source world and and uh and the and
1:18:24 the technical of things I'm I'm not
1:18:26 that's that's an area where I actually
1:18:29 am decently unqualified that said I
1:18:32 would go into um the AI Salon go to the
1:18:37 salon. um and hang out in some of the
1:18:40 technical guilds um pate's in there
1:18:42 Peter kaminsky's in there there's a
1:18:44 bunch of people in there that are good
1:18:45 at this technical stuff that you can
1:18:47 kick those ideas around with okay let's
1:18:50 get back to
1:18:52 Korea okay so what we're going to do
1:18:54 here
1:18:57 the the request was can we play with
1:18:59 Runway
1:19:00 ML and what I'm going to do is I'm going
1:19:03 to I'm going
1:19:05 to bounce around a
1:19:08 lot in this sort of add kind of fluid
1:19:13 way this is like idea fluid right
1:19:16 there's gender fluid this is idea
1:19:19 fluid where rather than trying to get
1:19:22 too restrictive and say use this tool
1:19:24 and this prompt structure and this that
1:19:27 what I'm going to do is I'm going to try
1:19:30 to take advantage of the fact that a lot
1:19:33 of these tools are really good right now
1:19:35 all
1:19:39 right if you don't know AMA maybe you
1:19:41 should have a conversation with chat GPT
1:19:44 actually that's a really good idea for
1:19:45 ever hooked yeah just um
1:19:49 jillo put chat GPT in conversation mode
1:19:53 if you don't know how to do that
1:19:56 in in chat GPT the app in the right hand
1:20:01 side of the
1:20:03 um the chat box there's these little
1:20:06 headphones and if you click that chat
1:20:09 GPD goes into conversation
1:20:14 mode hey chat gbt what's going on can
1:20:17 you what why are you making
1:20:22 beats do I have hey Kyle
1:20:25 not much just keeping things Lively and
1:20:27 creative over here making beats yeah
1:20:30 that didn't sound
1:20:35 normal hey can you tell me about AMA and
1:20:41 gpus sure thing Al llama is a large
1:20:44 language model framework specifically
1:20:46 designed to be efficient and lightweight
1:20:49 it's optimized for running llms large
1:20:51 language models on G
1:20:56 yeah that's great but can you tell it to
1:20:58 me as if you're a
1:21:00 [Laughter]
1:21:07 rapper all right here we go yo check it
1:21:10 out it's the Alama flow got to break it
1:21:13 down let the see alama's like a
1:21:15 lightweight champ in the ring built to
1:21:17 run on gpus yeah that's the thing gpus
1:21:20 are like the beat behind the
1:21:23 bars so you can have you have stupid
1:21:25 conversations too
1:21:28 okay all right so here's what we're
1:21:30 going to
1:21:35 do at least one of the video models um
1:21:39 dream
1:21:40 machine has key frames in it meaning you
1:21:45 can start you can you can have a start
1:21:46 image and you can have an end image and
1:21:48 then you can you can have a blank start
1:21:50 image and and have it move toward an end
1:21:53 image or you can have it start at an
1:21:55 image and go wherever it wants to go or
1:21:57 you can do both so here's what I'm going
1:22:00 to
1:22:01 do I'm going to go
1:22:05 to we're going to go to chat GPT and
1:22:08 we're going to have chat GPT give us
1:22:10 ideas and I think I'm even going to have
1:22:12 chat GPT make us some
1:22:15 images but we'll also take the prompts
1:22:18 that chat GPT gives
1:22:20 us and we'll go to Mid journey and we'll
1:22:22 go to Korea
1:22:26 so that'll give us three different image
1:22:27 generation
1:22:29 models so we'll have chat GPT do the
1:22:32 heavy lifting so we don't have to use
1:22:33 our
1:22:34 brains because wherever you don't have
1:22:36 to use your brain anymore you shouldn't
1:22:39 and here's why and this gets back to
1:22:41 this idea
1:22:43 of elevating where you apply your
1:22:47 creativity if you apply your creativity
1:22:50 to the low-level blocking and tackling
1:22:52 of
1:22:53 ideation then you've spent that energy
1:22:56 this this is one of the things I've
1:22:57 learned in my life at least for the way
1:22:59 my brain works in any given 24-hour
1:23:02 period I have some fixed amount of
1:23:04 Creative
1:23:06 Juice some days it's 15 minutes some
1:23:09 days it's 15 hours but it is fixed and
1:23:13 when it's gone it's gone if I spend my
1:23:16 Creative Energy on lower level [ __ ] then
1:23:20 it's not there when I want to do higher
1:23:22 level [ __ ] so what I've been trying to
1:23:24 get in into a discipline of is let the
1:23:27 AI do the ideation work let it do the
1:23:31 concepting let it let it just spit a
1:23:34 bunch of ideas out and then I as a
1:23:37 creative can operate at a higher level
1:23:39 of critical thinking and discernment and
1:23:42 taste and go ooh that's good that's good
1:23:44 oh that's really good let me dig deeper
1:23:46 on
1:23:48 that all right so that's something that
1:23:51 I want you all to start working on is
1:23:55 lose your
1:23:56 ego about how [ __ ] brilliant you are
1:23:59 yeah yeah yeah we're all brilliant and
1:24:02 chat GPT can give us 20 ideas in like 30
1:24:05 seconds so let it do that don't have ego
1:24:08 about wow you can't really do really
1:24:10 creative work really have you used
1:24:14 it okay so we're going to go to chat
1:24:18 [ __ ] and I'm going to show you something
1:24:21 really amazing that Chad GPT has gotten
1:24:23 good at in the
1:24:25 past two
1:24:27 weeks like really good at and it wasn't
1:24:31 good at this for a long time and it was
1:24:33 frustrating and it forced us to speak to
1:24:36 it in a different way oh thank you very
1:24:39 much jillo that's really really generous
1:24:40 really appreciate that
1:24:45 um you can now give chat
1:24:49 GPT um run on sentences run on run on
1:24:54 idea requests that are that are building
1:24:57 nested nested Concepts and it'll just
1:25:00 [ __ ] crank through them so I'm going
1:25:01 to do that right now this is going to
1:25:02 sound [ __ ] ridiculous but that's okay
1:25:05 oh Champ's dreaming
1:25:12 okay I want you to act as a
1:25:16 cinematographer and
1:25:19 visual
1:25:22 effects designer
1:25:26 and I want you to help me come up with
1:25:31 concepts for short video
1:25:37 Snippets that have a dramatic turn in
1:25:41 them period the Snippets will start with
1:25:46 a single key frame and end on a
1:25:49 different key
1:25:52 frame where there will have been
1:25:55 some dramatic thing happened
1:26:01 period I want you to start by giving me
1:26:05 20
1:26:08 concepts of
1:26:10 short visual
1:26:12 scenes that start with one image and end
1:26:15 with a different
1:26:19 image and for each of the
1:26:22 20 scenes
1:26:27 describe a visual style that will most
1:26:30 bring that concept to
1:26:35 life then I want you to
1:26:37 select
1:26:40 three of those scenes that you created
1:26:44 that you feel are the strongest and most
1:26:47 compelling and tell me
1:26:49 why and then I want you to
1:26:52 write an image prompt
1:26:55 for each start image and end
1:26:59 image that describes the image in
1:27:04 detail
1:27:06 including the visual
1:27:10 style and then I want you to
1:27:14 make the start image and the end image
1:27:17 based on those image prompts for each of
1:27:21 the three ideas
1:27:25 okay that is what we call in the
1:27:29 business a horrible [ __ ]
1:27:33 prompt it's actually not that bad
1:27:35 because it's got a lot of context in it
1:27:37 I gave the thing a roll right it's going
1:27:39 to be this this designer kind of
1:27:41 thing but watch what chat jbt does now
1:27:44 it didn't used to do that you couldn't
1:27:45 give it this sort of idea vomit kind of
1:27:51 thing so
1:27:54 here's 20 Concepts so it's giving us the
1:27:57 20 Concepts let's see we'll just look at
1:27:59 them the vanishing act a magician on
1:28:02 stage end image an empty
1:28:04 stage the final Checkmate two hands
1:28:07 about to move a White Queen on a
1:28:09 chessboard end image the black king
1:28:11 toppling
1:28:13 over all right so it's doing on and then
1:28:16 visual visual style hyper realistic
1:28:19 low-key lighting the mirror check so
1:28:22 again we don't have to use our brains
1:28:25 and and maybe it doesn't get this right
1:28:27 maybe this is not the right concept but
1:28:29 here's 20 and then I told it after it
1:28:31 does 20 select the best three here are
1:28:34 the three strongest Concepts The
1:28:35 Vanishing act the burning letter and the
1:28:38 Shadows reveal here are the image
1:28:41 prompts for the start image the end
1:28:43 image right continue generating it it
1:28:47 sort of crapped out because it was too
1:28:49 much but it's going to keep going here
1:28:55 it's going to keep going here come on
1:28:58 come on
1:29:00 CH or it's
1:29:04 [Music]
1:29:07 not I'm going stop it I'm going to say
1:29:11 continue
1:29:15 continue all
1:29:17 right now it's going to generate the
1:29:19 images
1:29:28 all right there's that I'm going to stop
1:29:30 it I'm going to
1:29:32 say um I forgot to tell you I want to
1:29:36 make the images 16 by9 wide aspect ratio
1:29:40 because they're going to be used to
1:29:41 create videos
1:29:55 notice memory enhancements are now
1:29:57 activated yep I noticed that um memory
1:30:01 is weird I don't know I think they went
1:30:04 from 1,200 tokens to 8,000 I hope that's
1:30:08 the
1:30:11 case all right
1:30:15 there's there was an error generating
1:30:17 response regenerate let's see
1:30:42 let's see
1:31:05 [Music]
1:31:14 oh that's
1:31:15 cool so we start there and it goes to
1:31:17 that that's actually really cool okay
1:31:20 cool and then I'll say um regenerate to
1:31:24 start
1:31:26 image for the vanishing
1:31:32 act Kyle are you writing my book LOL my
1:31:35 book has some mystery letters in
1:31:39 it that's awesome that's awesome okay so
1:31:43 there's the vanishing act he's got it's
1:31:45 a magician with a bird that goes to that
1:31:48 image that's pretty
1:31:51 cool these are not great the the letters
1:31:54 are not great like I don't think that
1:31:57 that transitioning to that is going to
1:32:00 be I don't think that's going to be good
1:32:03 but that transitioning to that is kind
1:32:05 of
1:32:14 cool all right so let's
1:32:20 go start image PRP the closeup okay
1:32:27 so let's cop so so now what I'm doing so
1:32:29 we've got we've got these images from
1:32:32 from chat GPT from Dolly
1:32:35 3 so now what I'm going to do is let's
1:32:38 go over let's grab the the
1:32:42 prompts and let's go over we'll go to
1:32:45 Mid
1:32:52 journey and I'm going to make these 16
1:32:55 by 9 they already
1:32:57 are and I think I'm going to put some
1:32:59 stylization and some weirdness in here
1:33:02 and I'm going to do there's my start
1:33:04 image for the magician thing am I going
1:33:07 to add anything else to this I'm not
1:33:09 right
1:33:12 now and then we're going to go back to
1:33:14 chat
1:33:19 J and we're going to grab the end image
1:33:24 and go back to Mid Journey oh that's
1:33:45 cool and Murphy's in the house was
1:33:48 shaking I Murphy hey how you doing I
1:33:51 Murphy yeah fantastic Bob tell them what
1:33:54 won he's not W anything it's not a game
1:33:57 show show
1:33:59 whatever tellan Murphy she's welcome
1:34:01 you're welcome man
1:34:03 Murphy hey everybody what are we up to
1:34:06 we
1:34:08 are
1:34:13 um the request was show Runway
1:34:17 ml so what I'm showing
1:34:20 is chat GPT for ideation
1:34:24 end image generation then we're going to
1:34:26 generate start and end images
1:34:30 in mid Journey chat GPT and Kaa using
1:34:35 flux so we're going to have three
1:34:37 different places to start and then we're
1:34:40 going to use three different video
1:34:41 generation models cling Runway ML
1:34:46 and I don't know what's the other one
1:34:49 Dream Machine by by Luma
1:34:51 Labs 32,000 tokens
1:34:55 oh great in the memory context window
1:34:57 now that's actually really great okay
1:34:59 good okay so there's start and end so
1:35:02 let's go
1:35:04 grab start and end for the burning
1:35:10 letter and actually I got to move all
1:35:12 these I've got all my tabs all over the
1:35:14 [ __ ] place it's driving me [ __ ]
1:35:17 Bonkers so we're going to put that right
1:35:19 next to there we're going to move Kaa
1:35:22 right next to it
1:35:26 uh all right so we got that we got mid
1:35:31 Journey okay so now this is the start
1:35:33 image for the second
1:35:36 concept and then we're going to go grab
1:35:38 chat jpt we're going to grab the end
1:35:40 prompt for the second
1:35:45 concept is this making sense what I'm
1:35:47 doing here
1:35:51 people and now we're going to go to the
1:35:53 third concept cep in chat
1:35:55 GPT and grab the prompt written by chat
1:35:59 GPT again I'm not using my brain here
1:36:03 I'm
1:36:04 just a conduit to
1:36:07 let these tools
1:36:10 be reveal their genius to
1:36:16 us now if you're like I hate this
1:36:19 process I want to be much more anal
1:36:21 retentive and I want to control
1:36:23 everything then you do you like
1:36:25 absolutely do that this is the kind of
1:36:28 Joy about these tools is I think the the
1:36:31 more anal retentive and controlling you
1:36:33 are the more frustrated you're going to
1:36:35 be with with generative AI right now the
1:36:39 state of generative AI is these tools
1:36:41 are janky as
1:36:43 [ __ ] the process that I'm showing you
1:36:46 lets them be janky it says [ __ ] it you
1:36:49 know the the J Leno Doritos commercial
1:36:51 eat all you want we'll make more that's
1:36:53 what generative AI is right now let it
1:36:55 generate a ship pile of content and then
1:36:58 just pick the good [ __ ] that's sort of
1:37:00 what I'm showing you right now you can
1:37:02 take a much more intentional approach
1:37:03 than this um it's going to take you
1:37:06 longer to get it dialed in once you get
1:37:08 it dialed in you'll you'll be really
1:37:10 [ __ ] masterful at it but okay so this
1:37:14 these are the start images these are
1:37:16 what they're going to trans transition
1:37:17 into that's super cool um so we've got
1:37:21 we've got those start and end images now
1:37:23 and let's go do the same thing in
1:37:26 Korea now do I
1:37:29 have
1:37:33 home do I
1:37:37 have hang on a
1:37:40 library does Kaa allow me
1:37:43 to generate
1:37:48 images yeah it does okay so here's my
1:37:50 history okay great all right so Kaa so
1:37:53 now let's go do the same thing we're
1:37:54 going to go
1:37:57 grab the start prompt and the end prompt
1:38:00 in Korea so now we're going to have
1:38:03 three different sets of
1:38:06 images Korea mid journey and Dolly
1:38:12 3 all right look how fast Kaa is that's
1:38:15 amazing oh except it gave me the wrong
1:38:18 aspect
1:38:19 ratio so now we're going to generate new
1:38:23 aspect ratio then we're going to go over
1:38:24 here we're going to copy second
1:38:28 prompt oops second
1:38:38 prompt
1:38:40 Crea paste
1:38:45 boom
1:38:47 Z oh I'm actually liking Korea this is
1:38:51 flux the flux model
1:38:56 those are pretty
1:38:57 slick all right I like it go back here
1:39:02 second closeup of the
1:39:06 letter
1:39:09 Korea paste
1:39:12 boom go back
1:39:16 here right here got
1:39:20 eight go back here did I let these
1:39:24 things emerge yes I did
1:39:27 okay are these done those are
1:39:30 done haste
1:39:33 boom go back over
1:39:37 here so we haven't started doing any
1:39:40 Discerning
1:39:44 yet let's let these things finish
1:39:47 rendering
1:39:52 [Laughter]
1:39:58 birds today feathers
1:40:00 tomorrow I'm looking for a man in fin
1:40:02 nights trust fun 65
1:40:21 Blas all right let's look those things
1:40:24 finish
1:40:25 rendering and paste the new prompt in
1:40:30 there do that
1:40:33 okay so let's go look I'm going to I'm
1:40:36 going to make an executive call I
1:40:39 think that the chat GPT images suck I
1:40:43 kind of like that
1:40:45 demon but that's kind of a
1:40:48 shitty I just feel like that's going to
1:40:50 be a bad
1:40:52 transition so so we're not going to go
1:40:55 with any of those let's go look at Mid
1:40:58 Journey so mid Journey we've
1:41:01 got the magician's too far
1:41:11 away
1:41:14 however this image is a start image with
1:41:17 the magician down
1:41:18 here transitioning into this this image
1:41:24 with the window and the feathers is kind
1:41:26 of
1:41:28 cool so let me do I'm going to do five
1:41:31 variations on this
1:41:33 one maybe it'll give us or four
1:41:36 variations on this one maybe it'll give
1:41:37 us one with a
1:41:39 window and then this one I'm just going
1:41:40 to upscale upscale
1:41:45 creative
1:41:47 okay all right so that's doing its thing
1:41:49 and then let's see
1:41:52 Shadow if we
1:42:01 start I don't think those are any
1:42:04 good I don't think they're going to
1:42:07 work and then what about the
1:42:09 letter there's your start
1:42:14 letter no those are all crappy I don't
1:42:16 like those okay let's go to
1:42:19 Kaa all right that's kind of interesting
1:42:22 yeah this will be good okay so these
1:42:24 because they're
1:42:25 flat so we'll start
1:42:29 with one of these we'll end with one of
1:42:32 these although this one I don't think I
1:42:35 let these finish
1:42:36 rendering that was my bad that was my
1:42:39 bad Bubba let's go make some more of
1:42:45 these
1:42:48 generate all right so that's doing its
1:42:50 thing so those are pretty good
1:42:54 all right these I think are actually
1:42:56 really good you have like this one here
1:42:58 is a start
1:43:05 image
1:43:07 actually this one is a start image okay
1:43:10 so let's go let's go I'm going to copy
1:43:12 this copy
1:43:14 image and we're going to go
1:43:18 to oh what the [ __ ] is it dream machine
1:43:21 so Luma Labs Luma Luma Labs I hope I
1:43:25 hope mods are are following along and
1:43:27 helping people I know I'm just rambling
1:43:29 right
1:43:30 now try now so we're at Luma labs.
1:43:35 a and now if I paste that no it doesn't
1:43:38 let me okay so I got to go back to
1:43:41 Korea so I'm going to save this image
1:43:45 as
1:43:47 env1 envelope one um this is file
1:43:51 management best practice
1:43:54 um save all of your files to your
1:43:57 desktop with nondescript names so that
1:44:00 you can never find anything ever
1:44:02 again you're welcome
1:44:07 env2 I know I did an image of an
1:44:09 envelope it was kind of a month ago
1:44:12 [Laughter]
1:44:14 is okay so now we're going to go back
1:44:16 here so what you do at Luma laabs is
1:44:20 there's this little image icon so I'm
1:44:22 going to click on that
1:44:24 and I'm going to upload image one this
1:44:29 one and then it gives you a little plus
1:44:34 thing so this is the start frame and the
1:44:37 end
1:44:38 frame and so now I'm going to go grab
1:44:40 inframe
1:44:44 two there it is and then we're going to
1:44:47 say
1:44:50 um envelope
1:44:56 magically
1:44:58 opens
1:45:00 and
1:45:03 bursts into
1:45:08 flams and we're going to let it enhance
1:45:11 the prompt and we're going to start
1:45:14 it all right and then we're going to go
1:45:17 to Runway
1:45:20 ml I think I have credits on Runway rway
1:45:47 ml you're already subscribed okay
1:45:51 fine so we're going to go to AI
1:45:55 video text to image
1:46:06 V2 let's
1:46:08 see AI
1:46:13 video oh this is I'm not on that weird
1:46:18 oh I see what I did Runway ml okay I
1:46:21 went to I typed in Runway ml and then it
1:46:24 took me to a whole another site okay gen
1:46:25 3 Alpha sorry I was at the wrong site
1:46:28 there so we're going to try run
1:46:30 Runway we're going to log in with
1:46:32 Google the goo Google single sign
1:46:38 on get started and now I think this is
1:46:42 just image
1:46:48 to can we
1:46:51 do start and end frames here I don't
1:46:54 know if we
1:47:05 can ah yeah so there's our start frame
1:47:09 here's our end
1:47:12 frame can I change it how do I change
1:47:15 it
1:47:18 trash select from assets no that's not
1:47:21 what I want back
1:47:23 uhoh
1:47:25 [ __ ]
1:47:27 plus Web
1:47:34 Two
1:47:39 oops
1:47:42 C
1:47:51 first um
1:47:56 remove image select from
1:48:00 assets
1:48:04 one I don't I I don't know how to use
1:48:11 this does anybody know how to use
1:48:16 this this is this is the this is the
1:48:19 bravery and
1:48:21 adaptability attribute of being an agile
1:48:24 human all right um what I should have in
1:48:28 here is humbleness um let's see
1:48:32 uh
1:48:34 [Laughter]
1:48:39 humility here's the thing about these
1:48:41 tools if you don't use them every day
1:48:43 every time you go back to use them
1:48:44 they've changed the interface so this
1:48:47 has got a first frame and a last frame
1:48:49 but I don't know how to change the last
1:48:51 frame and keep the first one anybody
1:48:53 does anybody know how to do
1:48:59 this oh this is [ __ ]
1:49:02 stupid use as first
1:49:08 frame use as first
1:49:12 frame all right
1:49:15 crop
1:49:18 no back
1:49:24 oh this
1:49:27 sucks I I honestly don't know how to do
1:49:32 this anybody I have faith in you what's
1:49:35 he on I'm on Runway ml right now okay
1:49:39 when in doubt bail and go back to the
1:49:41 last tool that you used oh wait I didn't
1:49:43 I didn't hit submit for this I'm such a
1:49:46 dumb
1:49:47 dumb or do I not
1:49:51 have account
1:49:55 your plan usage you have 65 out of 70
1:49:59 Generations okay
1:50:02 create oh crap
1:50:07 where shut
1:50:10 up I know I'm a dumb dumb all right uh
1:50:14 what did I say
1:50:16 um envelope magically opens itself and
1:50:21 bursts into flames
1:50:28 submit it's not working why is it not
1:50:31 working oh come
1:50:42 on all right let me copy this copy let
1:50:45 me reload the page
1:51:01 oh you know what this could
1:51:07 be this could
1:51:09 be [ __ ]
1:51:12 webp okay so you're like web P what what
1:51:16 the hell are you talking about okay webp
1:51:19 is a format an image format that
1:51:23 some brilliant [ __ ] engineer said you
1:51:27 know what jpeg is just not efficient
1:51:30 enough let's create a new format that no
1:51:33 one will [ __ ]
1:51:35 support okay so we're going to export
1:51:38 selected images
1:51:41 as a
1:51:43 jeeg let's see let's go to desktop
1:51:47 choose export um did I export them as
1:51:51 JPEG yes okay so let's go back
1:51:55 to
1:51:57 here and try this with our jpeg images
1:52:00 now there's number
1:52:04 one there's number
1:52:10 two let's paste in our prompt and hit
1:52:15 submit and I got
1:52:18 nothing what's happening
1:52:29 no prompt hit submit nothing
1:52:34 happens did it do these lower
1:52:38 no oh this is [ __ ] maddening
1:52:44 people embrace the Jank here's a
1:52:48 lesson this should work it's not
1:52:53 this should work and it's not let's see
1:52:59 use as first
1:53:04 image
1:53:06 okay and then I'm going to
1:53:10 go remove the image select from
1:53:17 assets there's that use as last image
1:53:25 I I honest to God don't know let's we'll
1:53:28 put in a different prompt
1:53:31 there we're going to generate that an
1:53:33 error occurred your credits will be
1:53:40 refurbished ain't AI
1:53:48 awesome um let's try cling cling cling
1:53:52 cling
1:53:53 cling
1:53:54 video cling
1:53:58 video I don't even know if cing does
1:54:00 [ __ ]
1:54:04 uh complete the security thing you got
1:54:07 to be kidding
1:54:13 me I don't know if clling does image so
1:54:16 AI
1:54:19 video
1:54:21 prompt 16 by 9 that's good image to
1:54:26 video okay
1:54:29 good we're going to grab envelope
1:54:35 one prompt can I do a second image add
1:54:40 end frame yes oh sorry so I I put one
1:54:44 image in here now I'm going to say add
1:54:45 end
1:54:47 frame upload number two
1:54:53 oh this is cool this is cool people okay
1:54:57 now I'm going to put in my
1:54:59 prompt envelope magically opens itself
1:55:02 and bursts into
1:55:04 frame creativity relevance standard mode
1:55:08 professional mode upgrade to premium
1:55:10 generate 10 credits and I've got 56
1:55:13 credits boom I think we're moving
1:55:16 people we've got one one of three one of
1:55:20 three of these tools is actually doing
1:55:22 something
1:55:25 something um I have a feeling that this
1:55:30 one Luma
1:55:31 Labs I have a feeling that my browser
1:55:34 has [ __ ] to bed cuz i' see all my tabs
1:55:38 all of you in Judgment of me because I'm
1:55:41 a tab
1:55:42 hoarder [ __ ] you I I enjoy my tabs yes
1:55:47 that's why my browser crashes and yes
1:55:49 that's why my life is a living hell but
1:55:55 but I have a feeling if I quit Chrome
1:55:57 and restart it I could get this thing to
1:56:00 work and Runway ml I this is this is
1:56:04 just bad
1:56:06 interface that I can't figure out how to
1:56:08 get it to do a start and end
1:56:15 frame and I don't know why it errored
1:56:19 out although it did say it's going to
1:56:21 give me my credits back
1:56:24 that's just [ __ ] stupid all right
1:56:26 where's cling cling
1:56:30 cing is this cling cling cling no this
1:56:34 is did I close cling what the [ __ ] did I
1:56:38 do
1:56:45 cling all right so it's generating it's
1:56:47 generating the video
1:56:55 while it's doing
1:57:01 that cling
1:57:04 cling
1:57:08 cling I think it's important for you all
1:57:10 to understand the
1:57:14 origin of that video generation called
1:57:17 cling cling cing you know what I'm
1:57:18 saying bust
1:57:21 it man you
1:57:23 o clean C clean C clean oh you capping
1:57:26 drip splash water drip splash splash
1:57:29 drip water
1:57:35 splashing you do not do that young
1:57:43 man all
1:57:46 right cling cling cling
1:57:53 [Laughter]
1:57:56 oh good lord
1:57:58 okay what have we learned here tonight
1:58:04 um I don't know probably not much other
1:58:08 than be willing to be a dumb
1:58:11 dumb let yourself feel stupid because
1:58:14 what what will happen we'll see if this
1:58:16 video turns out any good or not
1:58:19 um what will happen is it'll be
1:58:23 frustrating frustrating frustrating and
1:58:24 then all of a sudden you'll get it and
1:58:26 you'll figure out a workflow that works
1:58:27 for you and you'll figure out okay I
1:58:28 like this image generation tool I don't
1:58:30 like that one okay I can idate with chat
1:58:32 GPT or I can't or right a lot of what
1:58:36 I'm doing
1:58:39 is feeling around in the dark for the
1:58:42 edges of what AI makes possible AI
1:58:46 learning lab is like a box of chocolates
1:58:48 yeah tonight the box of chocolates were
1:58:50 all melted and they were all they were
1:58:52 all like toothpaste centers it was bad
1:58:56 we've learned bows are pretty exactly
1:58:59 while we wait for this to generate check
1:59:01 the mods please okay let me go check
1:59:07 mods cling cling
1:59:09 cling we be
1:59:13 capping I'm having some problem on dream
1:59:15 machine it won't even let me prompt I
1:59:18 even restarted okay oh great okay so it
1:59:20 looks like dream machine is just [ __ ]
1:59:22 up right now so that's the other thing
1:59:24 is these tools are just sometimes broken
1:59:26 and it looks like cling here is going
1:59:28 pretty [ __ ] slow Runway ml I don't
1:59:31 know like I I obviously don't know how
1:59:34 to use that interface and that shitty
1:59:40 interface like the the Luma the dream
1:59:42 machine one is really good interface it
1:59:44 just happened to not work tonight this
1:59:46 cling cing is pretty good by the way if
1:59:48 you if you haven't tried cling it's cing
1:59:51 ai.com k l i n g
1:59:55 ai.com this is the
1:59:59 Chinese sort of version of
2:00:01 Sora or that's what everyone says it's
2:00:04 as good as
2:00:05 Sora Sora is not out yet so we don't
2:00:07 know how Sora
2:00:11 is um all right we're up to 87% we'll
2:00:15 we'll get here soon we'll have a
2:00:19 video but what'll what'll end up
2:00:21 happening is is if you start out with
2:00:24 the sort of a like an add version of
2:00:28 production like I did
2:00:30 tonight you'll stumble upon things that
2:00:33 you like you'll be like oh I really like
2:00:35 the way Dolly made images or right I you
2:00:37 know like the way mid Journey did it the
2:00:39 way or the way Kaa did it there's 15
2:00:42 other image generation tools you can
2:00:43 play with it it literally doesn't matter
2:00:45 at this point they're all pretty
2:00:47 good what matters is what's what works
2:00:51 for you
2:00:53 what allows you to seamlessly
2:00:57 navigate these different kind of modes
2:01:00 of being a worker in
2:01:03 AI how can you move quickly from
2:01:06 Curiosity to oh let me go try something
2:01:09 to oh now I've got too many choices let
2:01:11 me refine those now let me let me dig in
2:01:15 and and try to make something
2:01:17 intentional out of this so so navigating
2:01:20 quickly between all these different
2:01:23 modes is what will give you this idea of
2:01:26 being an agile
2:01:27 human and and this is [ __ ] you can
2:01:29 practice right this is [ __ ] you can
2:01:31 practice by just doing it this video
2:01:35 better not
2:01:36 [Laughter]
2:01:40 suck I read an article that said chat
2:01:43 GPT 5 might just wait might just be
2:01:47 better at math I was like huh well we'll
2:01:51 see if it's got reasoning if Q if it's
2:01:53 got qar reasoning it will be better at
2:01:57 math um if it doesn't then it might be
2:02:00 better at math but it won't be great at
2:02:02 math we'll
2:02:06 see we will see can you check out Blaze
2:02:09 AI I think I looked at that the other
2:02:12 day didn't I Blaze
2:02:16 doai there's so many [ __ ] tools now
2:02:20 Blaze smart automation automate your
2:02:23 business processes with blaze
2:02:26 AI so is this
2:02:30 like create better content in half the
2:02:48 time this looks like a
2:02:51 lot
2:02:53 product brand voice train Blaze to sound
2:02:56 just like you all right so this is
2:02:58 making this is like automated content
2:03:01 creation
2:03:03 stuff so here's the thing about tools
2:03:05 like this there are going to be an
2:03:08 infinite number of these if you want to
2:03:10 discover a bunch of tools like this go
2:03:12 to futurepedia doio is that is that
2:03:20 it future
2:03:26 pedia
2:03:31 doio find the best AI tools and software
2:03:34 yeah so this has got how many tools are
2:03:36 they up to now it's got to be a [ __ ]
2:03:45 ton AI
2:03:49 tools they used to have an A total
2:03:52 number anyway this has got just all
2:03:54 sorts of tools within it image
2:03:57 generators 200 tools right design
2:04:00 generators 246 so there's just tons and
2:04:03 tons and tons of tools out there the
2:04:05 stuff that I'm trying to focus on are
2:04:06 kind of the the core the the the lower
2:04:10 level tools that a lot of these other
2:04:11 tools are built from um so that that
2:04:15 Blaze AI is they're what they're
2:04:18 essentially doing there is they're pre-
2:04:19 prompting they're they're stitching
2:04:21 together number of different models a
2:04:23 video generation model and they're
2:04:25 probably using 11 labs for audio
2:04:26 synthesis they're probably using hyen or
2:04:29 Synthesia for video synthesis um they're
2:04:33 using other things for design overlay so
2:04:36 what those sites are is they're pre-
2:04:38 prompting and pre-building a bunch of
2:04:40 stuff that's stitching together a bunch
2:04:42 of tools what I'm trying to teach here
2:04:44 is what are the core tools that make
2:04:46 those up and then you can look at a tool
2:04:48 like that and analyze it and say oh it's
2:04:51 cheaper for me to just buy the [ __ ]
2:04:52 they've done cuz it it does exactly what
2:04:55 I want it to do but that doesn't
2:05:00 mean like those tools are so specific
2:05:03 and that one in particular has a really
2:05:05 particular vibe to it some people are
2:05:07 going to love that some people are going
2:05:08 to hate that you can also kind of roll
2:05:11 your own if you learn the core
2:05:14 underlying uh engines J Hamilton how
2:05:17 does one become a mod you raise your
2:05:18 hand and say you want to be a mod and
2:05:20 then I'll introduce you to Vicky um
2:05:24 who's the community manager over on the
2:05:26 AI salon and then um she can walk you
2:05:29 through what it what it's about and then
2:05:32 if you're into it um and and you know
2:05:36 want to contribute awesome and then I'll
2:05:39 I'll basically just set you up on here
2:05:40 and you'll you'll have access to the mod
2:05:42 Channel there so um let if you're on the
2:05:46 if you're on the salon you can just ping
2:05:48 Vicki directly Vicky
2:05:50 Baptist uh what can Aid do today.com is
2:05:55 another one yeah that's
2:05:57 right um okay let's go back to clling
2:06:00 cing cing 97% we're almost there people
2:06:03 it's going to be so exciting it's going
2:06:05 to be so
2:06:06 good my voice is going I'm getting a
2:06:09 little raspy okay listen tomorrow at
2:06:12 11:00 a.m. mountain time I do AI office
2:06:15 hours on
2:06:17 LinkedIn um so it's a it's kind of a
2:06:20 it's what it's become is kind of a
2:06:21 really interesting Mastermind the same
2:06:23 group of people for the most part show
2:06:26 up every Friday at 11:00 a.m. so if you
2:06:29 go to my channel on Tik Tok or on on
2:06:33 LinkedIn Kyle Shannon on
2:06:35 LinkedIn um tomorrow morning I'll put up
2:06:37 a reminder for AI office hours you can
2:06:40 see my events but it's it's every if you
2:06:43 look at one of the past events the URL
2:06:45 for it is the same it's been the same
2:06:47 for like a year um and we just hang out
2:06:49 for an hour and talk about this AI stuff
2:06:51 so
2:06:52 it's it's really an interesting group
2:06:54 mod Channel thread for reference oh
2:06:56 great thank you Kyle the thread
2:07:04 please office hours tomorrow images and
2:07:06 Irregulars okay great thank you I like
2:07:09 this uh thread we're at 98%
2:07:12 there I like this mod thread you you
2:07:15 mods I'm digging it so I'm going to the
2:07:18 regulars
2:07:20 Channel by
2:07:25 a diagram of Kyle's Kyle's brains
2:07:28 workflow during Tik Tok this is totally
2:07:32 what tonight was okay but here's the
2:07:35 thing right
2:07:38 now because so many tools are so capable
2:07:42 I don't know that there's a better way
2:07:44 well there's a different way a different
2:07:46 way to get competent with these AI tools
2:07:49 is pick one
2:07:50 tool and get good at it and then pick
2:07:54 another tool and get good at it and then
2:07:56 pick another tool and get good at it
2:07:59 where I am right now is there's so many
2:08:01 tools to test out you literally can't
2:08:03 keep up so I'm right now at the point of
2:08:05 like just start getting good at being
2:08:08 Nimble and [ __ ]
2:08:10 throwing [ __ ] at the wall right and then
2:08:15 if something makes art great but yeah I
2:08:17 think that is an
2:08:20 accurate if you're brain is like I sorry
2:08:24 I came to the AI learning lab dude is
2:08:26 there an agenda for the evening this is
2:08:29 not going to be the channel for
2:08:31 you and and if if seeing your professor
2:08:36 look like this is disturbing to you in
2:08:39 any way this is not the channel for
2:08:45 you going to start an AI boot camp think
2:08:49 we could talk about that in a office
2:08:52 hours yeah
2:08:53 absolutely hop in there hop in there
2:08:56 tomorrow and that that's the whole point
2:08:57 the whole point is sort of I create I
2:09:00 create this space there and while it's
2:09:02 similar to this where there's a lot
2:09:04 going on in the comments down here
2:09:05 what's nice about office hours is you
2:09:08 know you can ask me questions but
2:09:10 basically you can just say here's what's
2:09:12 going on and then a lot of people in the
2:09:14 group will have insights about it so
2:09:16 it's become a bit of a mastermind it's
2:09:17 it's really quite a neat a neat group
2:09:20 all right let's look at cing cling cing
2:09:22 oh come on 99% really cling cling
2:09:26 cling we'd be
2:09:30 capping could you please make the video
2:09:32 now Oprah is interviewing Bill Gates and
2:09:35 Sam Alman on ABC and Prime Time on 912
2:09:38 yeah maybe
2:09:40 9912 I'll be back so on on the 10th of
2:09:44 September I fly to Washington DC on the
2:09:46 11th of September I'm testifying or
2:09:50 whatever I'm I'm speaking to a bunch of
2:09:52 senators and [ __ ] about Ai and I'll be
2:09:55 back on the 12th so
2:09:59 maybe it'll freeze at 99% yeah of course
2:10:03 um
2:10:06 maybe September 12th we'll we'll watch
2:10:09 the Oprah thing that would be kind of
2:10:11 fun we could watch it in here and talk
2:10:12 about
2:10:15 it jillo let's have dinner live in DC if
2:10:18 you're in DC yeah ping me DM me here or
2:10:21 DM me on the salon JSO happy to happy to
2:10:25 set something up if I've got time I
2:10:27 don't know if I do um but maybe when I
2:10:30 get in on
2:10:32 um on the 10th I think
2:10:35 I I think I might have an hour before I
2:10:39 have to go to my dinner or whatever
2:10:42 so all right how many likes are we up to
2:10:45 we're up to 30,000 you guys are [ __ ]
2:10:47 amazing cling cling cling cling clang
2:10:49 cling with the tral
2:10:51 okay I'm this is bugging me now let me
2:10:54 re reload the page cuz maybe it just
2:10:57 hung and it's actually
2:11:00 done sign in for cling cling
2:11:04 cling I like I can't leave here and not
2:11:06 show you a
2:11:08 video we went through all that come up
2:11:11 with the concepts and do the start image
2:11:13 and the end
2:11:16 image AI
2:11:18 videos still generating [ __ ]
2:11:28 99% this is what like when the old
2:11:30 Microsoft os's were at 99% installed on
2:11:33 their last minute exactly it happens all
2:11:36 the time it's part of the ride yeah I
2:11:39 know it's just a shitty part of the
2:11:41 ride I'm
2:11:44 hung give Kyle some Taps see if we can
2:11:47 get to
2:11:48 35,000 we
2:11:51 [Laughter]
2:11:53 anyway all right I'll come I'll come
2:11:54 back to that um holy [ __ ] what happened
2:11:57 to the hair I think the only thing you
2:11:59 can do if your hair is that bad you got
2:12:02 to distract people all
2:12:07 right to be continued maybe we'll see um
2:12:12 but let me let me wind some things down
2:12:14 here because it's getting late
2:12:17 um ai ai salon
2:12:22 I mean the
2:12:23 salon we're changing the name of the
2:12:26 salon or AI the AI Salon to the
2:12:31 salon so go to the salon. if you haven't
2:12:35 done it already tonight go there click
2:12:37 on join the
2:12:38 community and go sign up introduce
2:12:41 yourself follow the sort of onboarding
2:12:44 process join the
2:12:46 community why cuz a guy in a pink bow
2:12:49 told you to
2:12:51 it doesn't cost you anything and it will
2:12:54 be the future of your employment in my
2:12:57 opinion that group or groups like
2:13:04 it all right maybe open another window
2:13:07 of cling and see if it's ready in the
2:13:10 new window yeah I reloaded the page and
2:13:12 it it just it
2:13:20 didn't I can show you other
2:13:29 videos dragging with
2:13:34 birds this is cling cling cling video
2:13:37 here's cling cling cling
2:13:40 video this one was actually pretty
2:13:49 good I mean it's quite remark how how
2:13:52 much it retains the shapes and it
2:13:55 doesn't morph and [ __ ] everything
2:14:02 up these are some masks that one was
2:14:05 kind of
2:14:12 shitty yeah these were these didn't do
2:14:14 very good on these masks
2:14:22 [Laughter]
2:14:25 looks like he's got
2:14:31 TMJ this one was pretty
2:14:40 cool yeah it's still gone all right
2:14:43 whatever [ __ ]
2:14:45 it
2:14:48 um it'll have to be ready tomorrow okay
2:14:51 so
2:14:52 tomorrow we'll be here tomorrow's not
2:14:54 first Friday right
2:14:57 no so we'll do Friday night date night
2:14:59 tomorrow come back
2:15:03 tomorrow
2:15:04 and we'll do Friday night date
2:15:07 night and you're like what's Friday
2:15:09 night date night I don't know bring a
2:15:12 date and hang out in the AI learning lab
2:15:15 or if you don't have a date I'll be your
2:15:17 date just come hang out it'll be Friday
2:15:19 night date
2:15:20 night um and we'll do the same thing
2:15:22 maybe tomorrow I'll try to do something
2:15:24 a bit more focused this was like I on
2:15:27 purpose wanted it to be wide and messy
2:15:30 it was a little wider and a little
2:15:32 messier than I wanted it to be but
2:15:33 that's that's kind of the deal right now
2:15:37 there's even if you just pick a single
2:15:40 tool they keep changing them so much
2:15:42 they keep adding features and as the
2:15:44 models underneath them get better they
2:15:47 add things to it interface changes there
2:15:49 there's just like a lot to keep up with
2:15:52 so don't try to keep up with it just
2:15:54 accept that some nights are going to be
2:15:56 janky some nights are going to be
2:15:58 [ __ ] brilliant you're going to be
2:15:59 like holy [ __ ] it's going to change the
2:16:01 world and then two nights later you'll
2:16:03 be like this is the biggest piece of
2:16:04 [ __ ] I've ever seen good night everyone
2:16:06 have fun with the AI of your choice
2:16:09 exactly that's perfect super awesome
2:16:12 prompts um good
2:16:15 night um I'm 60 not scared by much but
2:16:19 if you knocked on my door right now at
2:16:21 the a pink bow I would be scared good
2:16:24 that's good this is I'm I'm here to
2:16:26 remind you that the world is not a safe
2:16:28 place that sometimes a a guy like me
2:16:33 shows up in a pink bow and nothing good
2:16:35 can come of that nothing good can come
2:16:37 of
2:16:38 that that's good I'm I'm glad we've got
2:16:41 some some appropriate anxiety in the
2:16:43 world that's good um follow my channel
2:16:47 if you would if you want to support the
2:16:48 channel directly you can subscribe to it
2:16:50 there's also in one of the corners here
2:16:51 there's a series that you can buy called
2:16:54 underhyped which talks about why thank
2:16:56 you very much Winston it's very generous
2:16:58 I appreciate that um this down here
2:17:01 talks about why I think AI is actually
2:17:04 underhyped not overhyped and how it
2:17:07 works and all sorts of [ __ ] so you can
2:17:10 do that um the other thing to do if
2:17:13 you're new here and and you're just like
2:17:15 I don't know why I'm still watching this
2:17:17 but I can't stop watching this is like a
2:17:19 [ __ ] train wreck you can you can
2:17:22 experience this train wreck five nights
2:17:24 a week for
2:17:27 free where
2:17:31 else but Tik Tock could you watch an old
2:17:34 guy stumbling through using technology
2:17:37 with semi humorous quips
2:17:40 occasionally right here um the whole
2:17:44 idea of this channel is to show up not
2:17:47 for what I do but for what you do while
2:17:50 I'm doing it
2:17:52 be in the conversation be learning about
2:17:54 AI be trying to do what I'm doing like
2:17:56 the fact that forever hook figured out
2:17:59 that um that uh dream dream machine
2:18:02 video is [ __ ] up for him it's not just
2:18:04 my browser like right you you kind of
2:18:07 learn stuff along the way be in this
2:18:09 conversation um you can subscribe to
2:18:11 this train wreck yeah if you subscribe
2:18:13 to these lives um that directly supports
2:18:15 me is appreciated um you know I don't
2:18:19 make much here I think if I average it
2:18:20 out at probably about 75 cents an
2:18:24 hour I put a lot of hours into this but
2:18:27 it isn't about the money but any any any
2:18:29 of that sort of support helps all right
2:18:31 I'm going to the woods of East Texas
2:18:33 tomorrow to ruminate on it all um I've
2:18:36 learned these past few months yeah it's
2:18:41 we're here's the thing if if you're
2:18:43 feeling like you're behind you're not if
2:18:46 you're feeling like you're ahead you're
2:18:48 not if you're feeling like you've got it
2:18:50 figured out you don't if you're feeling
2:18:54 like
2:18:56 me kind of in the middle of an
2:19:00 ocean just bobbing in the waves like not
2:19:03 there's no land in sight there's no wind
2:19:08 blowing if you're feeling like that like
2:19:11 just in this weird like I don't know
2:19:13 [ __ ] where I am or what's going on I
2:19:16 think that's where this industry is
2:19:18 right now it's it's it's a bit lost and
2:19:22 without Direction and that's okay
2:19:26 cuz within the next 60 days is my
2:19:30 guess um open AI is going to drop
2:19:33 something that's going to really change
2:19:35 things like really change
2:19:37 things and and this whole Cycle's going
2:19:40 to start over again we're going to get
2:19:42 all excited about whatever they drop
2:19:44 we're going to try to get our head
2:19:45 around it it's going to take us months
2:19:48 to get our head around it and then
2:19:50 everyone else is is going to try to
2:19:51 catch up to them I I think we're going
2:19:53 to we're going to have this
2:19:56 cycle repeat itself over the next 3 to 5
2:19:58 years so just it's okay if you're in a
2:20:02 in a Mala just keep coming back here and
2:20:05 and we can all be in this
2:20:08 together
2:20:09 um all right so let's do let me go see
2:20:13 if did my video No it's it's hung at 99
2:20:17 it'll it'll we'll we'll show it
2:20:19 tomorrow um
2:20:22 let's do our Romper Room magic mirror
2:20:23 joy py in the house jillo question most
2:20:26 jobs are asking for PhD and
2:20:30 masters of their room for
2:20:33 technicians yeah that that's that's
2:20:35 again that's this idea of if you want to
2:20:38 get into the infrastructure the
2:20:40 technical side of AI they're going to
2:20:43 ask for Masters if you're trying to get
2:20:46 into the how do I use chat GPT for my
2:20:49 product management
2:20:52 position um and how do I make things
2:20:54 more efficient or exciting or reinvent
2:20:57 things
2:20:58 entirely there you there's no PhD to get
2:21:01 like you can get a PhD in building
2:21:04 machine Learning Systems you can't get a
2:21:06 PhD in generative AI right now it's all
2:21:09 brand new so this is here here's the
2:21:12 equivalent if we were in
2:21:15 1995 you could get a degree in network
2:21:19 infrastructure internet Network design
2:21:22 nodes and servers and
2:21:26 [ __ ] you couldn't get a you couldn't
2:21:29 even get a a bachelor's degree in web
2:21:32 design because everyone was figuring it
2:21:35 out we we were trying to figure it out
2:21:36 in real
2:21:38 time by the time you get to like 2002
2:21:41 you could now get certificates in web
2:21:43 design by the time you get to 2010 now
2:21:46 you've got degrees in it
2:21:49 so chat GPT generative AI does for
2:21:53 machine learning and AI what the
2:21:55 worldwide web did for the internet so
2:21:58 it's the exact same metaphor so if you
2:22:01 want to get a job in the infrastructure
2:22:04 of machine learning yeah you're probably
2:22:06 going to need a master's or PhD or
2:22:09 equivalent to do that if you want to use
2:22:12 chat GPT in business and make money
2:22:14 there they ain't no degree in that
2:22:17 that's
2:22:18 this this chaotic [ __ ] show of a channel
2:22:23 is where that industry is right now it
2:22:26 just is people are going to tell you I
2:22:29 can you know I've got a course you can
2:22:31 make
2:22:31 money they're all [ __ ] because
2:22:34 everything's changing so fast and I'll
2:22:36 tell you what when open AI drops a
2:22:39 reasoning engine and a significantly
2:22:41 more capable chat gbt all the rules
2:22:44 [ __ ]
2:22:46 change and yet I feel like I've been
2:22:48 going through a PhD program in
2:22:50 generative AI yeah like the amount that
2:22:52 we're learning is staggering the amount
2:22:54 that we're going to actually retain and
2:22:56 use over time is going to be very small
2:22:59 but what we're learning Source camp and
2:23:02 I mean you're really good at this what
2:23:04 we're learning is how to be in this
2:23:08 conversation as the technology gets
2:23:10 better this group is going to know what
2:23:12 to do with it better than most people in
2:23:14 the planet that's what this Channel's
2:23:17 about I still feel like you know way
2:23:19 more than me and I've been watching you
2:23:21 for forever
2:23:23 now I you know I'm I'm sort of
2:23:26 pathologically obsessed with this stuff
2:23:28 so um yeah it's just it's part of I
2:23:32 don't know whatever it's it's part of I
2:23:34 retain some things really well I retain
2:23:36 most things not at all jow remember when
2:23:39 instructions were Ground truth yeah ULU
2:23:43 RX just say the word and pink bows for
2:23:45 all there you go pink bows for
2:23:48 everyone um
2:23:51 oh oh when when instructions were
2:23:53 groundbreaking yeah exactly yeah it's
2:23:56 this is a we're a mess right now you're
2:23:58 my date every night Kyle thank you Jay
2:24:00 Hamilton I appreciate that um okay so
2:24:03 who's in the house tonight we got uh
2:24:05 kickoff J silow good to see you Auntie
2:24:07 bubbles Amelio's wife
2:24:10 kakee Joey Jay
2:24:14 Hamilton Brandon thank you good to see
2:24:17 you as always Daisy May fan fantastic
2:24:21 fantastic to have you in here Bob
2:24:23 Winston Hardy good to see you
2:24:26 sir Um Papa JJ Source Camp you get a bow
2:24:31 and you get a bow and you get a bow bows
2:24:33 for
2:24:38 everybody oh my God it's crazy it's
2:24:41 crazy I tell you it's
2:24:44 crazy uh Mel good to see you we are not
2:24:47 a
2:24:48 cult we are not a cult we are not cult
2:24:51 put on your
2:24:52 [Laughter]
2:24:56 bows oh man Joker's in the house we are
2:25:00 one I know a bit but not enough am I too
2:25:03 far gone no here's here's the
2:25:06 deal you're not as far ahead as you
2:25:09 think you might be and you're not as far
2:25:10 behind as you think be as you think you
2:25:13 are probably at this point 45% of
2:25:18 Americans oh t-rock appreciate your
2:25:21 wiener thank you 45% of Americans have
2:25:26 not even heard of chat
2:25:28 GPT if you've heard of
2:25:31 it you're farther ahead than many many
2:25:35 people in the world if you've used it
2:25:39 once you're like in the top
2:25:42 10% if you use it more than an hour a
2:25:45 week that's down to
2:25:49 2% if you're using chat GPT for like 2 3
2:25:52 4 hours a week or some AI for 2 three
2:25:56 hours a
2:25:57 week you're in the
2:26:00 like 5% half a percent of
2:26:05 people so you're way farther ahead than
2:26:07 you
2:26:09 think and you're not nearly as far ahead
2:26:13 as you think you are because it's so
2:26:16 vast there's so much of it so you can't
2:26:19 possibly keep up with it
2:26:22 so don't try because it'll just make you
2:26:24 [ __ ]
2:26:25 crazy it'll it'll really burn you out
2:26:27 and I mean this channel for for a lot a
2:26:30 lot of months we were trying to keep up
2:26:31 with everything and we were actually
2:26:33 doing really good at it and then all of
2:26:35 a sudden like the wav started coming of
2:26:37 just more tools got more better and more
2:26:40 complicated and there's more of them and
2:26:42 you literally can't keep up with it
2:26:43 anymore so that's why we're sort of
2:26:45 shifting into it's not so much about the
2:26:48 tools but it's more about how we're
2:26:50 being being who we're being all right
2:26:53 groovy all right I'm going to get out of
2:26:55 here we'll do one last check no still at
2:26:57 99% that ain't happening tonight um I
2:27:00 appreciate all of you I will see you
2:27:02 tomorrow Friday night date night um
2:27:05 11:00 a.m. Mountain Time on linkon is AI
2:27:07 office hours check that out
2:27:11 and the
2:27:13 salon. is the community that you need to
2:27:16 be a part of all right peace out thank
2:27:19 you all very much tonight I know it was
2:27:20 kind of chaos but life is like a box of
2:27:24 chocolates peace