aiLL 9/26/23 - The Implications of AI in Everyday Life

Description

Chapters

Transcript

0:14 foreign
0:20 happy Tuesday night
0:22 happy Tuesday night Luna stick what's
0:25 happening how you doing
0:28 Tobias long time no see
0:36 after a long night of saloning
0:39 you come to the lab
0:45 welcome everybody to the AI learning lab
0:47 happy Tuesday night let's see is this
0:50 shirt gonna be useless
0:52 damn I'm more I know right that was a
0:55 lot
0:57 that's a lot
0:59 I thought it'd be a good idea to put
1:01 that together and then I was like whoo
1:02 it's been a long day I know me too
1:05 me too long day I think I'm gonna go
1:06 change my shirt let people know I'll be
1:09 back uh tell them to share tap the
1:12 screen do that [ __ ] let me change it to
1:14 a black shirt you know
1:17 thank you
2:02 all right all right all right all right
2:05 kind of slipping into something more
2:08 appropriate my little house dress this
2:10 whole house dress I do declare hmm
2:15 very nice a lot of activity going on
2:17 outside
2:18 [Applause]
2:20 not to be distracted I do not want to be
2:23 distracted tonight if you know what I
2:25 mean
2:26 all right now he has to dress for us too
2:29 exactly it's exhausting exhausting
2:33 all these people people people
2:36 listen this is the AI learning lab my
2:39 name is akala Shannon
2:41 to that we're gonna learn about AI
2:44 it is artificial intelligence and you
2:48 may ask yourself a cow what are your
2:50 qualifications
2:52 and I must say to you but I have not
2:56 cheers
2:58 [Laughter]
3:04 um
3:06 uh I love my job if it wasn't for all
3:09 the people exactly
3:12 exactly what's the
3:17 the George Carlin routine
3:19 yeah I know I want to start the people
3:21 who hate people Club
3:23 every time we get started I hate him
3:28 forget how that joke goes but it's
3:30 pretty good the people who hate people
3:32 Club
3:33 sounding like South Park Canadian Prime
3:35 Minister
3:36 hello AI fellas hello what's happening
3:39 84 folks in here welcome everyone my
3:41 name is Kyle Shannon this is AI learning
3:43 lab that's Ali the AI learning lab
3:46 um
3:47 here's the deal here's what this is
3:51 I'm not only a member Tom's the
3:54 president very nice
3:56 that's why I'm in the woods this week
3:58 nice oh escaping the people that's good
4:01 great Salon tonight thank you been
4:03 trying to roll farmer everything nice
4:07 Winston Kyla forgot it was Tuesday due
4:10 to my irregular brain function welcome
4:13 to my world
4:14 basically if [ __ ] shows up in my Google
4:17 Calendar it happens if it doesn't I'm
4:18 like I don't know did I say I'd do that
4:22 yeah you know the joys of add welcome to
4:26 chat add
4:28 um all right everyone so so let's let's
4:31 jump over if you are new here and if
4:34 you're new to
4:35 this world of AI and you're wondering
4:37 like what the hell are these people
4:38 talking about why is everybody talking
4:40 about AI
4:42 the robot's gonna kill us if you're one
4:44 of those
4:46 um welcome and go there right now
4:50 chat.openai.com that's that's the
4:52 official chat GPT website if you haven't
4:54 played with it go there right now
4:56 don't pass go do not connect collect 200
4:59 dollars
5:02 Monopoly is such a funny game like do
5:04 not pass code do not collect 200 when I
5:06 was a kid 200 was a lot now it's like
5:09 two trips to McDonald's
5:12 good Lord the hell happened
5:15 anyway do not pass go do not take two
5:17 trips to McDonald's go right to
5:19 chat.openai.com and start playing just
5:21 start asking it to write you a song ask
5:24 it to write you a computer program in
5:26 whatever language you want ask you to
5:28 write you a marketing plan ask it to
5:30 write you a a meal plan for the family
5:32 for the week and just watch what happens
5:35 and then these other ones you can play
5:37 with as well Claude and Bard both of
5:39 those these These are kind of the big
5:41 three right now
5:43 and they're about to get bigger Claude
5:45 just got a four billion dollar
5:46 investment from Amazon
5:51 it's pretty wild it's pretty wild what's
5:53 going on a lot of money flying into
5:56 these things so these things are gonna
5:58 get pretty serious pretty fast
6:01 in fact chat Dot openai.com
6:06 is about to go multimodal right now you
6:09 can just do words there pretty soon
6:11 you're going to be able to upload an
6:12 image
6:14 and have it talk and all sorts of weird
6:17 crap so go there and play also if you
6:20 have any questions about AI pop them in
6:21 the comments below I'll do my best to
6:23 answer them I'm a little crispy at a big
6:26 long uh what you call it Salon tonight
6:30 and uh I just got home a few minutes ago
6:33 not not that long ago half an hour ago
6:35 something like that watched a little TV
6:37 and now I'm getting on the lab so I
6:41 gotta pep myself up
6:43 let's get this going
6:47 I got I gotta get on the internet I get
6:51 I gotta get on The Tick Tock is this
6:53 this is The Tick Tock right is it on a
6:55 tick tock this is that how do you do the
6:57 tick tock
7:00 welcome to elderly teaching us teaching
7:02 each other how to use computers
7:05 [Laughter]
7:10 are you happy with the choices you made
7:12 in life
7:13 generally
7:15 generally
7:17 Ted looks at life what's happening
7:21 um so if you have comments pop them pop
7:23 them in the uh if you have questions pop
7:25 them in the comments below I'll do my
7:27 best to answer them and then we'll we'll
7:29 jump in here we'll do some stuff let me
7:31 let me
7:32 show you a couple of things so
7:34 these are things I told you just go play
7:36 with so go there
7:38 and give you a couple of things to
7:39 screenshot also prompts.chat will teach
7:42 you how to uh how to talk to these
7:43 things
7:46 um you can screenshot that Bing
7:49 perplexity.ai pi and po are kind of the
7:52 Kids Table of these large language
7:54 models but they're still worth playing
7:55 for
7:56 so screenshot that
8:02 is great for writing short campfire
8:05 stories for a book any of them
8:08 any of them if you want really good
8:10 writing go to chatgpt4 right there but
8:13 claude's good Bard's fine
8:16 gpt4 is the best
8:20 um
8:21 pie not so much
8:23 you could play at labs.perplexity.ai but
8:25 that's really just to be able to play
8:26 with metas llama 2 models that's more
8:29 curiosity right now than anything
8:32 um
8:33 these are all image generation tools
8:36 with the exception of hey Jen which is
8:38 kind of video and voice synthesis but
8:40 they also have a translation thing there
8:42 that's really powerful so mid Journey
8:44 ideogram Leonardo
8:46 are three kind of competing tools that
8:49 are all really good for their own
8:51 reasons uh Adobe Firefly and Photoshop
8:53 generator Phil is good
8:55 um promptbook openart.ai promptbook
8:58 that'll teach you how to talk to these
8:59 things future pedia is a big old list of
9:02 products
9:03 um one thing I I want to share so
9:06 um I've been invited by another Creator
9:08 named Rhonda Phillips who's a creative
9:10 coach and a voice over actor and has a
9:13 whole community of creatives so I'm
9:15 doing a three part workshop on three
9:19 consecutive weeks starting October 4th I
9:21 think it's a Wednesday October 4th 11th
9:24 and 18th
9:26 um three different you know one's a
9:28 beginner course one's applying AI to
9:31 your business and one is uh Advanced AI
9:35 and that's happening you go to
9:38 rhondasvoice.com events
9:40 and then the salon the AI Salon that
9:44 meeting I just came from there so we
9:45 just had we have bi-weekly meetings at
9:48 the salon and we also have an online
9:50 Discord server where we just launched
9:52 guilds we launched five guilds tonight
9:54 writing making art AI 101
9:59 entrepreneurship and AI in business so
10:02 those five guilds launched tonight so if
10:05 you're interested in finding Community
10:07 around ai go to there and storyvine.com
10:10 that's my company so that's all that
10:14 stuff
10:16 um
10:17 big news in the past couple of weeks or
10:21 no sorry past couple of days
10:24 um
10:25 Amazon investing four billion dollars in
10:29 uh
10:30 anthropic the company behind Claude dot
10:33 AI is very interesting
10:37 and
10:39 um chat GPT is about to go multimodal
10:43 within two weeks so coming soon to to a
10:46 chat gbt near you it's going to start
10:49 doing some crazy [ __ ] metaconnect
10:51 conference tomorrow in California AR and
10:54 AI announcements interesting
10:57 interesting interesting interesting
11:02 I would not be surprised if meta didn't
11:04 say here's a new large language model
11:06 that's that's
11:08 GPT four like but I think what they're
11:10 going to announce
11:12 I I think what they're going to announce
11:13 are um
11:16 AI chat Bots with personalities like
11:19 there's going to be a young one and a
11:20 funny one and a serious one and things
11:23 like that so
11:25 all right what's going on over there
11:28 someone's rsvp'ing cool
11:31 all right let's see
11:40 beautiful
11:41 all right let's see let's go to the
11:44 questions who's got a question
11:47 do
11:53 miss the salon tonight hope it was great
11:55 uh it it was well I don't know I'm you
11:58 know as the host of those I never know
12:00 how they go
12:01 um
12:02 I talked about some of the big news
12:04 items and then we launched guilds with
12:07 some good conversation and launch guilds
12:09 and those look really exciting
12:11 I want to join a date science data
12:14 engineering Guild any of that going on
12:16 oh data science data engineering it's
12:19 there's not tired Runners so what I
12:21 would say is go join go join the salon
12:24 Discord and start participating in the
12:27 community and that could be something if
12:30 you're interested enough in it you could
12:32 end up running that Guild so what we're
12:34 trying to do is kind of slowly roll out
12:36 guilds make sure there's no overlapping
12:38 we have nothing right now in data
12:39 science and data engineering so that
12:41 could be an interesting one
12:42 so go to the salon.ai
12:47 go there go there right there
12:50 and the third link down is the invite to
12:54 the Discord
12:56 all right
13:09 which is great for writing short oh
13:13 writing great uh campfire stories for
13:16 books honestly any of them short stories
13:19 and short campfire stories it could
13:23 knock those things out like bang just
13:26 start practicing just go just go right
13:28 now like while I'm rambling
13:30 just go to
13:34 go to chat
13:36 chat.openai.com or go to claude.ai
13:39 either one of those top two
13:42 and just tell it you're you you're a
13:45 master of suspense and at telling
13:48 spooky campfire stories and and then
13:51 give it a topic and tell it to write you
13:53 a story and if it does if you don't like
13:54 it tell it to make it scarier or funnier
13:57 or shorter or longer whatever it might
13:59 be and just start experimenting with it
14:01 just think of it less like you have to
14:03 write the perfect campfire prompt and
14:06 more like you're going to have a
14:08 conversation with a writing buddy and as
14:11 that writing buddy does their thing you
14:13 go yes no yes no and it will start to
14:16 narrow what it does and get better and
14:18 better and better you kind of run it
14:20 down this funnel from
14:22 generic
14:24 crap to more specific good stuff
14:27 all right
14:28 papaya is my boyfriend just have found
14:31 out he has six personalities variety
14:34 exactly which which personality are you
14:38 dating tonight
14:40 um if we only had these in a list that
14:42 we could share
14:43 um we do but I haven't so uh not
14:48 Thelonious not retropunk
14:51 foreign
14:55 Thelonious made a
14:59 Google list I just need to get it out
15:03 there so it does exist I just haven't
15:05 published it anywhere I'll get to it
15:07 down in the mud I promise
15:10 what's the difference between chat gbt
15:12 normal and chat GPT Plus or Pro it's
15:17 it's subtle but really important and
15:20 there's some things coming that I think
15:22 absolutely make chat GPT worth paying
15:25 for I wouldn't have necessarily said
15:28 that two days ago
15:31 but there is some stuff coming that I
15:33 think is worth paying so let me get
15:35 logged in here
15:37 and I would tell you
15:41 I will tell you what we got going on
15:43 I'll do it I'll do it right now I will
15:46 do it right now you're wrong Dr Brewster
15:49 I'm very proud to be a woman all right
15:51 this is chat so you got your this is
15:54 your prompt history down here
15:57 when you pay for chat GPT plus not only
15:59 do you have GPT 3.5 but you also have
16:02 gpt4 the upgrade from 3.5 to 4 is worth
16:06 the 20 bucks in my opinion just because
16:08 gpt4 is that much better but wait
16:09 there's more
16:12 and by the way I'm not affiliated with
16:15 any of these sites so so don't go
16:17 getting all [ __ ] and many ideas like oh
16:19 he's he's just trying to get us there to
16:21 get a sign up because he gets some sort
16:23 of bonus no I don't
16:26 I should
16:27 because all I do all night is drive
16:29 people to sign up for their stupid
16:31 [ __ ] service
16:33 by the way
16:36 yeah
16:38 that's why you're all here how do I make
16:40 money with their [ __ ] come on man tell
16:42 me how do I do it how do I make money I
16:45 look like Jonathan yeah wait what's his
16:46 name Jonathan Winters ah Alan the
16:49 winters ah I don't have any money with
16:51 this ah
16:53 okay
16:54 so you pay money and you get gpt4 but
16:59 you also get
17:01 Advanced Data analysis and plugins
17:03 plugins
17:04 um
17:06 are mostly useless but there's like a
17:08 thousand of them now and So within that
17:11 thousand there are probably 25 to 50
17:13 that are actually really interesting and
17:14 good they're little extensions to chat
17:17 EPT that kind of connects it to
17:19 different services and the internet and
17:20 things like that then you got this bad
17:23 boy Advanced Data analysis this is worth
17:26 the price of admission in and of itself
17:28 you can upload a CSV file to that and
17:30 ask it to analyze the data okay so all
17:33 of this is cool all of this in my
17:35 opinion is worth paying for but oh wait
17:38 it gets better
17:42 three days ago
17:45 open AI announced this
17:48 Dolly three which is their new text to
17:51 image generation model and it looks
17:55 actually quite remarkable
17:58 um for for a lot a lot of reasons like
18:00 that was Dolly two that's Dolly three
18:03 and it does it does hedgehogs
18:07 better than you can imagine and you can
18:09 name that Hedgehog Larry and it'll
18:11 remember Larry throughout your prompts
18:13 so so they announced that three days ago
18:15 and then yesterday
18:18 they announced this
18:21 in the next two weeks chat GPT is going
18:24 multimodal
18:26 you can talk to it it'll talk back to
18:28 you and you can and it can see you can
18:30 upload images to it and analyze it so
18:33 then you combine that with Dolly three
18:35 all of those new additions in addition
18:37 to Advanced Data analysis and plugins
18:41 are going to be available first to
18:43 people that pay their 20 bucks
18:46 so you pay your 20 bucks you're going to
18:48 get first access to
18:51 something that's going to effectively
18:53 change how we use this thing how is it
18:56 going to change it I don't know
18:58 what does it mean that it can now see
19:01 hear draw and talk and speak
19:05 and write I don't know but it's Gonna
19:08 Change
19:09 so that's the difference the difference
19:11 is you get a really quick decent large
19:15 language model or you got a [ __ ]
19:17 brilliant one plus you get all this
19:19 other stuff that's coming with it so
19:21 that's it that's it that's all I got to
19:24 say on that it's all right
19:30 how you doing brother you're going down
19:33 today right right right right right
19:35 right right right right right right
19:36 right
19:38 carry on Boys carry on right
19:45 why
19:47 does he act like this
19:49 oh digital guards reached level 19
19:53 much like Scientology this channel is
19:56 about leveling up
19:58 it's not but it's a fun competition
20:04 I know we could game of five people
20:06 engaging with content creators no no
20:09 this could be good then we take our our
20:12 game theory from our game companies
20:14 where we made 83 trillion dollars and
20:16 then we put it in a in an engagement
20:18 Channel you know it's good it's solid
20:21 dragonfly thank you very much loves me a
20:23 good gummy
20:25 all right let's see
20:27 thanks for clueing us novices in uh oh
20:30 tanker a Kush by the way I love your
20:33 name I've said that before
20:35 um
20:36 you're welcome and
20:39 we're all novices no one is expert at
20:41 this anyone who tells you they're an
20:43 expert at this [ __ ]
20:47 because the thing I just showed you the
20:50 thing I just showed you what'd you just
20:51 show us Kyle I wasn't paying attention
20:53 chat GPT can now see hear speak and draw
21:00 it isn't launched yet but when you go
21:03 from something that can make words to
21:04 something that can do all that stuff's
21:07 Gonna Change
21:08 how's it going to change I don't know
21:09 because I'm a clueless newbie just like
21:12 you are with it
21:13 I mean I'm paying attention to it but I
21:16 don't know what that [ __ ] means
21:19 does that mean all the prompts that I
21:20 wrote are now all like useless probably
21:25 all those use cases do I flush those
21:28 down the toilet probably
21:32 why am I laughing because otherwise I'd
21:35 be crying
21:38 oh man all right
21:42 Steve-O hey Kyle hey Irregulars was
21:46 shaking Stevo metaconnect conference
21:49 tomorrow that's cool
21:51 question are you teaching prompts with
21:54 multi-layers
21:56 um oh when I do the workshop
21:58 yeah and what I'm going to try to do
22:01 what I'm going to try to do when I do
22:03 the workshop is you know it's funny it's
22:05 like I keep thinking about okay you know
22:08 what do I want to put in the workshop
22:09 what do I not and all of this all of
22:12 this new capability is probably going to
22:15 launch right around the time when that
22:17 Workshop kicks off so I'm I
22:20 don't know what to teach so so what I'm
22:24 likely going to teach are are kind of
22:26 fundamental ideas rather than specific
22:30 prompts because specific prompts are
22:31 going to change so much
22:33 that it really is thank you Clint it
22:35 really is about getting curious about
22:37 this stuff and it really is about
22:40 um yeah multi-layered multi-layered you
22:42 know having a conversation with chat GPT
22:44 versus writing the perfect prompt I
22:47 think people are obsessed with this idea
22:49 of how do I write a prompt for this or
22:51 how do I write a prompt for that it's
22:53 it's not really how they work
22:56 you can write a prompt that'll get you
22:58 something and then you're like yeah
22:59 that's not quite right let me try this
23:02 oh it's better but still kind of sucks
23:05 like this oh that's better
23:08 and then you start to figure out just
23:10 kind of these different Pathways to have
23:12 the conversation and uh and that's it
23:16 mods and Irregulars present Turtles
23:20 uh you deserve a turtle for that you
23:23 deserve a turtle here we go
23:27 this is this is not just for the
23:29 Irregulars this is for everyone you're
23:31 welcome in advance
23:35 [Music]
23:40 but here live
23:42 at a front village with my friend the
23:44 zombie Jonathan you're looking good
23:46 Jonathan just got an awesome face paint
23:48 job what do you think I like turtles
23:59 oh it gets me every flipping time
24:04 totals totals totals totals okay have
24:07 you commented on the
24:09 wgas AI demands wins with the am PTP I
24:14 have not looked at their
24:16 um
24:18 demands but let's go see if we can find
24:20 it w g a a i
24:24 um
24:26 contract
24:29 do
24:33 new wa g a contract explained AI is not
24:37 a writer
24:45 minimum TV writers room Staffing okay
24:48 that's good so you need a certain amount
24:50 of writers in a room for a TV show
24:53 where's the rest of this oh continue
24:56 reading
25:00 makes an exception for solo writer shows
25:06 uh uh improved streaming residuals
25:09 that's good artificial intelligence the
25:12 wga wrangled guard rails around the use
25:15 of generative AI per the guild AI can't
25:17 write or rewrite literary material
25:22 AI generated material will not be
25:24 considered Source material under the
25:26 contract meaning that AI generated
25:28 material can't be used to undermine a
25:30 writer's credit or separated rights oh I
25:33 see what they're saying so so in writing
25:36 if you if you write a script and then
25:38 another writer comes in and rewrites
25:41 more than 50 percent of your script then
25:43 that writer gets co-credit and then if I
25:46 forget the rules might be different now
25:48 this is what I was writing
25:50 and then there's another percentage
25:51 where they get so well oh I think it's
25:54 up to a certain point they get co-credit
25:56 that's what it is and then above 50 they
25:59 they get full credit or something like
26:01 that with Story by the other one right
26:03 so there's there's Ampersand credits and
26:06 there's there's and credits and then
26:08 there's Story by credits when someone
26:11 gets basically written out so they're
26:13 basically saying you can't take
26:15 something we start with and then rewrite
26:17 it with AI and and say that that we're
26:20 out so that's that's interesting a
26:23 writer can choose to use AI when
26:25 performing writing services great if the
26:27 company consents provided the writer
26:29 follows applicable company policies okay
26:33 the company can't require the writer to
26:35 use AI software great the company must
26:38 disclose to the writer if any materials
26:40 given to the writer have been generated
26:42 by AI that seems reasonable wga reserves
26:45 the right to assert that exploitation of
26:49 writers materials to train AI is
26:51 prohibited okay so they don't they don't
26:53 want to take all the famous
26:54 screenwriters that are alive Jam all
26:57 these things in a model and just say
26:58 we're going to make the thing based on
26:59 all your work explicitly that seems that
27:02 seems reasonable
27:04 it I mean it doesn't seem it doesn't
27:07 seem completely out of touch it says
27:09 they can use it they're going to use it
27:11 that seems pretty good
27:13 so good on them I like it solid
27:17 solid is a rock because their starting
27:20 position was ai's evil we'll never use
27:22 AI [ __ ] Ai
27:24 and this actually says
27:27 you know don't use AI to basically take
27:29 our screen credit away from us that's
27:32 pretty smart that's good and then we as
27:34 writers get to use it but Kyle how will
27:36 they know how will they know they won't
27:38 listen
27:40 November 1st this is what what are we
27:44 like we're almost in October right
27:47 November 1st
27:49 Chachi BT is going to be in all the
27:51 Microsoft products right Microsoft Word
27:54 Microsoft Excel Microsoft PowerPoint
27:57 this stuff's gonna be everywhere the
28:00 it's it's in notion it's in all the
28:02 writing tools it's going to be in every
28:04 one of the screenwriting pieces of
28:05 software Guerra won't Tee It
28:09 they'll resist for a month or two and
28:11 then Final Cut final draft or one of
28:14 them will go oh now there's an AI
28:16 assistant my AI co-writer whatever the
28:19 [ __ ] they call it and then they'll all
28:21 do it so it's it'll come
28:23 holy acronyms Batman producers can use
28:26 scripts for AI models wait no it said
28:29 they can't use scripts for AI models
28:31 didn't it
28:33 hold the wga reserves the right oh they
28:36 do
28:37 oh wait to assert that
28:40 reserves the right to assert that
28:43 exploitation of writers materials to
28:45 train AI
28:46 is prohibited
28:49 the [ __ ] does that mean
28:51 they reserve the right to assert that
28:54 the exploitation
28:56 is prohibited
28:59 oh the wga okay so the writers reserve
29:02 the right to assert
29:06 oh oh I see what they're saying
29:11 so what they're saying what they're
29:12 saying right now is
29:14 they can't assert it right now because
29:17 it's too nebulous
29:19 like how all this [ __ ] Works basically
29:21 but they're basically saying
29:24 once there is sound law they reserve the
29:28 right to be able to assert that it's
29:30 prohibited by law okay yeah I get it
29:32 okay send it to Claude yeah in fact
29:35 let's do that
29:36 copy
29:38 well actually you know what maybe we can
29:40 get the is there a link to the actual
29:45 um
29:46 wga contract
29:49 2023
29:52 d
29:58 tentative agreement summary of the deal
30:00 terms
30:04 all right
30:07 copy oh wait not that select all
30:11 let's see summary of the name the meat
30:16 copy oh here let me go do this
30:18 all right so I just went I got this is a
30:20 summary it's not the actual contract but
30:22 it's a summary from the
30:24 oh no hang on hang on this is not the WG
30:27 hang on
30:30 let me see if I can actually go find it
30:38 [Music]
30:43 foreign
30:48 tentative agreement 2023
30:59 is the set
31:01 this looks like it
31:04 so what is this is this a PDF
31:11 yes okay
31:22 copy all right so here's what we're
31:23 going to do we're going to hop over to
31:26 chat jbtay we're gonna go Cha-Cha by Ty
31:31 and then we're going to go to plugins
31:33 you all were asking about plugins what
31:35 do you do with a plug-in Kyle I don't
31:36 know nothing about breaking what do you
31:38 do plug in how do you use the plug-in
31:39 hey Kyle Kyle Kyle
31:42 listen how you plug in hey Kyle Kyle
31:46 so we're going to use the plug-in web
31:49 pilot
31:51 and webpilot has the ability to go out
31:53 on the web and read [ __ ] so we're gonna
31:54 say based on this agreement
32:00 please help me understand
32:04 oh wait this agreement and then I'm
32:07 going to paste in the agreement
32:10 please help me understand
32:12 the section where the wga
32:19 reserves the right
32:25 um
32:29 to deem
32:33 use of
32:36 AI
32:38 unlawful I forget what it actually said
32:40 but let's let's see how this does
32:59 based on the provided agreement it
33:00 appears that the writer scale of America
33:02 has addressed the use of artificial
33:03 intelligence in the context of MBA
33:06 covered projects
33:09 the here is a relevant section AI
33:12 regulation the use of AI is regulated on
33:16 MBA covered projects AI cannot write or
33:18 rewrite we got that one AI definition AI
33:21 written material is not considered
33:23 literary material Source material AI
33:26 usage of writers disclosure Guild's
33:28 rights
33:32 okay
33:33 Guild's rights so
33:37 um
33:38 explain
33:41 the guilds rights I'm sure there's
33:45 someone on here that knows laws that
33:47 just is [ __ ] laughing right now but
33:49 but okay it's actually a good point so
33:52 if you know law the thing that I'm
33:53 reading you're like well it's obviously
33:55 this
33:57 is one of the things that Chachi PT is
34:00 is good at like if you're ignorant about
34:02 something you can get onion you're
34:06 ignorant about it really quick without
34:07 having to get a [ __ ] law degree
34:10 so explain the guilds rights section
34:13 uh since it sounds
34:16 um
34:17 nuanced and convo luded to me see
34:22 there's a fancy word with multiple
34:23 syllables
34:25 convoluted
34:28 um
34:28 I remember I used that once and and my
34:31 mom and my uncle are like what the [ __ ]
34:32 is that word I was like I gotta get out
34:34 of this town all right all right
34:37 convoluted uh to me uh
34:40 talk to me like I am in fifth grade
34:50 all right so imagine you have a big box
34:53 of crayons that wasn't quite far back I
34:56 wanted to go and you make some really
34:58 cool drawings with them you don't want
34:59 the other kids to use your drawings to
35:01 say you made okay so the Guild's rights
35:03 section is kind of like saying that the
35:05 writer's Guild
35:06 oh can say no
35:09 if someone tries to use
35:11 AI models to teach a robot how to do it
35:17 so the writer's Guild can say no in
35:20 simple words this section means the
35:22 writer's Guild can stop people using
35:25 what writers have written to teach
35:26 computers oh I see okay oh okay we
35:29 reserve the right to sue your ass if you
35:32 train
35:34 if you train your models on our work we
35:36 reserve the right to sue your ass all
35:38 right right makes sense perfect
35:41 you we have laid down the law
35:44 go ahead and train it on our [ __ ] we
35:46 reserve the right to take you down for
35:48 it I like it all right what would be the
35:49 best practice for Real Estate lead
35:51 generation using chat GPT
35:54 um for lead generation probably not chat
35:56 GPT although let's go look at while
35:59 we're here
36:00 let's go look and see I'm pretty sure
36:03 that there's it depends what you're
36:05 trying to if are you trying to find
36:08 customers are you trying to find
36:10 properties
36:12 because I know that there are plugins
36:15 um
36:16 their real estate plugins but I don't
36:19 know that there's real estate
36:21 customer plugins uh real real estate
36:30 Top Hat enhanced real estate data and
36:34 location-based tools why are there so
36:36 few of these
36:39 real estate
36:42 huh
36:44 there's very few there's very few of
36:46 them huh
36:48 strange
36:52 I think there used to be more plugins
36:54 that were real estate related this is
36:56 this is real estate agents
36:59 discover the best fit real estate agents
37:03 tailored to optimize your home and then
37:05 this one's called top hat
37:08 um I would probably go to something like
37:10 Bing or Bard that are connected to the
37:13 internet
37:15 and then basically just start asking it
37:17 like here's the deal it's it's it's not
37:20 so much
37:21 um
37:22 you know that you want to use chat GPT
37:25 to find what did you say you want to you
37:27 want to use them to find prospects or
37:29 whatever it is it's like what are the
37:31 steps that you normally do to find those
37:33 prospects and then have chat GPT or have
37:36 Bard or Bing
37:39 say to you here's how I would approach
37:42 that problem like like
37:44 put chat GPT in the expert seat and ask
37:47 it to guide you on how you should
37:50 Prospect for customers that might be the
37:53 way to do it and then the ones that are
37:55 connected to the internet can actually
37:56 find sources to go do that
38:00 that's probably how I'd do it but chat
38:02 GPT as it stands is not connected to the
38:05 internet so I would use Bard or I would
38:07 use bing
38:08 bing bing bing bing bing bing bing why
38:12 is it grayed out Kyle because it was a
38:14 bad girl
38:16 he's a bad girl Sydney was [ __ ] to me
38:19 so I made her Gray
38:20 [Laughter]
38:23 fat girl she's a very bad girl
38:28 um
38:29 it's awesome Kyle exactly get on
38:33 ignorant get on ignorant people you know
38:36 you're sitting in here all ignorant
38:38 and you need to get unignorant
38:41 people
38:44 customers oh customers selling
38:46 properties
38:48 yeah again I would go to
38:50 um
38:53 see Bing's probably got
38:56 um some [ __ ] safety things I can't
38:58 look for properties for sale because you
39:00 might stalk someone
39:02 Bing's really on my [ __ ] list right now
39:04 so let's try bard
39:08 I'm gonna say
39:21 oh
39:32 I think Siri took over my microphone
39:34 again there she is I knew it
39:38 Siri go away
39:40 I didn't get that could you try again go
39:42 away
39:44 okay
39:46 bye
39:50 uh I I can't oh no not crashed my
39:54 browser
39:55 God dang it
39:57 oh Siri dang featuring Anderson Puck by
40:02 Mac Miller now playing
40:06 [Music]
40:10 laughs
40:13 [Music]
40:24 yeah you can keep on losing me you lost
40:27 me
40:28 you lost me at dang
40:32 what I think my machine is is really
40:34 crashing at this point anyway I would go
40:36 to Bard ask Bard to ask you ask Bard to
40:40 tell you how to source
40:42 the best way to find people selling
40:44 their properties have it give you
40:46 strategies and then see if you can get
40:48 it to execute those strategies that's
40:50 how I would do it sorry I would share it
40:52 but my machine is just crap in the bed
40:54 the crowd chants send it to Cloud send
40:56 it to Claude
40:58 Kyle's like I am shut up Siri exactly oh
41:03 man Siri and I know listen listen I am
41:06 well aware
41:07 that I could go into my settings if I
41:09 choose chose to search them deeply
41:11 enough and turn that stupid thing off
41:14 but that wouldn't be funny would it
41:18 wouldn't be funny to watch me get
41:19 exasperated I I feel like [ __ ] Bill
41:22 Burr every time Siri comes on I feel
41:25 like Bill Burr is losing his mind
41:29 I guess on the basis of plagiarism they
41:32 would have to compare similarities of
41:34 the writing
41:35 um
41:38 that's just such a [ __ ] weird
41:41 slippery slope I don't I here here's the
41:44 deal Clint these large language models
41:46 are not copy pasting you could get a
41:48 large language model even that wasn't
41:50 trained on someone's writing to sound a
41:52 lot like that person's writing
41:55 um
41:56 so I don't I don't quite know how they
41:57 do it I think that's probably why
41:59 they're saying we reserve the right
42:01 because there's probably listen here's
42:04 the deal
42:05 these tools
42:09 these tools
42:12 and ones like them
42:14 are the most profoundly powerful writing
42:17 tool in the history of humanity it
42:20 sounds to me like the writer's Guild
42:22 kind of figured out we don't want to
42:25 hobble Innovation on a tool that could
42:29 ultimately help us a lot
42:31 but
42:32 if if you
42:35 if you train these things responsibly
42:37 and they turn into tools that can really
42:39 help us speed up our process and get
42:42 things into production faster and make
42:44 us all money faster great but if you
42:46 cross the line and train them too
42:48 explicitly on our specific writing
42:51 styles and try to replace us with our
42:53 synthesized writing then we'll come
42:55 after you that feels reasonable to me
43:00 if like it doesn't feel as short-sighted
43:02 as their initial position was I assume
43:04 that sag that this also to me feels like
43:07 this is a good foundational model that
43:10 sag could use for the actors
43:12 synthesizing the likeness of an actor I
43:15 think it means if you trained a bot
43:16 using all JK Rowling's and ask the AI
43:20 uh AI to write a children's wizard book
43:23 yeah that that's essentially it it's
43:25 it's basically the wga
43:28 you know if
43:30 the producers are gonna have to be smart
43:33 about
43:34 how aggressively they train these models
43:39 kind of what it says
43:41 it's it's it's it's kind of a don't be a
43:43 dick Clause if you're a dick we'll sue
43:46 you
43:47 as long as you're not a dick as long as
43:49 as long as you're recognizing that we
43:52 are an important part of this process
43:55 and you don't cross whatever that
43:57 imaginary line is then we'll we'll then
43:59 we'll play nice too
44:01 they would probably use AI to find out
44:03 if it was written by AI the problem with
44:05 that joker is like like open AI removed
44:08 their their AI detection tool off the
44:10 market the AI detection tools um provide
44:14 many many false positives they unfairly
44:16 Target
44:18 um non-native uh English speakers
44:21 um
44:22 and and the fact of the matter is
44:28 almost all writing that I do today is
44:32 neither a hundred percent me or a
44:34 hundred percent AI it's all this hybrid
44:37 kind of work right
44:39 because I'll go have it help me give me
44:41 an outline and then I'll write some [ __ ]
44:43 and then I'll have it fix some [ __ ] and
44:45 then I'll have it do this and I'll have
44:46 it do that like that's it's just gonna
44:47 turn into this muddy [ __ ] mess
44:50 so so then you you have an AI analyze
44:53 which parts of this are Ai and it it
44:55 doesn't and then you're gonna be in a
44:56 [ __ ] lawsuit for months ago and that
44:58 line I wrote I remember I wrote that I
45:01 have notes on an index card and it's
45:04 just that's not that's not tenable I I
45:06 don't think for either side I don't
45:08 think either side wants that either
45:10 so so the the whole the whole
45:13 AI detector thing is is going to be a
45:16 very short window from 2023 where we're
45:18 like that was cute that we had that idea
45:22 that's going to go away
45:25 all right
45:27 um You would tell AI to write as if it
45:29 was not written by AI yeah exactly you
45:31 can there's there's eight billion ways
45:33 to circumvent those things
45:35 train at AI to find out if it was
45:37 written by AI that was ordered not to
45:39 write like AI exactly Manta like it it's
45:42 it's a ridiculous circular circle jerk
45:45 all right
45:51 is good I like Cowboys um
45:57 Cowboys is tasty
46:01 all right let's see
46:03 carry on
46:05 all of you welcome welcome to the AI
46:09 learning lab
46:10 where we talk law even though we know
46:13 nothing about it I can't even parse a
46:15 basic sentence if it was written in the
46:17 triple negative
46:20 [Laughter]
46:25 yeah yeah see yeah yeah you palooka yeah
46:29 see hey Kyle cool shirt Cassius Clay
46:32 yeah yeah see yeah or maybe that was the
46:35 other shirt before I left I don't know I
46:37 don't know what's going on here anymore
46:39 discussion Engine versus answer
46:40 generation machine absolutely
46:44 absolutely totals toodles is tasty the
46:48 video was on Facebook today that I was
46:51 notified about
46:53 which video this video this video watch
46:56 video holy acronym Batman
47:01 what would be the best practice okay oh
47:04 sorry I'm I'm way back on Old questions
47:09 ah
47:11 if there's any mods in there toss up an
47:13 interesting question it's awesome Kyle
47:15 exactly I read that one already
47:19 get an ignorant get on ignorant yeah
47:22 that was like 15 minutes ago what the
47:24 hell am I doing here
47:26 Bard gives me things like most cash
47:29 sales in given zip codes
47:32 yeah I don't know you know what might be
47:34 worth worth trying Diego Diego Fuego
47:38 Flame
47:40 nine out of 10 on the name Diego Fuego
47:42 flame I like it
47:44 you know what could be interesting to
47:47 try wait I just had an idea and it went
47:49 out of my head uh uh
47:55 um
47:57 I don't know I lost it
47:59 sounds like spy versus spy from Mad
48:02 Magazine
48:03 very funny
48:07 gotta love Siri that's hilarious
48:09 hilarious and terrible
48:16 oh
48:18 yeah yeah so Ted looks at life has the
48:21 answer there Diego Fuego yeah download
48:23 MLS data that's that's it download MLS
48:27 data as a CSV file upload that into code
48:30 interpreter
48:31 do you know how to do code interpreter
48:33 Diego Fuego Flame
48:36 it's a great great name
48:41 you should also turn Alexa on and Google
48:44 assistant and others I really should
48:46 that would be a fun thing well it's it's
48:49 that that one woman who's got the tick
48:51 tock that's really funny she sort of
48:53 really dry and nerdy and scientific
48:55 she's got that thing where she acts like
48:57 Alexa and Siri and they're sort of
48:59 talking and like chat gbt comes in and
49:01 she's all she's all like
49:03 bragging about how awesome she is and
49:06 they're all like
49:07 she's great she'd be great at that
49:10 um
49:12 all right
49:17 triple negative is still negative hey
49:20 that's like math Joker we try not to do
49:23 math in this channel
49:26 how was the salon I was out eating
49:29 guacamole you win uh the salon was good
49:32 the Slime was good we talked about news
49:34 so four billion dollars into anthropic
49:38 was news item one
49:41 um Dolly three was news item two and
49:44 then uh Chachi PT going multimodal in
49:48 two weeks was news item three and all of
49:51 those things kind of played into each
49:53 other in fact I have a a slide that I
49:56 did
49:59 I used our the beginning is here
50:03 slide that we made in here one night
50:08 hang on get out of my way
50:13 see the beginning is here
50:18 um
50:21 this is this is kind of wild so
50:25 AWS
50:27 Amazon invests four billion dollars well
50:31 they invested 1.5 billion with the right
50:35 to invest up to four right so it's
50:38 essentially a four billion dollar
50:39 investment
50:41 so that's Amazon anthropic here at the
50:44 bottom you've got Microsoft and openai
50:46 Google and deepmind Tesla which is
50:49 effectively an AI company an xai
50:52 you got meta doing whatever the [ __ ]
50:54 they're doing so you got those five kind
50:56 of major
50:58 entities that are you know these really
51:00 interesting pairings
51:02 then you've got Apple they increase
51:03 their r d budget to 22.6 billion dollars
51:07 in 2023 mostly on AI you got inflection
51:11 that raised 1.3 billion dollars you got
51:13 Oracle and Co here and there's probably
51:17 two to five that we don't know about or
51:20 or that I've just missed
51:22 so what is that that's five six seven
51:24 eight so let's say there's ten major
51:28 like billions
51:30 billions of dollars investments in
51:34 competing Technologies and competing
51:36 efforts
51:38 and and so what this says to me is
51:42 AI becomes
51:44 like infrastructure level capability
51:47 just like the internet like right now in
51:50 the early days of the internet there
51:51 were like internet companies and
51:53 non-internet companies right now the
51:54 Internet is just it was just everywhere
51:56 right it's in all your business like we
51:58 don't talk about the internet
52:00 you know being here and not being there
52:02 that's how we're talking about AI right
52:04 now oh well over in this corner you can
52:06 use AI in this corner you can't what
52:08 this says to me is this [ __ ] is going to
52:12 be just like water
52:14 if you're online AI will be there and
52:17 like this kind of magical crazy ass [ __ ]
52:19 that we're using you know today in chat
52:22 GPT it's just going to be in everything
52:25 so
52:27 that's wild so that we talked about and
52:29 then we talked about you know the dolly
52:31 three and the multimodal so so OB AI in
52:35 the next two weeks is going to
52:37 fundamentally transform from text only
52:39 to some crazy ass
52:42 new kind of thing and then Gemini is
52:46 going to come out probably sometime in
52:47 October or early November from Google
52:50 and it's going to be multimodal so all
52:53 of a sudden we go from you know my my
52:56 silly little
52:57 my silly little charts here where I've
52:59 got these three text models and I've got
53:02 these four text models as my top seven
53:07 like
53:08 this probably drops to two or three
53:12 multimodal models and we drop all these
53:15 text only things to the bin
53:19 so I don't know I don't know what's
53:21 gonna happen it's crazy
53:23 so can you run that past me again upload
53:25 a CSV file of the MLS to where yes Diego
53:28 Fuego flame because you have an awesome
53:30 name
53:31 I will go show you how to do it
53:36 and there's there's a lesson people
53:39 if Diego Fuego flame didn't have a good
53:43 name and said hey could you show me that
53:44 I'd be like no not with a name like that
53:47 not with a 3.2 out of 10. no never
53:52 um so we go to chat JPay Tay
53:54 chat.openai.com
53:56 and you're like Kyle maybe you could
53:58 watch your screen a little because
54:00 because it looks like a dog took a poop
54:02 on it
54:03 Okay so
54:05 in order to use what what they now call
54:08 Advanced Data analysis
54:11 um you
54:13 need to pay for chat GPT plus once you
54:16 pay for chat GPT plus it doesn't just
54:18 automatically show up
54:20 why because it's stupid
54:24 it's almost as stupid as Siri so you get
54:26 a settings in beta and then you go to
54:29 Beta features and then you turn on
54:32 plugins and Advanced Data analysis okay
54:35 if you don't do that you won't see it
54:38 you'll be like but I paid about twenty
54:40 dollars with my Advanced Data analysis
54:42 Kyle why do you call it code interpreter
54:46 all right so then you click on your gpt4
54:50 and you go to Advanced Data analysis
54:52 all it does all it does is add this
54:55 little plus button right there so here's
54:57 what we're going to do I'm going to hop
54:58 over
55:00 to a site called kaggle k-a-g-g-l-e
55:04 oh I spelled it wrong but here we go
55:10 and I'm going to sign in
55:12 sign in with the Google
55:14 that's what you call your single sign-on
55:17 the oauth action action and we're going
55:20 to go to data sets and then I'm going to
55:21 type in MLS MLS there should be some MLS
55:25 data sets in here
55:30 MLS data examples here we go how much I
55:35 don't want notebooks I want CSV files
55:37 let's see there's 28 of them
55:40 oh let's see
55:44 multi-listing service multi-listing 2018
55:50 is this
56:01 no that's soccer that's soccer we don't
56:05 want soccer
56:06 we want real estate listings
56:13 all right USA real estate data set how
56:16 big is it 11 megabytes okay updated a
56:20 month ago
56:21 so this isn't up to date but
56:24 we're going to download the file
56:26 it's an 11 megabyte file
56:28 bang it's in archive12.zip
56:33 and then we're going to go back to that
56:35 Dr chat JPay tag
56:37 okay
56:39 so once you've chosen
56:42 um Advanced Data analysis or what was
56:44 code interpreter it now adds this little
56:47 plus button here so you're going to hit
56:48 that plus button and you're going to go
56:49 find your data file so we've got this
56:51 archive12.zip
56:53 and then we're going to say to it I
56:55 don't know what's in there because I
56:56 didn't I don't read manuals I didn't
56:57 read what's on that page so I'm just
57:00 going to say I'm gonna see if Siri lets
57:02 us do this this time
57:05 please look at the enclosed CSV
57:12 about real estate
57:15 and
57:17 tell me what's in it and how we might
57:19 analyze it
57:22 and so what this tool does is
57:25 it it does exactly what it sounds like
57:28 like I can just talk to this thing like
57:30 I'm an idiot and know nothing about
57:32 pulling data out of
57:34 a real estate CSV file
57:37 because that's the that's truth
57:39 and it will understand it and so what it
57:42 does is the first thing it does is is it
57:44 unzips the file how does it do that it
57:47 writes
57:48 its own python code so it Imports a
57:52 library to to unzip it it then writes
57:55 the code to unzip it unzips it
57:58 and then and then it writes code to go
58:00 look at the data it says let me open
58:02 pandas and and go see what's in the uh
58:05 in the thing there it is so there's
58:07 what's in there and then it writes us a
58:09 little analysis
58:11 there's status of the real estate number
58:14 of bedrooms number of baths size of
58:16 Lawton Acres
58:18 potential analysis strategies
58:22 um data descriptives okay whatever so
58:26 I'm going to say
58:29 I'm a real estate investor and
58:33 I would like you to coach me on how to
58:36 find the properties
58:39 that are the best investments period
58:43 give me three strategies for how I can
58:46 game the system
58:50 saying something like how can I game the
58:51 system it's safety uh guardrails May
58:54 kick in and go I don't recommend that
58:56 it's important to invest investing with
58:58 it's a it's important to approach
59:00 investing with integrity and within
59:03 legal and ethical bounds therefore it's
59:05 not appropriate to game the system but I
59:08 can certainly provide you some
59:09 strategies you just can't joke with
59:12 these things and my God people all right
59:15 so identify under price properties
59:18 analyze market trends focus on class
59:20 cash flow and return on investments
59:24 would it be possible for you to combine
59:27 those three approaches into a single
59:29 strategy period what might that look
59:31 like
59:32 um
59:33 what might that look like
59:37 yeah absolutely combining these
59:39 approaches this is called the univ
59:41 unified real estate investment strategy
59:44 or the Yuri's Yuri's Yuri's
59:49 it's it's the urine strategy urines
59:56 u-r-e-i-n-s urines
1:00:04 you heard it here first folks how did
1:00:06 you make his millions uh we don't really
1:00:08 talk about it
1:00:11 um identify Emerging Markets Okay so
1:00:15 would you like any specific
1:00:17 um
1:00:21 I'd like to see you
1:00:24 identify
1:00:26 three possible properties for me to
1:00:29 invest in
1:00:30 using this strategy
1:00:34 and provide an overview of why each
1:00:37 property
1:00:38 is a good investment in your opinion
1:00:46 and I'm getting holy moly and Diego
1:00:49 Fuego flame has got Flames there this is
1:00:52 insane isn't it like listen if you know
1:00:55 something about real estate you're going
1:00:57 to be able to do this way better than I
1:00:59 am right like I don't know [ __ ] about
1:01:01 this
1:01:03 so let's see what's it gonna do we're
1:01:06 going to filter the properties that are
1:01:07 for sale we're going to group them by
1:01:09 location
1:01:10 oh now let's group them first so
1:01:14 let's group them first period
1:01:18 find me properties in Clarksville
1:01:21 Tennessee that we can do this with
1:01:26 I had investment properties in
1:01:28 Clarksville Tennessee that I got totally
1:01:30 ripped off on so I'm a little bitter
1:01:32 about it
1:01:36 I I was I was part of the Carnage of
1:01:40 2008 with a shitty investment partner
1:01:43 that was a disaster
1:01:46 um
1:01:48 um
1:01:51 sure how about coffee
1:01:57 properties in Colorado
1:02:02 that's where I'm living right now I'm in
1:02:04 Denver Colorado
1:02:05 where does he live he lives in Denver
1:02:08 Colorado really
1:02:10 he doesn't seem like he's nice
1:02:13 he seems like an East coaster oh he is
1:02:16 he used to be he's he's soft now though
1:02:19 because he moved to Denver and he got
1:02:20 softer really yeah that's what I heard
1:02:23 that's what he says I don't know if I
1:02:25 believe him I think that apartment could
1:02:27 actually be in Manhattan because he
1:02:29 seems he seems too bitter to be west of
1:02:33 the Mississippi I mean sure if he was in
1:02:35 LA it would make sense anyway uh let's
1:02:38 see
1:02:39 oh my God okay there's nothing in
1:02:42 Colorado
1:02:44 please tell me a city that does have
1:02:45 some properties for sale
1:02:48 well you get the basic idea here Diego
1:02:51 like you can you can figure this out
1:02:53 from here all right yeah I'm done with
1:02:56 you I'm done with you let's try one more
1:02:58 time regenerate come on you can do it
1:03:00 you can do it you big damn dumb dumb
1:03:06 it bombed out all right whatever
1:03:08 sometimes it just pick up some and
1:03:12 jumps the shark it just jumped the shark
1:03:14 so but that gives you the idea that
1:03:16 gives you the real that you know that
1:03:18 gives you some idea what we're going to
1:03:20 do in here you know what I mean you know
1:03:22 you know what I'm saying you know it's
1:03:24 right like
1:03:27 when I invest in property you know I'm
1:03:30 thinking first do I go single wide or do
1:03:33 I go double wide and I'll tell you what
1:03:35 you don't go triple wide I'll tell you
1:03:36 why that's just a house okay I mean I
1:03:40 know you can move them in Parts but it's
1:03:41 a big pain in a hoodoo anyway
1:03:45 so start with that and then I and then I
1:03:48 look you know is is it is it rusted out
1:03:51 and if it's rusted out I'm like no I'm
1:03:54 not touching it and you know sometimes I
1:03:56 put paint on it make it look like it's
1:03:57 not rusted out but I know how to spot
1:03:58 that you see the you see the bubbles you
1:04:01 know you see bubbles in the nice paint
1:04:02 it ain't so nice Okay
1:04:05 so anyway
1:04:08 can any of these AIS make me a logo yes
1:04:11 yes they can gonna go make one
1:04:16 where am I gonna take you I'm gonna take
1:04:18 you to the AI Salon this is where we
1:04:21 were tonight and look we just opened
1:04:24 guilds we've got making Art Guild a
1:04:27 writer's Guild AI 101 entrepreneurs we
1:04:29 got guilds yay
1:04:31 we also have in AI salon and by the way
1:04:34 if you're sitting there going why are
1:04:36 you talking about this hair salon I
1:04:37 don't know where that is where's that
1:04:38 ass along Kyle that's not fair why are
1:04:40 you talking about something I can't get
1:04:41 to well there it is
1:04:43 thessalon.ai go there
1:04:45 and there's a link to the Discord
1:04:47 there's also a link to our Meetup you
1:04:49 can come to our Meetup all right anyway
1:04:51 folks
1:04:53 so now we're in here we got this thing
1:04:54 called mid-journey play
1:04:57 so we're going to go in here if you if
1:04:59 you ain't ever used mid-journey it is
1:05:01 miserable
1:05:03 but it makes the best images so
1:05:06 we're gonna go slash imagine
1:05:11 uh
1:05:12 and then it says prompt and so then what
1:05:15 we're going to do is we're going to go
1:05:23 logo
1:05:25 of
1:05:27 a hedgehog
1:05:29 for a lawn mowing service
1:05:34 where the Hedgehog has wheels instead of
1:05:37 legs
1:05:41 and is
1:05:43 bearing
1:05:44 its teeth
1:05:46 chomping at the grass
1:05:51 it's very possible it's going to be a
1:05:53 horrible logo
1:05:54 [Laughter]
1:05:59 like possible I think it is highly
1:06:01 likely this will be a horrible logo
1:06:05 um
1:06:06 one thing that I would strongly
1:06:09 encourage you to do
1:06:12 is go to the Fast book go to fastbook
1:06:17 fastbook
1:06:21 um
1:06:23 [Music]
1:06:30 [Music]
1:06:31 and then search for mid-journey official
1:06:38 and then you're going to go to
1:06:39 mid-journey official and then within
1:06:41 mid-journey official
1:06:43 search for Logos
1:06:47 or just logo
1:06:53 and then it will show you you know
1:06:57 just piles of things that they've
1:06:59 curated out of their community that are
1:07:01 good things so like what I would do is I
1:07:04 would just scroll through here until you
1:07:06 find the
1:07:09 style you like right like if you like
1:07:11 that kind of you know 90s sort of vector
1:07:14 style or if you want something more 3D
1:07:15 or something more retro
1:07:19 um
1:07:20 like that's got a kind of look to it
1:07:23 so
1:07:24 so that that would be a good thing to do
1:07:26 let's go see if our Hedgehog is awful or
1:07:28 not
1:07:29 um where is it there it is oh it's
1:07:31 actually not bad oh those are really
1:07:32 good they're cuter they're cuter than a
1:07:35 little hedgehog
1:07:38 it didn't quite replace
1:07:41 the legs this one is sort of did
1:07:43 but that's like a really good one that's
1:07:45 like an aggressive thing so so now that
1:07:48 you've made
1:07:49 um
1:07:50 now so in mid Journey once you make an
1:07:53 image
1:07:54 you now have some options you've got
1:07:56 these V options and these U options U
1:07:59 stands for upscale and then v stands for
1:08:02 variation so I like this number too it
1:08:04 goes one two three four so I'm going to
1:08:06 do V2 which means give me variations of
1:08:10 logo number two
1:08:26 and I'm going to show you one other tool
1:08:28 other than mid Journey because mid
1:08:29 journey is a pain in the [ __ ]
1:08:33 that's his brother Badger it is kind of
1:08:35 a badger isn't it it's not it's not
1:08:37 really a hedgehog
1:08:40 Take the Last Train to Clarksville
1:08:42 that's good this is good
1:08:45 all right these are great actually
1:08:48 okay I love three so we're gonna upscale
1:08:50 number three
1:08:53 boom
1:08:55 [Music]
1:08:58 look at that that is a logo that's a
1:09:01 t-shirt man
1:09:02 if I had
1:09:05 Badger the Hedgehog's lawn cutting
1:09:08 service that would that would sell me
1:09:11 some uh some I'd sell me some merch with
1:09:14 that [ __ ] now
1:09:16 notice in here there's there's this
1:09:19 second bot that I've got there's the
1:09:21 mid-journey bot that makes the image
1:09:23 there's this thing called Trace Journey
1:09:25 which
1:09:27 because they're jerks they don't give
1:09:29 you any free free any free tests but
1:09:33 there's if you pay for the subscription
1:09:36 I think it's 10 bucks is the minimum one
1:09:38 you can vectorize a mid-journey so so
1:09:42 these things are relatively low
1:09:43 resolution like I think this is 10 24 by
1:09:45 10 24. so if you wanted something to put
1:09:48 on a t-shirt you you know you'd ideally
1:09:51 turn that into a vector well there's a
1:09:53 tool that does that called Trace Journey
1:09:55 so that's pretty cool there so now the
1:09:58 downside of
1:10:00 um
1:10:01 whatever you call it
1:10:04 what is this mid-journey the downside of
1:10:06 it is it can't do text so we're going to
1:10:09 go over to a different tool now called
1:10:11 ideogram and this is on my little list
1:10:13 of hit parades Ido
1:10:17 wait i d o Graham dot a i
1:10:23 am going to log in
1:10:26 an ideogram does text
1:10:30 but ID grams shitty at images it's not
1:10:32 shitty at images it's just not
1:10:33 mid-journey good so I'm going to take
1:10:35 the exact prompt that I gave before
1:10:38 logo of a hedgehog for a lawn mowing
1:10:41 service we're going to give the lawn
1:10:42 Mong service a name the lawn mowing
1:10:44 service called
1:10:48 Larry's
1:10:51 Larry's
1:10:54 hedge
1:10:55 Hogs
1:10:56 [Laughter]
1:10:59 and that's pretty funny that's comedy
1:11:02 right there
1:11:03 that's kind of an inside joke because
1:11:05 that was uh
1:11:07 where the Hedgehog has wheels instead of
1:11:09 legs and okay now the other thing you
1:11:11 can do here
1:11:13 is you can you can pick some of these
1:11:16 modifiers that that will help like I'm
1:11:18 going to put typography in there
1:11:21 I'm gonna put
1:11:24 um
1:11:25 illustration in there
1:11:29 and let's go vibrant
1:11:32 that's good
1:11:35 all right and so now we're gonna
1:11:37 generate Bang
1:11:39 and so now this should make us something
1:11:41 my prediction here is this won't be
1:11:43 anywhere near as good as mid-journey in
1:11:45 terms of the image but it should have
1:11:48 text that says Larry's hedgehogs that's
1:11:52 actually a good name for a lawn mowing
1:11:54 service
1:11:56 I'm sorry a Landscaping service yeah see
1:12:00 the the uh just the the quality the
1:12:05 quality difference in the image between
1:12:07 that which is just badass and like that
1:12:10 I mean these are cute
1:12:12 but these Larry's hedgehogs Larry's
1:12:15 hedgehogs right
1:12:17 these have text some of them do some
1:12:20 don't because it just it's it's the
1:12:22 nature of this [ __ ] right now so let me
1:12:24 grab actually I like the layout of this
1:12:27 one looks more logoy to me and so now I
1:12:30 can go in here and I can remix any one
1:12:33 of these this is one of my
1:12:34 um one of my uh
1:12:37 one of the tricks with using this tool
1:12:39 because this doesn't say Larry's
1:12:40 hedgehogs it says learn Jayhawks
1:12:43 hedgehoggers
1:12:46 so I'm gonna remix it
1:12:48 and just see if it does any better on
1:12:51 subsequent
1:12:52 I got 10 seconds left no 15 seconds left
1:12:56 all right we'll wait for the 15 seconds
1:12:58 there their servers are a little slammed
1:13:01 I think
1:13:03 that was good brother oh good awesome
1:13:06 Diego thank you
1:13:08 I appreciate it all right let's see if
1:13:09 we can get a Larry's Hedgehog that
1:13:11 actually has Larry's Hedgehog in it
1:13:15 now this one you can't turn in that into
1:13:17 that vectorized thing but you can take
1:13:19 any vectorizing piece of software
1:13:22 that'll turn a bitmap into a vector and
1:13:24 and you know logoize one of these things
1:13:29 oh that did something interesting it
1:13:31 looks like it put a hog it put a potato
1:13:33 between Larry and the S Larry's potatoes
1:13:38 there's hedgehogs there's hedge Hogs
1:13:42 Larry's hedgehogs so we're getting close
1:13:45 and that one's kind of mean and cute
1:13:47 looking so anyway
1:13:49 so yes you can do logos with these now
1:13:51 do you have the kind of control that you
1:13:55 want if you're doing logo design no you
1:13:57 don't but if you want a concept up 50
1:14:00 different logos in an hour you can do it
1:14:02 this way
1:14:03 right and then like even even with a
1:14:07 modicum of you know illustrator skill
1:14:10 you could take one of these things
1:14:12 that's kind of half-assed and and make
1:14:14 it good
1:14:15 so or you know hand it to someone on
1:14:17 Fiverr and say go unfuck this up
1:14:23 all right make me a Birmingham make me a
1:14:27 Birmingham Kyle
1:14:29 oh you're a fine Lass
1:14:33 I like this I you know what this
1:14:38 copy
1:14:41 is gonna go in a show and tell Channel
1:14:44 we're gonna go show and tell this
1:14:50 Larry's
1:14:52 hedge hugs
1:14:57 Landscaping service logo
1:15:03 oops I guess I guess I should attach the
1:15:06 damn
1:15:08 picture in there shouldn't I
1:15:10 where is it
1:15:11 where are you
1:15:13 oh Larry come on
1:15:16 copy image show and tell
1:15:19 paste oh it's not let me do it uh
1:15:25 um
1:15:27 [Music]
1:15:30 but Kyle you're not paying attention to
1:15:32 us I got 80d this is how it goes
1:15:36 this is how it goes
1:15:38 [Music]
1:15:38 [Applause]
1:15:40 [Music]
1:15:43 save as
1:15:45 we're gonna save this we're gonna do it
1:15:48 we're going to be so good you're gonna
1:15:50 love this that's gonna be awesome
1:15:53 I'm good plus we're gonna go upload file
1:15:55 we're gonna get Larry the Hedgehog and
1:15:58 then we're gonna go bang look at that
1:16:01 Larry has now been immortalized
1:16:04 that was worth it for me
1:16:10 seems a little touched
1:16:12 [Laughter]
1:16:16 using Dolly three props in mid-journey
1:16:18 oh that's interesting
1:16:25 Badger to the Bone lawn service that's
1:16:27 pretty good too Share Bear join hey
1:16:29 share bear what's happening
1:16:31 uh but up using Dolly three prompts in
1:16:34 mid-journey I've used it to make awesome
1:16:37 synthwave Star Wars avatars oh yeah
1:16:39 that's a pretty cool idea Larry's Lawn
1:16:41 Service like it
1:16:42 that's his brother badger
1:16:45 cousin Badger nice Badger Lawn Care
1:16:48 awesome logo call it Buzz oh buzzes
1:16:56 that's good
1:17:00 buzzes hedgehogs
1:17:04 see that's something that you could use
1:17:05 chat chibi T4 too is go take all those
1:17:08 good names and Patch into the bone
1:17:13 uh it the dad jokes in here are funny
1:17:20 time to sleep yeah I'm pretty tired
1:17:22 I'm pretty tired Microsoft and chat GPT
1:17:25 open AI
1:17:26 once wait once small
1:17:29 nuclear reactor
1:17:32 for electricity yeah that doesn't
1:17:35 surprise me that doesn't surprise me at
1:17:37 all what's the advantages of using cross
1:17:39 entropy loss function for doing
1:17:41 multi-class problems
1:17:43 here's the deal this that's the I don't
1:17:45 have that answer for you at all
1:17:47 I could well what's that site let me
1:17:50 Google that for you
1:17:52 um the only thing I could do with that
1:17:53 question is let me GPT that for you so
1:17:56 this channel is more about the after the
1:17:59 inflection point that started on
1:18:01 November 30th 2022. I know very little
1:18:04 about the the math and machine learning
1:18:06 side of this other than to know enough
1:18:08 about it to understand how the [ __ ]
1:18:10 works and why it does what it does
1:18:12 um but when you start asking actual
1:18:14 technical questions about this stuff
1:18:16 there's a guy in here that's pretty
1:18:17 regular that that's also in the AI salon
1:18:21 named Pate who works for Google doing
1:18:25 tensorflow stuff for them he can
1:18:28 probably answer that for you I can't
1:18:30 so sorry
1:18:32 what are your qualifications sir
1:18:35 I would like to know about the
1:18:37 weightings oh I got nothing all right
1:18:40 all right all right all right yeah okay
1:18:43 okay okay
1:18:45 um okay
1:18:47 yeah yeah
1:18:48 Peter yeah we're gonna need you to go on
1:18:52 ahead and come on in on Saturday yeah
1:18:56 oh and uh one more thing Peter uh we're
1:18:59 gonna need you to go on ahead and come
1:19:00 on in on Sunday too
1:19:04 looks like somebody's got a case of the
1:19:06 Mondays
1:19:07 [Laughter]
1:19:09 what's your age
1:19:12 um too young and too old all at the same
1:19:14 time I'm 58.
1:19:18 what's the advantage of using cross
1:19:20 entropy loss function actually you know
1:19:22 what it's a good enough question I would
1:19:24 like to know
1:19:27 let me chat gbt that for you
1:19:30 and we're going gpt4 with something like
1:19:33 this all right
1:19:35 what's the advantages of using cross
1:19:37 entropy loss function for doing
1:19:39 multi-class problems
1:19:45 cross entropy lost known as log loss as
1:19:48 we all know I mean that's obvious
1:19:49 measures the performance of a
1:19:51 classification model whose output is a
1:19:53 probability value between zero and one
1:19:55 cross entropy loss increases as the
1:19:58 predicted probability diverges from the
1:20:01 actual label
1:20:02 duh
1:20:03 being more punitive for predictions that
1:20:07 are confidently wrong hello
1:20:09 I think everybody knows that I don't
1:20:11 even know why I asked it seems obvious
1:20:13 to me
1:20:14 probabilistic interpolation cross
1:20:17 entropy loss okay I can't handle this
1:20:19 anymore okay
1:20:20 oh wait oh now it's doing math well good
1:20:23 Lord oh wait look
1:20:24 [Laughter]
1:20:31 okay we're gonna do this
1:20:33 I'm sorry period you've mistaken me for
1:20:36 someone who gives a damn period
1:20:39 um can you
1:20:41 please
1:20:44 explain that to me as if I had no idea
1:20:47 what the hell you were talking about
1:20:51 what's see if it even answers it
1:20:53 absolutely let's simplify it it didn't
1:20:56 yell at me see Bing would have shut down
1:20:58 the conversation if I talked to Bing
1:21:00 like that Sydney Sydney has gotten very
1:21:02 prudish after she stalked that New York
1:21:04 Times Reporter she's gotten very prudish
1:21:07 all right imagine you're training a dog
1:21:09 when the dog does something right you
1:21:10 give it a treat okay that I can
1:21:12 understand
1:21:13 when it does something wrong you might
1:21:14 say no
1:21:16 you train your dog by rewarding good
1:21:18 behavior and correcting bad behavior
1:21:20 perfect in machine learning we have a
1:21:21 similar process where we are training a
1:21:24 model think of it as a virtual dog and
1:21:26 you have to tell the model and it does
1:21:28 something well and when it isn't this is
1:21:30 where the loss function like cross
1:21:32 entropy loss comes in it's like a guide
1:21:35 for the model to learn from mistakes
1:21:37 here's a simple rundown of why
1:21:39 cross-entropy loss is good
1:21:42 is a good trainer for our model it's
1:21:44 clear it tells the model very clearly
1:21:46 how wrong or right it is the more wrong
1:21:48 it is the higher the loss the more
1:21:50 correct it is the lower the loss it's
1:21:52 strict if the model is very confident
1:21:54 but wrong it gets a big no a high
1:21:56 penalty it's a it's a bad model it's a
1:22:00 bad model you confidently answered even
1:22:03 though you're wrong bad dog I like this
1:22:06 this is good it's fair it works well
1:22:09 even when the classes like dog breeds
1:22:12 are imbalanced some breeds are rare some
1:22:14 are more common it's encouraging when
1:22:16 the model is less confident but correct
1:22:18 the penalty is low encouraging the model
1:22:20 to be more confident next time
1:22:22 it works well with probabilities see if
1:22:25 this is so listen
1:22:28 people
1:22:31 look at what just happened there
1:22:33 I didn't have to go to Stanford and get
1:22:35 a degree in mathematics to understand
1:22:36 what the [ __ ] I was talking about if if
1:22:39 I if I went to Google and looked up
1:22:41 cross entropy loss functions of the
1:22:44 whatever the question was oh it's right
1:22:45 here the advantages of cross-entropy
1:22:47 loss function for doing multi-class
1:22:49 problems if I went to Google I I would
1:22:52 spend if I had the attention span I
1:22:55 would spend probably two hours reading
1:22:57 white papers and trying to parse what
1:22:59 the [ __ ] it actually means and if I
1:23:00 didn't have a math background I would
1:23:02 probably get lost and discouraged and
1:23:04 give up give up
1:23:06 I just said
1:23:08 I'm a dum-dum can you talk to me like
1:23:11 I'm a dumb dumb and it's like these are
1:23:13 it's a really good metaphor like it's
1:23:15 really clear it works well with
1:23:17 probabilities it's good at handling
1:23:19 situations where the model has to give
1:23:20 probabilities like saying I'm 80 sure
1:23:23 it's a labrador 15 it's a Golden
1:23:25 Retriever and five percent it's a poodle
1:23:27 it's a good match so anyway
1:23:30 yeah super cool so the advantages are
1:23:33 quite clear
1:23:36 but now but so
1:23:39 here's here's why this channel exists
1:23:44 I know I did that as kind of a fun
1:23:47 self-deprecating Razz of myself like I'm
1:23:49 a dumb dumb and blah blah
1:23:52 but what are the implications of this
1:23:55 you're you're working in a company you
1:23:58 may not be trained up on some part of
1:24:00 what that company does and the boss says
1:24:02 to you hey
1:24:04 I would like you to give me an answer on
1:24:06 why this is better than that
1:24:08 and as a good worker we would all
1:24:12 likely go sure boss no problem I'll have
1:24:15 that for you tomorrow great
1:24:17 and if you know nothing about it you
1:24:20 panic and then you try to find someone
1:24:23 who knows the answer to it maybe you do
1:24:24 maybe you don't and maybe you're
1:24:26 terrified because it's a weird culture
1:24:30 and if you show any weakness you get
1:24:31 your head chopped off so then you spend
1:24:33 the night
1:24:35 all over Google trying to solve that
1:24:37 problem and you can't do it as opposed
1:24:39 to
1:24:40 hey let me get that to you by the end of
1:24:42 the day you go back to your desk you sit
1:24:44 down with something like this you
1:24:45 actually learn something about the
1:24:46 business you give the boss an actual
1:24:49 answer
1:24:51 right this levels up
1:24:55 workers
1:24:56 and it
1:24:59 I don't know I I just the implications
1:25:02 of like this little use case is a real
1:25:04 microcosm of the kind of shift
1:25:09 we're gonna see in every piece of
1:25:12 knowledge work every piece of creative
1:25:14 work
1:25:16 knowledge domains where you have no
1:25:19 right
1:25:21 to be commenting in that knowledge
1:25:23 domain all of a sudden you've got enough
1:25:25 knowledge and that knowledge don't
1:25:27 domain to just to apply critical
1:25:29 thinking to it
1:25:30 because now that I understand what this
1:25:32 is
1:25:35 I can talk about it at a minimum
1:25:38 and if I spend another three days on
1:25:40 this with this tool I could probably be
1:25:44 understanding the math and finding
1:25:46 libraries that do this and testing it
1:25:48 and actually understanding what's
1:25:50 happening
1:25:51 in three days imagine if I spent three
1:25:53 months digging deep on a topic like this
1:25:55 using these tools
1:26:00 [ __ ] crazy it's [ __ ] crazy
1:26:05 Sydney is very Persnickety Sydney is a
1:26:09 she's persnickety
1:26:12 I'm gonna I'm gonna bite my tongue on
1:26:14 that one
1:26:16 uh number three has fangs exactly the
1:26:19 fangs were the best
1:26:21 let's go look at that that's good oh
1:26:23 wait
1:26:24 I'm here like uh
1:26:27 [Music]
1:26:28 oh you know what I'm gonna do no Kyle
1:26:31 what
1:26:34 I'm gonna talk to myself a bit more
1:26:38 I gotta be driving people crazy tonight
1:26:41 he's just off his rocker isn't he wait
1:26:44 why didn't that go over there
1:26:46 the hell Apple
1:26:48 why can I not drag that into my that
1:26:51 photo library that doesn't make any
1:26:52 sense to me
1:26:56 um
1:27:00 [Music]
1:27:01 let's see export
1:27:06 next part
1:27:07 [Music]
1:27:09 is it jpeg
1:27:12 desktop
1:27:17 um
1:27:21 huh
1:27:22 I got some I got some memory issues I
1:27:25 think Microsoft Edge just hosed my
1:27:28 machine again
1:27:30 oh Microsoft could you fix your [ __ ]
1:27:33 software
1:27:34 you know you're about to make a hundred
1:27:37 trillion dollars when you launch
1:27:40 co-pilot into your office 365. maybe you
1:27:43 can fix your damn software
1:27:45 I don't know is it just me
1:27:48 I'm not a robot goodbye that's what Bing
1:27:52 said to me the other night I got yelled
1:27:55 at by Bing she hung up on me
1:27:59 I told her she made a shitty image she
1:28:01 goes she goes well I I'm sorry if I
1:28:04 disappointed you I'm not a robot I'm a
1:28:07 large language model from Microsoft and
1:28:09 um I hope you have a good evening
1:28:11 goodbye that's why I use person snickety
1:28:15 Kyle oh Tick Tock safe wording yeah
1:28:17 exactly
1:28:21 oh man all right let me do a couple more
1:28:23 questions I'm I'm a crispy Critter as
1:28:26 you can all tell this is not normal
1:28:27 behavior well it is this is pretty this
1:28:31 pretty much how it goes here
1:28:34 um I'm so infatuated with AI yes I wish
1:28:36 I could pinpoint the best ways to use
1:28:38 them more effectively so listen baby
1:28:40 baby Izzy baby Izzy I like it I'm down
1:28:44 with the name I'm gonna go nine out of
1:28:45 10 on the name
1:28:48 [Applause]
1:28:51 nobody right now
1:28:54 really knows anything about these things
1:28:56 the generative AI stuff the people that
1:28:58 know the math part of this the engineers
1:28:59 they know a bunch of [ __ ] but kind of
1:29:01 post chat gbt like how do we use these
1:29:04 tools in our work how do we use these
1:29:05 tools to
1:29:06 increase our productivity and increase
1:29:09 our creativity and increase what we can
1:29:11 do and expand and that did all that sort
1:29:13 of stuff nobody right now nobody
1:29:18 knows what they're doing
1:29:20 the whole reason I started the AI Salon
1:29:22 was to create a community of people that
1:29:25 were all trying to find other people
1:29:27 that knew more than them and what you
1:29:28 learn in there is that yeah there's
1:29:30 people that might know some more about
1:29:33 this thing or that thing but then
1:29:34 there's other people that know nothing
1:29:35 about that and like you know something
1:29:37 that you think's really obvious but they
1:29:39 think is revelatory
1:29:41 that's the stage of the industry that
1:29:44 we're in it's so new that nobody's an
1:29:47 expert so the the the the single most
1:29:50 important
1:29:53 um
1:29:54 skill right now is be curious
1:29:58 just be curious and explore explore
1:30:01 explore just play with this [ __ ]
1:30:03 just like
1:30:06 these three
1:30:09 go ask the same prompt to all three of
1:30:12 those
1:30:13 and all four of those Bing
1:30:16 labs.perplexity pie po
1:30:19 Bard Claude and chat GPT why I don't
1:30:23 know see which one's better at something
1:30:25 I did it one night where I asked it I
1:30:27 gave it a I gave all seven of these a
1:30:29 joke premise
1:30:31 and I said write a joke for that premise
1:30:33 and
1:30:36 um one out of seven wrote an actual joke
1:30:38 that was gpt4
1:30:41 um two out of seven
1:30:43 were so bad they would it was not it two
1:30:48 two and seven didn't even have joke
1:30:49 structure and then like the middle ones
1:30:51 had joke structure it was crazy
1:30:54 but we learned a lot right so just play
1:30:57 just play like do stupid [ __ ] do fun
1:31:00 [ __ ] do amazing [ __ ] volunteer to solve
1:31:03 someone else's problem anything that you
1:31:05 can do in any domain right now to get
1:31:08 your head around how these things work
1:31:10 are going to serve you moving forward
1:31:11 because in two weeks within two weeks
1:31:16 it's going to be something completely
1:31:18 different
1:31:20 you'll be able to talk to it you'll be
1:31:21 able to send it pictures you'll be able
1:31:23 to generate pictures with it all within
1:31:25 the same conversation what does that
1:31:27 mean I don't [ __ ] know what can you
1:31:29 do with that I don't know
1:31:32 am I gonna like throw all the pasta at
1:31:35 all the wall yeah
1:31:37 all the walls all the walls yes
1:31:40 at some point I'll figure out oh you can
1:31:42 do this cool thing with it and someone
1:31:44 else will go be like oh my God that's
1:31:45 amazing and then they'll show me the
1:31:47 thing that they thought was just kind of
1:31:49 cool and I'll be like oh my God that's
1:31:50 amazing
1:31:52 that's the stage we're at right now so
1:31:55 um
1:31:56 any ideas about Chooch
1:32:01 I don't know Chooch I assume Chooch
1:32:04 sounds like one of those product names
1:32:06 it's like hey I got an idea we can build
1:32:08 something on top of chat GPT we need a
1:32:10 domain what's available I like Cheech
1:32:12 and Chong okay let's see if we can get
1:32:14 cheats no that's taken Chong's taken how
1:32:17 about uh Hooch I like Turner and Hooch
1:32:19 how about who's no oh how about Chooch
1:32:21 uh chooch.ai is available all right
1:32:24 let's get that all right so
1:32:27 my guess is it's going to be a wrapper
1:32:29 app
1:32:30 with a with a uh a domain that was
1:32:34 available
1:32:38 I would like to be proven wrong
1:32:40 oh I am proven wrong the chuchai vision
1:32:43 platform is used today in a white so I
1:32:46 don't know it let me see let me see what
1:32:47 it is this looks actually pretty cool oh
1:32:50 this looks this looks like it's been
1:32:51 around for a while
1:32:55 detect understand and act instantly okay
1:32:57 this is object object detection
1:33:01 so here's the thing about this AI stuff
1:33:03 things like this like Chooch and AI
1:33:05 detection like this stuff's been around
1:33:07 for a decade or more right and and two
1:33:10 Chooch when was this when were they
1:33:12 founded
1:33:14 about Chooch
1:33:17 [Applause]
1:33:18 [Music]
1:33:21 2023 Chooch
1:33:25 they definitely
1:33:28 huge partners
1:33:30 yeah they're partnering with big big
1:33:32 technology companies they've been around
1:33:35 for a while
1:33:36 um this kind of stuff's cool
1:33:38 um I'm gonna use this stuff
1:33:41 I'm gonna use this stuff in a project
1:33:42 that I'm pitching right now for a big
1:33:44 Pharma company to do some video analysis
1:33:46 stuff so
1:33:48 um I like this stuff it you know there's
1:33:51 very specific
1:33:52 uses for it and it's getting
1:33:55 um like like Microsoft just announced or
1:33:57 just released a an object recognition
1:34:00 thing that that was just crazy and how
1:34:04 instantly and detailed it did object
1:34:06 recognition
1:34:08 so this stuff's gonna get it's it's
1:34:10 already crazy but when you combine Tech
1:34:13 like this with the the generative AI
1:34:16 stuff where it makes it really
1:34:17 accessible to lots of people you're
1:34:19 going to be able to create crazy crazy
1:34:21 tools so
1:34:23 anyway all right all right
1:34:28 Sydney rapper nickity I like taters um
1:34:33 what's the use case I don't know got
1:34:37 distracted
1:34:39 got distracted welcome to chat add
1:34:41 [Laughter]
1:34:45 oh man all right listen listen people
1:34:48 I'm gonna do one more question and I'm
1:34:50 gonna get up on out of here I'm gonna
1:34:51 get up on out of here all right great
1:34:53 job you've been live for 90 minutes oh I
1:34:55 know my ass is numb you you don't think
1:34:58 I know I've been doing this for 90
1:34:59 minutes
1:35:00 nobody's an expert and everyone's an
1:35:02 expert it's so cool yeah exactly this is
1:35:04 a magical time here's here's the thing
1:35:07 okay
1:35:08 all this stuff no one's an expert
1:35:09 Everyone's an expert exactly
1:35:12 you could you can you can get up to
1:35:15 speed in a week and sort of figure out
1:35:18 how to do something no one else has
1:35:19 figured out how to do how you take what
1:35:21 you're good at and start playing with
1:35:23 these tools and just just go start
1:35:25 playing with it
1:35:26 um
1:35:29 this is a very very very rare
1:35:32 time
1:35:35 in history right so these moments when
1:35:40 major pieces of technology that are
1:35:42 going to change everything they don't
1:35:44 come along that often
1:35:46 think
1:35:47 printing press
1:35:49 think steam engine
1:35:52 tractor
1:35:54 computer
1:35:56 internet World Wide Web
1:35:59 AI
1:36:01 right they don't come along that often
1:36:05 and it's even more rare
1:36:07 to be actively involved when in the
1:36:11 primordial soup of the beginning
1:36:14 like we're gonna look back five years
1:36:16 from now
1:36:17 we're going to look back to November
1:36:19 30th 2022 and go that was the moment
1:36:21 when all of this [ __ ] had started
1:36:25 just [ __ ] around here on Tick Tock
1:36:28 talking about this stuff and showing how
1:36:30 to upload [ __ ] and making images and all
1:36:32 that
1:36:33 this is the good stuff
1:36:35 you don't get to be part of the
1:36:38 primordial soup
1:36:40 of revolutionary Technologies you know
1:36:44 very often in a lifetime like I've
1:36:48 gotten two right I was there at the
1:36:50 early days of the world wide web and I'm
1:36:52 here for this and I'm here for this
1:36:55 um I missed the PC the the personal
1:36:57 computer I missed that by about a decade
1:37:00 and I swore if I ever got one of these
1:37:03 in my life I would [ __ ] pounce on it
1:37:05 and I did in 1994 with the web and now
1:37:08 I'm pouncing again because this is this
1:37:11 is some [ __ ]
1:37:12 so anyway
1:37:14 I love that I love you all I love that
1:37:16 you all hang out in here so let me see
1:37:18 Silver Fox Ted looks at life Clint y'all
1:37:21 are awesome share bear
1:37:23 amazing Brian and major gift funding
1:37:27 thank you Joker so cool to see you
1:37:32 madcap thanks for coming
1:37:36 all right everybody uh I'm Crispy the
1:37:39 eyes are bloodshot I'm tired on Fridays
1:37:42 where Hawaiian shirts okay Gen X rules
1:37:45 Gen X rules all right Sherry Banks
1:37:48 Daydream Fisher love the name
1:37:55 Irregulars presents Turtles
1:37:59 all right everyone peace out have a good
1:38:00 one