aiLL 9/7/23 - Navigating the Wild West Early Days of AI

Description

Chapters

Transcript

0:00 [Music]
0:19 foreign
0:25 welcome
0:27 Thursday happy Thursday
0:30 welcome to the AI learning lab
0:34 see if we can get this thing get this
0:36 show on the road let's get that thing
0:38 rolling let's get it rolling they do
0:40 that
0:41 they do this it's gonna be so good it's
0:44 gonna be so good
0:50 [Music]
0:51 I got
0:53 Sweeney Todd song stuck in my head
0:59 [Music]
1:03 Tobias in the hot tub it's having a
1:06 device
1:12 welcome welcome so there is that just
1:15 joined
1:16 [Laughter]
1:23 like it so there is that
1:26 it's good
1:33 evening good evening gotta get this back
1:35 better lots of stars oh nice
1:39 Stevo welcome welcome Nancy hello hello
1:42 hello
1:44 I got the so so to buy us out with the
1:47 Stars I got the little crescent moon
1:48 over the over the hot tub Waters right
1:51 here
1:51 in uh in honor of Tobias healing is back
1:54 sending the good Juju waves your way
1:57 Ted looks at life with shaken
2:00 how's life looking huh get it oh
2:04 such a card
2:13 if you would be willing uh share the
2:17 live it looks like we're starting to
2:19 like train we're up to 190 likes maybe
2:22 tonight we can get to a thousand let's
2:24 see what we can do
2:27 and Anne Rando
2:32 man was I disappointed by the
2:33 prompt.chat site extremely generic
2:36 instructions yeah I know it's other
2:38 places to learn prop things um that that
2:40 one's
2:41 um very basic on purpose right it's just
2:43 it's just to give people a primer just
2:45 to get them basically started
2:48 um
2:52 I don't actually go to a lot of prompt
2:54 sites I I mean I'm
2:58 it's probably stupid of me but
3:00 for me it's about
3:03 seeing what I can do to crack it myself
3:07 so I just start with the problem and see
3:10 what I can do to crack it I've gotten
3:12 better at when I finally get something
3:13 cracked I I I started a Trello board
3:16 with prompts in it now so that's good
3:20 um what did I do today that was was
3:22 really good
3:24 oh I'm working on it I'm working on a
3:26 site for prompting images and I came up
3:29 with a
3:31 with a series of prompts that I liked
3:33 and then I use chat GPT to iterate like
3:36 50 of them and put them in a spreadsheet
3:37 that was pretty cool
3:42 um I use task Persona context Exemplar
3:45 and format template yeah see you're
3:47 you're more sophisticated than I am at
3:48 the prompting I I tend to be
3:51 it's it's an 80d thing like like
3:53 if I get too formulaic I like I have to
3:55 I either have to be completely formulaic
3:57 or not formulaic at all so I do it more
4:01 like I'm having a conversation with a
4:02 co-worker that's kind of my My
4:04 Philosophy with prompting is is less
4:06 systematic uh which probably means it's
4:08 less better but but uh that's cool
4:11 Tyrant truth what's happening
4:14 task Persona context Exemplar and format
4:18 template yeah that's pretty cool that's
4:19 a that's a good template I mean it's a
4:21 good framework
4:23 um Claude announced quad Pro for 20
4:24 bucks yep I think it throttles me down
4:27 now well claude's been throttling me I
4:29 hit I hit the quad throttle at about I
4:33 don't know it's like 20 or 25 prompts
4:36 um I've noticed that pretty consistently
4:38 uh no you're not late Emilio's wife just
4:40 starting Silver Fox here with shaking
4:44 how many Irregulars with one on I on the
4:46 NFL game I've got one eye on the the US
4:49 Open so Coco goth just got into the
4:51 finals 19 years old
4:54 um and now it's Madison Madison Keys
4:56 playing savalenko savalenko
4:59 um
5:00 yeah I don't know that Madison's gonna
5:03 take her down but if she does then we
5:05 got a uh two Americans in the finals but
5:08 savalenko is like a she's a machine man
5:16 sabalenka not sabalenko sabalinka anyway
5:20 welcome everybody
5:22 this is the AI learning lab as evidenced
5:24 by the little balloon here my name is
5:26 Kyle Shannon uh I'm an entrepreneur and
5:29 a co-founder and CEO of a company called
5:31 storyvine which is an automated video
5:33 storytelling platform I love talking
5:35 about this AI [ __ ] Clint just showed up
5:38 give me a little heart on the head a
5:39 little hearty heart on the head
5:42 um I really like the balloon you created
5:43 last night thank you very much yeah I
5:46 did this in by the way if you're if
5:48 you're new to this generative AI stuff
5:49 and you're wondering where I got that
5:51 fancy illustration with AI learning lab
5:53 on it I must have paid someone for that
5:55 oh no
5:56 that's ideogram ideogram dot a i Clint
6:01 Copperhead was shaking was shaking bacon
6:05 um so here's what we're gonna do if
6:07 you're new to this AI stuff and you just
6:09 stumbled upon this old dude sort of
6:11 rambling
6:12 randomly
6:14 uh that's me
6:15 uh ask any questions you have about AI
6:18 below and I'm going to give you some
6:20 some places to start
6:23 so like I said that's that's an image we
6:25 did in ideogram I'll show you that URL
6:27 in a second
6:30 we're going to start with these though
6:32 we're going to start I got a top seven
6:34 list and these are these are all uh open
6:38 source or not open source these are all
6:41 large language models like chat gbt so
6:43 the very first one there is Chachi BT
6:45 that one right there
6:47 chat.openai.com that's chatgpt if you've
6:50 not played with these go there right now
6:51 just do it just do it do it do it
6:55 go there go there right now
6:58 don't just go yeah I'll do it tomorrow
7:00 no you won't I know you
7:02 I've seen this before you're like yeah
7:04 I'll get to it now right now go there
7:06 now
7:07 and then start playing thanks dragonfly
7:09 I loves me again
7:11 um
7:12 and then prompts.chat will teach you how
7:14 to do that as Anne Rando pointed out
7:17 it's not super sophisticated that
7:18 document but it'll get you started
7:20 there's some primers in there that'll
7:23 talk about how these large language
7:24 models work Claude and Bard are these
7:27 these three chat GPT cloud and barter
7:29 are the three that I kind of cycle
7:30 between when I'm when when one fails me
7:33 I'll sort of cycle around to the next
7:34 one and I just keep going in circles
7:36 like this
7:37 um that tends to do it pretty good so
7:39 this is the uh this is the adults table
7:41 at Thanksgiving here's the Kids Table
7:44 so so these are the ones that didn't
7:46 quite make the cut and not for any
7:49 reason other than just sort of pure
7:51 usefulness Bing is is actually pretty
7:53 good it's Microsoft's gpt4 connected to
7:56 the internet it makes images my
7:58 challenge with it is the safety guard
8:00 rails on it are so high that it's just
8:02 it's like being in [ __ ] Romper Room
8:05 um perplexity.ai is a good research site
8:08 and then labs.perplexity is they're
8:10 hosting meta's llama 2 models and their
8:13 code llama model that lets you play with
8:15 those for free to see what that's all
8:16 about Pi dot AI is chatty GPT
8:20 they probably wouldn't like it if I
8:22 called it that to their face but that's
8:24 what it is if you want to talk to a chat
8:26 gbt like with your voice and have an
8:28 interesting conversation it's really
8:30 good at that if you want to do actual
8:32 research with it oh is it a pain in the
8:33 ass and then poe.com oh that's cool
8:36 thank you very much I put my hat Under
8:38 the thing there but
8:40 um
8:41 all right and then poe.com is you can
8:44 play with lots of different models I
8:46 think there's five or six there now and
8:48 then
8:49 um they've also got a bot maker so you
8:50 can make your own Bots and practice your
8:52 prompting and tell all your friends I
8:54 made a bot
8:55 it's actually really easy
8:58 um on the image front if you want to
9:00 make pitches if you want to make pretty
9:02 pitches like the the AI learning lab
9:04 thing I did here
9:07 um here are the things to play with
9:09 mid-journey is the best it's the one
9:12 that's the most sophisticated the
9:14 pictures are just prettiest they're just
9:16 the prettiest
9:18 ideogram can do text mid-journey can't I
9:21 do grams like 10 days old this shit's
9:23 moving so fast that like this here's
9:27 here's here's how fast this stuff's
9:29 moving what used to be in this spot was
9:32 Adobe Firefly so so ideogram this 10 day
9:36 old entry booted Adobe off the list and
9:40 then uh Leonardo and clip drop are both
9:43 things sitting on top of stable
9:45 diffusion clip drop is from stability Ai
9:48 and then Leonardo's just some other site
9:50 that's a really good site sitting on top
9:52 of stable diffusion and then
9:55 uh prompt book will teach you how to
9:57 talk to these image things okay and
9:59 future pedia is a list of like 4 000
10:01 sites so go play with that stuff one
10:04 thing you should know about this
10:06 particular Tick Tock live this
10:09 particular channel is it's so popular
10:12 that even the scammers have copied us so
10:14 so there's a scammer out there who put
10:17 an extra l in front of the word learning
10:19 so he's got a stutter AI learning lab
10:23 and he's trying to scam the subscribers
10:27 of this so go check your subscription if
10:29 you subscribe to my content make sure
10:31 you're not subscribing to his content
10:33 because he'll reach out to you and try
10:35 to sell you some crypto [ __ ] little
10:37 [ __ ] so feel free to uh uh whatever
10:40 you call it report him I've reported him
10:42 many times uh report him for a copyright
10:45 violation apparently they take those
10:47 more seriously than they're
10:49 impersonating someone so he has copied
10:51 all my videos over or a pile of them so
10:54 it's absolute copyright violation so
10:56 [ __ ] him all right fantastic oh by the
10:59 way I swear sometimes
11:01 all right so let's get this up here
11:04 we'll show this so people have a little
11:07 reference to play with and I'll kind of
11:09 bounce back and forth between pretty
11:12 pictures and that all right all right is
11:15 the scammer back yes the scammer's back
11:21 stuff when you're stupid
11:23 is that right yeah um yeah I think so
11:28 okay pop any comments down below I'll do
11:31 my best to answer them I will get behind
11:33 I promise you that uh the Irregulars
11:35 there's a bunch of people in here that
11:36 show up all the time they're super cool
11:38 they'll help you out if I haven't
11:40 answered your question uh you know help
11:43 each other out do that I'll do what I
11:45 can and let's get rolling let's get
11:48 rolling rolling rolling I think they
11:51 should pay you for the right to chatty
11:52 GPT come on inflection
11:55 they just raised 1.3 billion dollars you
11:57 don't think they could drop me a hundred
11:59 grand for coming up with chatty GPT come
12:01 on I don't think I came up with it
12:03 though I it was probably you Emilio's
12:05 wife it was one of the Irregulars came
12:06 up with that
12:07 it wasn't me but I took it don't get me
12:11 wrong I also didn't come up with
12:12 mansplaining as a service so if you've
12:14 if you've not played with these tools
12:15 they do lie so I like to call it
12:17 mansplating as a service like like a
12:20 good man in your life they will
12:21 confidently answer you even if they're
12:23 wrong
12:24 laughs
12:27 all right
12:30 old dude with the great hair I don't
12:32 know a great now I wouldn't um great not
12:35 the right at least there's some here
12:38 goes up Kyle what's up Kyle what's up
12:42 cow what are you doing cow
12:46 um did you get your moderators uh I have
12:48 so
12:50 oh that's the other thing
12:52 um
12:53 if you want to join I created a
12:54 community called the AI salon and we
12:57 meet every other Tuesday and our next
13:00 meeting is next Tuesday
13:01 and
13:03 um
13:04 we we also have a Discord we have an
13:06 online Discord and there's there's a
13:07 group in there for for the Irregulars so
13:10 I've got people in there who have uh
13:13 said they were going to be Vol or
13:14 moderators here we just haven't had the
13:16 conversation about when to actually do
13:18 it and what they actually do and all
13:20 that sort of [ __ ] so I have volunteers I
13:23 just have not activated them yet I know
13:25 is that the right word I have not we've
13:27 not said I have not set it up there you
13:29 go
13:31 that sounded like your dad voice just do
13:33 just do it do it
13:36 just do it
13:38 I'm I'm I stole that from uh Matt Rife
13:41 you know the comedian Matt Rife the
13:43 super hot dude where he's what was he
13:46 someone someone's doing something and he
13:48 just he got all like Dad voice on him it
13:51 was pretty funny all the women were like
13:56 do it
14:00 oh can you check out Claude tonight to
14:02 see if they are limiting under 100K
14:05 tokens
14:06 really got an error huh that's weird
14:10 let me go look and see if there's
14:11 anything different there
14:15 I used club today I mean I got I got
14:18 rate limited like after 25
14:22 queries see here's what I was working on
14:24 wait what is that relay oh no that was
14:27 something I was doing in here
14:29 um
14:34 let me attach a file here what's in here
14:50 what's this
14:58 oh here's a so this is this is the uh
15:02 this is the cover
15:03 oh wait is this this must be in in
15:05 iCloud
15:09 exploding yeah 605k
15:13 this was a book so so there's a uh
15:16 there's a project called GPT author that
15:19 will write you a full novel so I just
15:21 uploaded my my novel
15:24 um
15:25 please
15:27 um
15:28 summarize
15:32 this classic novel
15:37 written by me I bought
15:39 [Laughter]
15:47 don't turn around
15:49 based on the chapters provided this
15:51 appears to be a fictional novel that
15:53 explores the Journey of Malcolm an
15:55 ambitious and Innovative artist who
15:57 pushes the boundaries and disrupts Norms
15:59 through his avant-garde creations
16:02 it seems to be working that was a pretty
16:04 big
16:05 605k it wasn't huge it wasn't quite a
16:08 megabyte but you know
16:10 it seems to be Cloud seems to be working
16:14 um
16:16 but yeah Claude did announce Claude Pro
16:19 today so so yet another service asking
16:22 for 20 bucks a month
16:23 it
16:26 is it worth paying the 20 bucks a month
16:29 to Claude I I don't know if you if you
16:32 regularly do deal with long documents it
16:34 is because I think it's like twice as
16:36 fast and
16:38 five times as many interactions
16:43 which is probably enough
16:47 um be right back Kyle okay welcome Vicky
16:50 even in France even in Vicky what's
16:52 shaking What's going day on day on down
16:54 down down
17:01 this was a limit on input length
17:04 I don't have a big giant ass file
17:08 oh you know what I can do let me go get
17:09 hang on
17:12 let's go
17:14 um
17:15 screenplay
17:19 PDF free Scarface
17:34 read the script
17:38 all right can I copy
17:51 copy let's see if we can copy it
17:55 uh
18:01 yeah this I think this is the whole
18:05 thing of Scarface uh
18:08 write a sexy Hollywood
18:13 log line
18:16 or lake El Cog line
18:38 um
18:48 seem to have stumped it I wonder if
18:50 something I copied had some weird ass
18:53 code in it I bet it did
18:55 foreign
19:04 Tony Montana
19:07 don't get high on your own Supply
19:11 when a ruthless Cuban Refugee crimes a
19:13 top Miami's cocaine Empire his
19:16 unquenchable thirst for power money and
19:18 women leads him down a deadly spiral
19:21 that's actually pretty good that's
19:23 Scarface in a nutshell it did it
19:26 [Laughter]
19:28 yay you did it
19:31 um
19:34 give me a character breakdown but
19:37 describe the characters as if they were
19:40 in a funny cartoon
19:46 ah let's see
19:49 the only Montana Tony Montana
19:53 and Tony
20:02 let's see what it does by the way Claude
20:04 what I'm that I'm using right now what
20:06 what Ted looks at life is talking about
20:08 so it's got a hundred thousand token
20:11 context window which basically means
20:13 about 75 000 words
20:15 and they just announced a pro version of
20:18 this today and what he's wondering is if
20:21 they're limiting how much non-paying
20:23 members can put in
20:24 I haven't I haven't hit that I used it a
20:27 decent amount today and I used it with a
20:28 pretty pretty big document
20:31 Tony Montana
20:33 a tough topic talking Cuban Refugee with
20:36 a scar across his face hence the title
20:39 of a film smoke cigars snortz Coke and
20:42 blast bad guys with hilarious quips
20:44 think Cuban Bugs Bunny mixed with Elmer
20:47 Fudd
20:52 Tony's handsome easy going best friend
20:54 and partner in crime he's a lady killer
20:56 who can wriggle his tongue at the
20:58 regular his tongue like a motorboat
20:59 remember that scene
21:03 oh that's good chichi
21:07 Gina Montana
21:10 yeah so it works seems to be working
21:13 I don't know what to tell you I don't
21:16 know what to tell you
21:18 all right all right all right all right
21:21 all right all right all right my name is
21:24 Rosanna Rosanna
21:30 all right
21:32 oh Madison Keys
21:35 up a break in the first set
21:40 wait oh what was that did I get yelled
21:43 at oh some comments in live were
21:45 filtered to protect the community's
21:47 experience are you all being nasty to
21:49 each other
21:50 don't be nasty in here this is a nice
21:52 Zone we have fun in here this is fun we
21:55 have fun now we're gonna have fun
21:56 tonight you're gonna have fun we're
21:58 gonna have fun so don't be nasty don't
22:00 be all like nasty gramming it up
22:03 you're crazy people you're all crazy
22:06 y'all just crazy
22:08 you're all crazy crazy crazy crazy crazy
22:13 all right there we go
22:16 there we go there we go yeah yeah time
22:19 for Wapner yeah yeah Charlie Babbitt
22:22 yeah I'm an excellent driver yeah go to
22:25 Walmart go ahead let's go to Walmart
22:27 yeah
22:31 I'm Charlie Babbitt Charlie Babbitt
22:33 [Laughter]
22:41 I'm chasing people away there's like no
22:44 one left uh if you would go ahead and
22:46 share the live because I'm just chasing
22:48 them out tonight
22:52 oh good lord what are we gonna do with
22:56 me
22:57 I don't know I thought so too just a
22:59 slap I can be snarkier you can be all
23:03 right retro punk I know that guy so
23:05 retro Punk wait was it retro Punk
23:11 no I you both have purple you and
23:14 Thelonious both have purple icons and
23:17 you have cool
23:20 um cool names that are about the same
23:22 length so Thelonious and retropunk I
23:24 always I always mix you two up
23:28 but Thelonious sent me something from
23:30 hey Jen today that that they released
23:33 the beta of their video translation tool
23:36 and he sent me an example that he did of
23:38 him he just recorded himself speaking in
23:40 English and then translated it into him
23:44 speaking French
23:45 and it like re-sinks the lips and
23:47 everything it was really flipping good
23:49 so
23:50 crazy so that's coming it's called hey
23:52 Jen I don't have access to the uh
23:54 translation thing even in AI with Kyle
23:58 hi Uncle Kyle headshot drecker what's
24:00 happened or Troy Decker Troy Decker how
24:03 you doing hope you're doing good you
24:05 doing good yeah welcome in welcome into
24:09 the lab all right we're gonna have fun
24:11 tonight what time on Tuesday so Tuesday
24:14 night for the AI salon
24:17 it is
24:19 um 5 P.M to 7 P.M Mountain Time
24:22 so
24:24 four to six Pacific 7-9 East Coast
24:29 you can figure out Chicago you Chicago
24:31 people are smart figure it out
24:34 you're in St Louis you get you can do
24:36 the math you're somewhere in between
24:37 here and there
24:40 five to seven Mountain Time
24:42 and uh
24:44 go if you go to the salon.ai
24:48 the second link is the Meetup the first
24:50 link talks about what the salon is what
24:52 our values are that's important
24:55 it's basically like a list of values
24:57 like be curious be adventurous don't be
24:59 a douche it doesn't say that but that's
25:02 sort of what it means there's like
25:03 empathy and inclusion and like just be
25:05 nice be nice support each other which is
25:08 what this channel is about too odd odd
25:10 how that works right
25:12 um
25:13 so go there sign up for the Meetup and
25:15 then you'll get access to the uh to the
25:17 URL for the Meetup and we have a good a
25:19 good grand old time in those things
25:24 everyone gets to introduce themselves
25:26 it's actually pretty cool hey hey
25:29 Sam corn what's happening buddy
25:31 welcome welcome welcome to goss not sure
25:34 if I'm an AI Salon welcome or kicked out
25:37 or both
25:38 well togas it it remains to be seen
25:43 you gotta you gotta you can get in the
25:45 door but do we let you stay I don't know
25:48 I don't know
25:49 but join us Tuesday
25:51 just go sign up on the on the Meetup
25:53 there and uh
25:56 and then RSVP for next Tuesday's meeting
26:00 I need more queries on Claude or a paid
26:02 plan so they came out with a paid plan
26:04 today Vicky 20 bucks a month I don't
26:07 know if it's live let me let me look
26:09 let's see
26:11 close that close that uh uh Claude dot a
26:17 i k s
26:20 account set subscribe to Pro
26:24 level up your Claude 2 usage Priority
26:27 Access during high traffic periods Early
26:30 Access to new features what are those
26:32 new features pay your 20 bucks and we
26:34 might tell you
26:37 [Music]
26:39 hey
26:43 learn about Cloud Pro let's go learn
26:45 about it shall we
26:47 do
26:48 [Music]
26:51 does Cloud Pro have usage limits yes but
26:53 they're five times the usage limits
26:55 compared to our free service which I
26:57 think was about 25.
26:59 well actually it's not so it says the
27:02 number of messages you can send will
27:03 vary based on the length of the message
27:05 including the length of the files you
27:07 attach
27:08 we will provide a warning before eight
27:11 messages or ten messages remaining
27:14 and we'll reset every eight hours
27:16 practically what do these limits mean
27:20 if your conversations are short you can
27:23 expect at least a hundred messages
27:26 if they're long you can expect less yeah
27:28 that's that's what I've been
27:29 experiencing
27:31 if you upload a copy of The Great Gatsby
27:34 you may only be able to send 20 messages
27:36 in that conversation within eight hours
27:39 as each time you send a message Claude
27:41 re-reads the entire conversation
27:43 but that's actually so even though
27:45 that's you know
27:47 20 doesn't seem like a lot
27:49 if you're uploading a 50 000 word novel
27:53 or a 50 000 word annual report or a four
27:56 hour transcript of a of a conversation
27:59 you had the fact that you can just do it
28:02 without having to chop up your your
28:04 documents it's it's a pretty big deal
28:06 clouds are pretty clouds are pretty
28:09 useful tool
28:10 um I don't I don't know that I'm gonna
28:12 pay the 20 bucks for it because
28:16 I haven't like the the limit hasn't
28:19 really hasn't really hurt me yet but
28:21 we'll see
28:22 thank God for these new phones that are
28:24 dunk waterproof yeah I slipped
28:28 like you be careful on your back dude
28:32 who cares about your phone yeah I know
28:34 these uh it's amazing I take I watch
28:36 tick tocks in the shower I was saying
28:38 that the other night occasionally a
28:40 water drop will hit it and it'll turn on
28:42 the live I just see myself naked I'm
28:44 like
28:45 scares the [ __ ] out of me
28:49 uh oh man
28:52 Claude Pro damn I know right I picked I
28:55 paid Claude I use it a lot very cool
28:57 Silver Fox yeah quad's pretty good I
29:00 gotta say man it's I we we got some some
29:03 really cool Wars wars coming up this is
29:05 like you know we're entering what's the
29:07 equivalent of the browser wars in the
29:09 90s where um
29:12 you're gonna have open AI they're gonna
29:14 They're Gonna Keep swinging because they
29:15 have so much money from Microsoft and
29:17 they they now have Revenue
29:19 um you got inflection with uh with pie
29:23 you got anthropic with Claude you got
29:25 Google they're Gemini their Gemini
29:28 Project seems like it's going to be
29:29 pretty Kick-Ass so that's four right
29:32 there then you got whatever Elon Musk
29:34 comes up with x dot a i and then if
29:37 you're on Apple devices you're gonna
29:39 have whatever Apple comes up with
29:41 so it's gonna be it's gonna be a pretty
29:44 Saucy mix out there
29:45 probably by Christmas all I want for
29:48 Christmas is gpt5
29:55 how many 20 months can we hit 20 bucks a
29:58 month can we handle I know it's it's it
30:00 really is exhausting right now and it's
30:02 like
30:07 I think I'm up to about
30:11 120 bucks a month on this [ __ ] it's like
30:16 but at the same time like I don't want
30:19 to not use it I'm just kind of waiting
30:21 for like you know all the free things in
30:23 the open source Community to get their
30:24 [ __ ] together and then someone will
30:26 start a foundation and host like a
30:28 hundred and 80 trillion parameter model
30:32 that's open source and free and it does
30:34 everything in a microsecond and you know
30:37 we never have to pay anything ever again
30:39 but you know that's probably four or
30:40 five years away
30:42 so in the meantime
30:44 Jesse from Austin hopefully everyone's
30:45 investing into AI it's a big and only
30:48 going to get bigger
30:50 that's Jesse from Austin
30:52 going in with the Investments
30:55 um yeah I I tend to agree I haven't
30:58 invested but
31:03 it's true it is true
31:06 um yeah it's it's uh I think I I I would
31:11 tend to agree [ __ ] is gonna go crazy
31:14 Claude says five files 10 megabytes each
31:17 that sounds about right by the way
31:19 daughter just got back from burning man
31:21 best 10 days of her life the Mudfest
31:28 that looks so [ __ ] miserable didn't
31:30 it it just I saw that thing it's like
31:33 like I mean I know I'm old but like
31:38 I don't know two three years you know
31:40 what it was pre-covered I would see
31:42 pictures of burning man I'm like yeah
31:43 I'm gonna go there one day and then in
31:45 covet I was just saying I don't think so
31:47 and then I saw the [ __ ] mud storm of
31:49 Burning Man this year and I'm like I'm
31:51 [ __ ] good no I don't I don't need to
31:54 pretend survival and then actually get
31:57 trapped in the middle of a [ __ ]
31:59 drowning lake bed and have to survive
32:02 I'm good I'm good I'm good I'll just be
32:05 right here I'll be here in Denver y'all
32:06 y'all go get mud caked
32:10 Claude says message is 51 over length
32:13 and it's just 34k
32:18 that's very strange Ted
32:22 because I just upload I just uploaded a
32:24 605k file and a
32:29 whatever whatever the the
32:32 um
32:38 the screenplay
32:40 the screenplay was 205k
32:43 so if you're 51 over input length
32:47 you sure you're at claude.ai
32:50 I've never had
32:52 an error message on input length I would
32:54 I would I would restart your browser try
32:57 a different browser restart your machine
32:59 you might have some weird caching thing
33:00 in there or if you haven't done it click
33:03 on the like I assume you know this if
33:05 you click on the the anthropic logo in
33:08 the upper left hand corner that resets
33:10 your that resets the the thing if you're
33:12 still having it from kind of a a
33:15 pristine
33:17 um chat then something's up you must
33:18 have some weird cache thing going on in
33:20 your browser
33:23 um be right back okay okay so hello any
33:27 yellow friends including the oh hello
33:30 any yellow friends including the
33:32 submarines very cute though guys
33:35 just kidding not leaving oh Emilio's
33:38 wife you're a card
33:41 um
33:42 Winston's in welcome Winston I think the
33:45 limit is tokens but it's not context
33:48 wait it's not context window limited and
33:51 it's based on usage I think I think it
33:53 is based on usage
33:56 um but he shouldn't he shouldn't be
33:58 getting your 51 over with the 34k file
34:01 you can put 100 megabytes in that thing
34:03 or 50 megabytes five five files of 10
34:06 megabytes each right so 50 megabytes so
34:09 he shouldn't be anywhere close
34:12 so that's just weird man you know what
34:15 it is it's probably conspiracy he's
34:17 probably the the feds sniffing in on
34:19 your machine and they they put in like a
34:22 a nine and a half like a 49 and a half
34:26 megabyte file that's like your your
34:29 record man and you know they're probably
34:32 in there like just watching everything
34:34 you do holy crap Madison Keys up three
34:36 nothing against sabalenka dang
34:40 could be an All-American latest final
34:45 very cool all right
34:47 oh Tick Tock is jumping to the bottom
34:50 jumping to the bottom
34:54 we got Tic Tac didn't promote our little
34:57 live here tonight and feel free to
34:59 um feel free to share this if you want
35:03 and uh
35:05 you know do the likey like thing
35:09 we're nice Irregulars the Irregulars are
35:12 very nice I like the Irregulars
35:14 foreign
35:19 all right trying to figure out where I
35:21 was
35:24 thanks Kyle Claude just hates me yeah
35:26 it's you know there's I I I uh there's
35:30 something like seven planets in
35:31 retrograde I think Venus just came out
35:33 of retrograde so it's probably six now
35:35 or five I don't know there's a whole
35:36 bunch of basically the planets are all
35:38 hiding from whatever retrograde means
35:40 but it's not good it means communication
35:43 sucks so I think Claude is probably
35:45 programmed by an astrologist and you
35:49 know that's why you're having these
35:50 issues so if you just wait for the
35:52 planets to align everything will be fine
35:53 Ted everything's gonna be just fine it's
35:57 dead he's upset with Claude
36:01 Gavin Newsom in California just signed a
36:04 bill to investigate AI what does that
36:06 mean
36:07 what does that mean
36:10 Newsome
36:12 AI investigate is that really what it
36:15 says
36:18 Newsome signs executive order to study
36:21 AI
36:22 oh just to study it okay AI use and
36:25 risks good
36:26 I mean Nissan wants to shape is for
36:29 future can California lead the way
36:34 uh the executive order sets the stage
36:37 for potential regulation around what's
36:38 known as generative AI technology which
36:40 Saudi raised concerns about plagiarism
36:42 bias and child safety the governor of
36:44 California the governor in California
36:46 lawmakers
36:48 thus far have been cautious about
36:49 regulating technology they might not
36:51 fully understand nobody [ __ ] fully
36:53 understands it what are they talking
36:55 about hindering business Innovations we
36:57 recognize both the potential risks and
36:59 tools that's not a bad thing they're
37:02 just like listen
37:04 governments need to figure this [ __ ] out
37:06 because this is going to change a lot
37:08 like when this starts [ __ ] with
37:10 employment numbers they're going to need
37:11 to have their [ __ ] together
37:13 Madison is up for love in the first set
37:18 good Lord she is destroying sabalenka
37:24 um
37:25 833 Kyle time don't be nasty in here
37:27 yeah people don't be nasty in here Dr J
37:30 good to see you
37:33 bubbles
37:34 is there a ritual to be an irregular yes
37:37 show up show up regularly the more
37:41 regular you are the more irregular it
37:44 proves that you are and then
37:46 you'll look at images like this one
37:49 and you'll be like yeah that's me
37:52 I'm not weird I'm irregular
37:58 if you find stupid [ __ ] like that funny
38:00 then yeah that's that's the you just
38:03 show up you show up
38:06 it's like the old Lotto commercial you
38:08 gotta play to win gotta play to win so
38:11 if you're playing bubbles you're
38:13 irregular
38:16 that's why we sell Metamucil in the
38:19 store
38:25 all right
38:27 no one left there are 40 Irregulars well
38:31 I know I I meant like the like normally
38:33 tick tocks got us there's normally like
38:36 a 40 to 50 sort of regular bunch in here
38:38 and then normally Tick Tock will like
38:41 throttle it up to like 75 or 100 or 120
38:44 and then back down to 75. but tonight it
38:46 it started at like I don't know 90 and
38:49 went quickly to 50 and now it's just
38:50 been slowly going to do it but we're
38:52 slowly getting back up see I think it's
38:54 your like train
38:57 um no that was me oh retro okay it was
39:00 you retro Punk um yeah retro Punk was
39:03 the one that did the video converted it
39:05 to French with uh with hey Jen it it
39:09 looked it was quite impressive because
39:11 there's a video out of a thing called
39:16 what was it called
39:18 um
39:21 let's see
39:23 um
39:26 um AI
39:28 dub
39:30 video
39:36 um sync
39:40 not the pro
39:43 what was it called
39:51 I'm gonna find it in x
39:57 [Music]
40:04 oh Twitter X
40:08 um Twitter's updating their terms of
40:10 service now normally a I wouldn't care
40:13 about this because I've never read a
40:15 terms of service in my life but
40:17 she's up five love in the first set
40:19 sevenenko's just [ __ ] melted down
40:21 apparently
40:23 um
40:25 but now we have things like Chachi BT
40:29 and Claude that can read the terms of
40:31 service for us so
40:34 let's go figure out what the hell
40:35 they're doing with all our data people
40:42 oh this was effective May 18th there's
40:45 nothing new here why did that just pop
40:48 up
40:56 huh I wonder if they just didn't have to
40:58 update the date let's go to Claude so
41:01 here's another big document we'll pop in
41:03 Claude and see if it works
41:06 [Laughter]
41:14 three messages remaining until 1am
41:17 yeah they definitely are rate limiting
41:19 this pretty good so it's 80k
41:22 um
41:23 please give me a bulleted
41:30 list
41:33 of the 10 most important
41:38 clauses
41:41 here
41:44 limit
41:46 bullets
41:48 to 15 words
41:52 or bellets
41:57 um
41:58 [Music]
42:06 yeah I have a feeling Ted you're right I
42:08 have a feeling I have a feeling oh
42:10 claude's gonna get shitty for us
42:12 non-payers now that now that they have a
42:14 paid thing okay Twitter you must be 13.
42:19 we can collect and use your personal
42:21 data okay that's pretty clear you own
42:25 content you post and grant us the
42:27 license to distribute follow all the
42:28 rules and laws when using it we can
42:31 remove content or suspend accounts
42:32 violating rules we can place ads and
42:35 access your data to operate the service
42:37 your use of the services at your own
42:39 risk our liability is limited to a
42:42 hundred dollars
42:45 we can revise terms at any time
42:47 continued use equals acceptance
42:50 basically if you use it you accept
42:53 whatever terms they give you so so if
42:55 they just change them to say you owe us
42:58 all of your salary but and you're using
43:00 the service then you've agreed to it
43:03 and it's it's Cali governing laws
43:05 California for non-eu users
43:09 non-eu
43:12 does that mean America no I suppose it
43:14 means non-eu oh because eu's got their
43:16 super stringent [ __ ] on their own all
43:19 right whatever
43:20 that was a snore
43:22 you see
43:25 that's why I don't that's why I don't
43:28 read those things
43:31 janky
43:33 um
43:35 [Music]
43:42 all right
43:45 I can't remember handshake for everyone
43:47 I need one
43:51 virtual you can just do like that's my
43:54 handshake
43:56 my hand is shaking
44:00 oh
44:03 Claude logged out back in all good yeah
44:05 okay that was a caching issue you had
44:08 something there was something weird in
44:09 there
44:10 its memory just burped actually you know
44:12 what it could have been is they you know
44:14 they probably did some sort of server up
44:16 upgrade update and it
44:18 something was different on the back end
44:20 versus your your cast version in your
44:23 browser or something like that some
44:25 wacky ass programmer [ __ ]
44:27 it's all about being nice be a decent
44:29 human exactly it's not that hard
44:33 just don't be an [ __ ]
44:37 have empathy everyone's going through
44:39 [ __ ]
44:41 as much of a dick as they are there's
44:43 something going on so just give them
44:45 Grace chill out take a breath
44:48 and then at some point
44:52 it'll all be cool
44:53 but probably not for a while
44:56 because we've screwed up the country all
45:00 right what are your thoughts on open
45:02 interpreter do you mean code interpreter
45:05 wait is open interpreter like oh that's
45:07 that new thing that's the new code
45:09 interpreter
45:10 uh interpreter I haven't played with it
45:13 yet meet open interpreter yeah an open
45:17 source locally running code interpreter
45:19 locally running implementation of
45:23 opening eyes code interpreter by
45:25 naharika Singh September 6th
45:29 in the Everett in the ever evolving
45:31 realm of programming and software
45:33 development the demand for efficient
45:35 flexible and accessible tools has never
45:37 been higher
45:42 I can't say his name I could say it when
45:45 I wasn't saying it was a stupid accent
45:46 okay
45:50 does he have it hosted anywhere
45:53 open and collab
45:57 [Music]
46:00 oh this is funny let's look at this so
46:04 if you haven't seen Google collab before
46:06 I'll show it to you here's what it is
46:09 um
46:11 we're gonna allow that
46:13 connect let me see if I've still paid my
46:16 bill connecting so you got to connect so
46:18 what we're doing here initializing Ram
46:20 disk okay so what this is is this is
46:24 some sort of virtual machine in the
46:25 cloud and I know you Geeks are probably
46:27 like well Google collabs technically not
46:29 a virtual machine it's an instance of it
46:32 doesn't matter there's a computer
46:34 somewhere running some [ __ ] okay and
46:36 it's got RAM and disk so so as far as
46:38 I'm concerned it's a virtual machine I
46:41 don't care I don't care about the
46:42 technical details you can run [ __ ] on it
46:45 and then what a go what a Google collab
46:48 notebook is is someone has been kind
46:50 enough to write a bunch of code python
46:52 code I think and just put it in these
46:55 little cells with play buttons and then
46:57 so like this says paste your open AI key
47:00 below so I would paste my open AI key
47:02 there
47:03 and but notice what they did they sort
47:06 of copied the open ai's little graphic
47:08 look and feel so and they're toying
47:12 around with uh they're toying around
47:14 with trademark and copyright stuff here
47:17 um
47:18 let's see
47:22 um
47:26 hello world math
47:30 yeah this thing looks pretty cool
47:32 um I haven't played with it yet so so
47:35 basically what you do though
47:37 to run these things is you just
47:38 basically keep hitting the play buttons
47:41 and it'll it'll sort of load this piece
47:43 of code and then you put your API key in
47:45 here and you hit play and it'll load
47:47 that and so you're you're loading a
47:49 sequence of pieces of code and then when
47:52 you get to the end of it you've got an
47:53 application running you can either open
47:55 a web browser and see it or you can
47:57 execute things within the notebook that
47:59 all of all of them are different but
48:01 it's basically like a super convenient
48:03 way for
48:05 people to share little setups that they
48:08 have setting up things like
48:10 um automatic 1111 for stable diffusion
48:14 or in this case open interpreter so what
48:17 my thoughts on it are uh
48:20 wait DV deviator deviator oh yeah
48:24 deviator I like it 10 out of 10. it made
48:27 me think it's solid it makes me smile 10
48:30 out of 10 on the name
48:31 um I think it's I I think this is cool
48:33 and I think we're going to continue to
48:35 see open source projects mimic
48:38 um things in the real world and it's
48:40 it's a good thing
48:43 because the less we have big companies
48:46 you know consolidating power they're
48:48 going to do that [ __ ] anyway so like
48:49 over the next five years they're all
48:51 going to be fighting to subsume one
48:53 another so they're more of a solid
48:55 foundation we can get in the open source
48:57 Community the better off it is for LS
48:59 chickens
49:03 I don't know if it's any good
49:06 because I haven't used it after seeing
49:07 your YouTube life video I understand
49:10 what our share shared knowledge
49:13 communities values cool fantastic
49:16 thoughts on open interpreter we just did
49:18 that I like Claude the best I don't pay
49:20 for chat EBT that's interesting Vicky
49:22 I'm I'm still paying for chat GPT but
49:25 you know it's funny
49:28 they took away internet access and they
49:30 haven't returned it and
49:33 um
49:34 these other models are getting better
49:36 and better all around them it's kind of
49:38 like gpt4 was significantly above
49:41 everything else
49:42 and then these other models are just
49:44 getting better and better they're
49:45 creeping up to it and like Claude and
49:49 and some others are
49:52 consistently solid
49:55 and I feel like chat I feel like open AI
49:58 is
49:59 um distracted with
50:02 business deals
50:04 for lack of a better term because they
50:07 have they had a giant Target painted on
50:09 their face and another one on their back
50:11 and another one on their ass and then
50:12 they had a Target on each kneecap and
50:15 then they had Targets on their elbows
50:16 and on each knuckle is a target like
50:20 any company that kind of comes out of
50:22 the gate like open AI did when they
50:24 launched chat CPT I know they've been
50:26 around for years before that they were
50:28 decently visible but when something like
50:30 chat gbt hits as as hard as that hit got
50:34 to 100 million users in six weeks
50:37 [ __ ] everyone's coming after them it
50:39 was just you know it's taken the better
50:40 part of a year for people to start
50:42 catching up but you know they're they're
50:45 probably spending way more in lawyers
50:47 than they were anticipating as my guess
50:51 so yeah so it's really interesting it's
50:53 I I'm wondering if I'm going to be
50:55 paying for Chachi EBT much longer I
50:56 think I will because I
50:58 I can't imagine they're gonna they're
51:01 they're gonna let people pass them but
51:04 you know this is this is all about
51:06 execution
51:08 Kyle tries to go through each comment in
51:10 order so it's usually not necessary to
51:11 repeat your question thanks Dr J
51:15 DD
51:19 Saucy llms
51:21 nope it was awesome I'm older than you I
51:25 don't know what that was in oh Camelot
51:28 Camelot
51:31 um you know you may not know this show
51:33 use chat GPT several times today good
51:35 info produced
51:36 can't get the bugs off my exhibition
51:41 you mean physical bugs
51:45 you have an art exhibition where are you
51:47 are you up in like the mountains in the
51:49 mountains okay farmer
51:51 you know Tobias is in the hot tub right
51:53 you know he you know he dropped his
51:55 phone in the hot tub right you know that
51:57 he can't be trusted to be on his own
52:00 what are you doing
52:08 thank you oh I have invested in the
52:12 golden button what's the golden button
52:14 you ask you could make money
52:18 how do you make money with your DVD Kyle
52:21 tell us please there's a secret right
52:24 isn't there a secret tell us there's a
52:26 secret you've got the secret I'll pay
52:28 you for the secret
52:29 I'll only charge you a 19.95 an hour for
52:32 the secret not the secret the secret
52:35 secret no that that's one you just tell
52:38 the world what you want the universe
52:39 what you want you get it that's that's a
52:41 different secret my Secret's an actual
52:42 secret you have to pay for 19.95 a
52:44 minute night night one two 200 299 yeah
52:50 yeah yeah that's a ticket yeah see yeah
52:55 all right
52:56 Microsoft says AI tools will take
52:59 them to 10 billion faster than any other
53:03 Biz they've launched at the Goldman
53:04 conference that's interesting
53:07 so
53:08 yeah fascinating
53:10 um
53:12 yeah a couple of weeks ago people were
53:14 talking about the demise of open AI
53:16 because they were spending seven hundred
53:17 thousand dollars a day
53:19 but now they're on a billion dollar run
53:22 rate so they're making about
53:24 what is that
53:26 um if if seven hundred thousand dollars
53:29 a day is
53:31 um like 250 million a year
53:34 so if they're making a billion then
53:36 seven that's
53:38 1.4 2.8 they're making almost three
53:41 million dollars a day and it's costing
53:43 them 700 000 a day
53:45 I'd take that
53:47 I'd take that math
53:53 um that's really interesting I just
53:54 realized I can join on my laptop low
53:56 where have I been oh yeah you can
53:57 exactly you can join these on your
53:59 laptop and that's why I was saying
54:01 um one of the things when I when I do
54:04 when I show these things wait not those
54:07 when I show these things and people want
54:09 to screenshot them and you can't
54:10 screenshot them on a Samsung you can
54:12 probably screenshot them on your on your
54:14 laptop
54:19 jump into the bottom
54:22 Tick Tock jump me to the bottom and I
54:25 hate the tick tock
54:28 got a lot
54:34 all right I'm back to where I was
54:38 I'm so happy with all the good
54:39 information
54:41 out there and also here
54:44 oh thank you Winston I appreciate that
54:47 Joker's in hey everyone Mercury is in
54:50 microwave exactly that's good I'm gonna
54:53 have to remember that one
54:58 oh good lord
55:00 Mike Rowe is active now I don't know if
55:02 he's in here but if you are what's up
55:04 dude sorry Kyle I didn't get to hear
55:06 your answer about if you got moderators
55:09 um I have moderator volunteers I just
55:11 haven't activated them yet because it
55:13 requires me to actually think about
55:15 stuff and learn how to do that so
55:18 at some point I'll get annoyed enough
55:21 with not having them that it'll push me
55:23 over the add edge and I'll make it
55:25 happen
55:26 but I do appreciate everyone who has
55:28 volunteered and I I will not um I will
55:31 take you up on it I promise
55:34 um he has a list but not activated it
55:36 yet correct my new density over
55:40 subscribing to everything
55:43 and every service may be a potential uh
55:46 casualty yeah
55:48 there's so much you could I just stopped
55:51 like I was sort of three months into
55:53 this thing I'm like I gotta subscribe to
55:54 that it's like when you get into nfts
55:56 you're like I'm gonna buy that Nifty I'm
55:57 gonna buy that one I'm gonna buy that
55:58 like where's my money and they're like
56:00 wait how do I sell these things oh you
56:01 can't sell them they're worth nothing no
56:03 oh but I gotta buy more these are gonna
56:05 be the ones that are gonna be worth it
56:07 it's like that with these AI tools it's
56:10 ridiculous
56:14 bubbles gotcha I'm not weird I'm
56:16 irregular
56:18 that's see now you're getting it you've
56:21 got the secret handshake bubbles and by
56:24 the way you're like 90 there with the
56:26 name bubbles right you're 10 out of 10
56:27 on the name so you're essentially a
56:30 regular from the get-go
56:32 Vivek Buca irregular gang
56:35 there are so many irregular shout out to
56:37 my irregular numbers by the way you're
56:40 up to 17 000 likes your regulars you
56:42 dang weirdos with your carpal tunnel
56:44 screen tapping nonsense
56:47 22 7 moment I don't know what that means
56:50 but I think it's funny
56:51 oh 22 no I thought that was likes I
56:54 don't know what that is hi friends
56:57 hey Jen beta is live too
57:01 but hey Jen beta for just like normal
57:05 wait
57:06 um
57:13 beta
57:18 um
57:18 [Music]
57:20 I signed up with this walking around
57:22 Costco I was at Costco today that's like
57:24 twice in one week that's that should be
57:26 against the law
57:28 try hey gen for free sign in
57:31 DT
57:40 [Music]
57:48 [Music]
57:50 business owner I'm a business owner
57:54 oh I don't feel like answering [ __ ] okay
57:57 never mind whatever Tobias are you
58:00 tapping Tobias couldn't tap because he
58:02 was in the hot tub and then he dropped
58:04 his damn dumb phone in the hot tub
58:06 laughs
58:08 it's not good it's not good
58:12 it's not good okay farmer he's not okay
58:15 I we haven't heard from Tobias since he
58:18 dropped his phone uh
58:25 let's see
58:27 yes I'm lagging too oh well it's not too
58:30 bad for me oh that sucks how many
58:32 messages can you send now since the pro
58:34 version I don't know I don't know it's
58:40 my guess is that they're gonna drop
58:42 normies down to 10 10 or 15 messages
58:45 and if you send something long they'll
58:47 probably do like two or three
58:49 so it's gonna be annoying
58:52 oh you know what they're gonna do you
58:54 know what claude's gonna do those
58:56 [ __ ] [ __ ] the [ __ ] let
58:58 me tell you something let me tell you
59:00 something Claude with your little
59:02 anthropic name your little anthropic
59:05 company your ex open AI people we
59:08 invented gpt3 and now we got a better
59:10 way to make things you know what they're
59:12 gonna [ __ ] do
59:14 they say you pay 20 bucks a month and
59:16 you get five times as as many messages
59:19 when you pay the 20 bucks a month five
59:22 times what
59:24 I bet they're gonna cut the free thing
59:26 in into a quarter and then give you five
59:29 times
59:30 one-fourth of what you had before so
59:32 you're basically gonna pay 20 bucks to
59:34 get the same [ __ ] you had before
59:36 you [ __ ] I knew it
59:39 I knew it
59:42 yes scumbags
59:45 now listen
59:47 I just made that [ __ ] up it's total
59:49 speculation but I'm cynical
59:52 do you do you know when I was a kid
59:54 that that cereal boxes used to be like
59:59 as wide as your [ __ ] head when you
1:00:02 turned it sideways they were they were
1:00:03 Broad and you turned it sideways it was
1:00:06 as big as your head you turned a cereal
1:00:08 box sideways now it's like one corn
1:00:10 flake wide
1:00:12 you want to know why I'm bitter
1:00:15 cereal
1:00:18 and everyone's doing the same [ __ ]
1:00:20 thing they'll just make it a little bit
1:00:21 smaller make it a little bit smaller oh
1:00:23 you know what we could do now we'll make
1:00:25 a king-sized version that's the same
1:00:27 size as the original one was 20 years
1:00:30 ago but we'll charge five dollars for a
1:00:33 candy bar
1:00:34 they're all doing the same [ __ ]
1:00:37 it's disgusting [ __ ] capitalism
1:00:45 [ __ ] anthropic
1:00:46 see now they just got on my [ __ ] list
1:00:49 I think I I might move Claude to the
1:00:51 Kids Table just out of spite
1:00:52 [Laughter]
1:00:59 I'm gonna start I'm starting rumors here
1:01:01 tonight we're gonna take some companies
1:01:03 down
1:01:04 you give us our tokens back
1:01:07 [ __ ] anthropic
1:01:09 all right well well if we I figured from
1:01:12 a percentage basis if we if we uh if we
1:01:15 reduce the the amount of tokens for the
1:01:16 free tier two to uh approximately uh uh
1:01:19 20.32 of of what the of what the
1:01:22 standard is today and then uh we we
1:01:24 produce a product uh at approximately I
1:01:26 mean you could call it uh uh twenty
1:01:28 dollars uh per month per thirty day
1:01:30 period which on some days it's 31 days
1:01:32 uh but for for twenty dollars for the
1:01:34 for the thirty on average Thirty point
1:01:36 two days period uh we could then uh
1:01:39 claim that they get a five times the
1:01:41 amount of tokens as as the non-paying
1:01:43 version and everyone will assume it will
1:01:45 be the same and but but but in the end
1:01:47 we we get twenty dollars for exactly the
1:01:48 same thing we're giving away for free
1:01:49 now
1:01:51 so so that that would be my proposal
1:01:56 do you think you're [ __ ]
1:02:04 you're [ __ ] I just see now I'm upset
1:02:07 I need water
1:02:09 Madison Keys up 2-1
1:02:12 in the uh
1:02:15 in the second set
1:02:19 although
1:02:21 seven link has got an ad
1:02:28 all right
1:02:30 Turtles chapter not Raptor
1:02:33 anyway I I know it's a conspiracy theory
1:02:37 right and I know I'm throwing anthropic
1:02:39 under the bus and I know that was very
1:02:41 stereotypical computer engineer making
1:02:44 fun of them I'm sure there's an engineer
1:02:46 that's triggered right now not making
1:02:47 eye contact with me but listen
1:02:52 that's what they're [ __ ] doing
1:02:59 a manual car wash do you know where I
1:03:01 can find one yeah it's called a hose
1:03:03 okay farmer get a hose get a sponge
1:03:06 it was funny I used to I used to bartend
1:03:08 in in downtown Manhattan and there were
1:03:10 these two guys that came in these these
1:03:12 it was two black guys a young guy named
1:03:14 Adam and the older guy was his name was
1:03:17 Willie
1:03:19 and Willie would come in and so so they
1:03:22 were they were both alcoholics and
1:03:26 um
1:03:27 Adam was so much of an alcoholic that he
1:03:30 had ulcers and so he would come in and
1:03:32 he would he would order a scotch and
1:03:33 milk he'd go hey Kyle scotch and milk
1:03:36 inside making a scotch and milk and then
1:03:39 Willie would come in and he'd go I want
1:03:42 a beer and he go a little beer and then
1:03:45 so I had these little like four ounce
1:03:47 glasses and I was pouring a tap of four
1:03:50 ounces of beer and they would sit there
1:03:52 and they would milk those and then
1:03:53 Willie would would ask for like a pint a
1:03:56 pint glass like a regular beer glass
1:03:59 and then I put water in it it goes put a
1:04:02 couple of drops of soap in that and he
1:04:04 would go out he would wash his entire
1:04:06 car with just the water in a single pint
1:04:10 glass he took great pride in that so so
1:04:14 okay farmer so you can
1:04:17 wash your car with a pint of water I
1:04:20 have seen it happen
1:04:21 in in New York City
1:04:24 of all places
1:04:26 one shot GPT requires as much energy as
1:04:29 boiling one liter of H2O that probably
1:04:33 that that doesn't surprise me
1:04:37 um
1:04:38 you know but what do you want you we get
1:04:40 robots that that are gonna you know do
1:04:43 do all of our work for us the Energy's
1:04:45 got to go somewhere
1:04:48 what are your thoughts on Long Shot AI I
1:04:51 don't know it let's go look
1:04:53 um
1:04:54 [Music]
1:05:01 oh generative AI for fact check blogs
1:05:08 think chat GPT for teams and Enterprises
1:05:12 research generate optimized content
1:05:14 upload your docs fact check ensure zero
1:05:18 hallucination integrate anywhere
1:05:22 um
1:05:27 semantic SEO fact GPT
1:05:37 um
1:05:39 I like the promise I have no idea how
1:05:42 they're [ __ ] doing I'm talking to a
1:05:44 developer right now
1:05:45 that's part of the AI salon
1:05:48 about doing a project for me and he's
1:05:50 he's figuring this out right now for a
1:05:52 bank
1:05:54 [Applause]
1:05:56 um
1:05:58 I'll sign up for it I'll check this out
1:06:01 this looks interesting so so for those
1:06:03 of you that are new to this [ __ ] um
1:06:06 the the the chat GPT things of the world
1:06:09 they do this thing called hallucinating
1:06:11 which basically just lying it will lie
1:06:13 boldly to your face it'll go here's your
1:06:16 answer and you're like um I'm pretty
1:06:17 sure that's not right oh I'm sorry
1:06:19 you're right that's not right here's the
1:06:21 right answer and then you look at that
1:06:22 and you're like no no that's not right
1:06:23 either it's like oh you know you're
1:06:25 absolutely right here's the answer and
1:06:27 if you get it in the wrong Loop it will
1:06:29 never get the answer right that doesn't
1:06:31 happen very often but it does happen so
1:06:33 this is this is interesting I'll check
1:06:34 it out um I don't know anything about it
1:06:36 and I'd be curious to know how they're
1:06:39 ensure zero hallucinations uh maybe if
1:06:43 you're just doing a semantic search I
1:06:45 don't know how they're guaranteeing zero
1:06:46 zero hallucinations
1:06:48 unless they're unless they're saying
1:06:50 it's zero hallucinations when you're
1:06:52 training a when when you're gpting a
1:06:56 document basically and then you're
1:06:58 you're interacting with that document
1:06:59 that I could see
1:07:01 but I don't know that's interesting
1:07:04 very interesting I like it I like it
1:07:07 I've learned some new things here today
1:07:08 oh you know what I'm gonna do
1:07:10 I'm going to drag that into my little
1:07:12 folder
1:07:13 and then didn't I see something else
1:07:15 didn't someone else tell me about
1:07:16 something else oh hey Jen I knew about
1:07:18 that though
1:07:19 open open interpreter
1:07:22 that was the one
1:07:24 all right
1:07:28 AI learning lab Up Up and Away in my
1:07:33 beautiful my beautiful balloon remember
1:07:37 that as kids that was one of those songs
1:07:40 that just made me
1:07:42 I hated that song and it played all the
1:07:45 [ __ ] time
1:07:47 for like a decade
1:07:51 smooth Rock smooth jazz easy listening
1:07:54 Up Up and Away in my beautiful we just
1:07:58 here your shot stop it
1:08:02 [Applause]
1:08:07 so yeah down east affiliate sorry I
1:08:09 don't have a better answer for you but
1:08:10 it looks pretty cool
1:08:12 uh Franklin hey bro what's shaking
1:08:13 pumpkin spice Cheerios
1:08:16 yeah pumpkin's face Cheerios the box is
1:08:19 like it's it's just just [ __ ] awful
1:08:22 is that why one of the local Bots is
1:08:24 called chat H2O
1:08:26 [Laughter]
1:08:28 uh what do you recommend wait
1:08:31 question what do you recommend to see
1:08:33 what type of AI tools your Biz will need
1:08:37 huh
1:08:40 that's interesting
1:08:43 um
1:08:44 I don't know that there's a service like
1:08:46 that you could you can you can sort of
1:08:48 roll your own at futurepedia right or or
1:08:53 you know uh services like future pedia
1:08:57 there's lots of directories out there
1:08:58 with with AI Tools in them I don't know
1:09:01 that there's
1:09:03 it would be interesting to do a
1:09:05 directory like future pedia where you
1:09:07 basically answer questions about your
1:09:09 business and then
1:09:11 the the the directory recommends a suite
1:09:14 of tools
1:09:16 but I think it's so specific
1:09:21 Siri
1:09:24 hey
1:09:26 stop it
1:09:32 um
1:09:33 that would be interesting but but it's I
1:09:35 mean it's so dependent like do you have
1:09:36 a small business do you have
1:09:38 a medium business do you have a large
1:09:40 business are you in a pharmaceutical
1:09:42 company do you work with pharmaceutical
1:09:44 companies I've got a very small business
1:09:45 but most of our clients are major Pharma
1:09:47 companies so so I've got a suite of
1:09:51 tools that I can use for my [ __ ] it's
1:09:53 different than the suite of tools I can
1:09:54 use for their [ __ ]
1:09:56 um
1:09:57 so I I don't know that there's you know
1:10:02 quite quite frankly Franklin didn't mean
1:10:05 to do that but
1:10:06 quite frankly Franklin Franklin
1:10:09 my dear I don't give a damn um
1:10:14 um
1:10:15 um what was I saying before I got stupid
1:10:19 um uh
1:10:22 the tools are changing so fast and
1:10:25 they're so so
1:10:26 I kind of jokingly put up this image
1:10:28 earlier
1:10:30 that one
1:10:32 um
1:10:33 a lot of these generative AI tools are
1:10:36 janky pieces of [ __ ] right now so things
1:10:39 are changing so fast that that I think
1:10:42 the most important thing to do for your
1:10:44 business is just do what you can to get
1:10:46 AI literate and what does AI literate
1:10:48 mean it basically means in in my humble
1:10:51 opinion
1:10:53 it means start playing with
1:10:56 you know Chachi PT or Claude or Bard and
1:10:59 you know Bing any of these start playing
1:11:01 with them regularly and don't
1:11:04 play with them is kind of the right word
1:11:05 get curious about them
1:11:07 do anything on them do personal projects
1:11:10 do [ __ ] at home do stuff for your kids
1:11:12 do stuff for the wife for the church
1:11:14 group for the community center it
1:11:15 doesn't [ __ ] matter just just like
1:11:17 start using them to solve problems to
1:11:20 write marketing copies
1:11:23 and in doing that you'll kind of figure
1:11:24 out where the boundaries of them are and
1:11:26 then you can start to say okay let me
1:11:29 now that I'm kind of getting what these
1:11:31 things are good at what they're not good
1:11:32 at where they hallucinate where you know
1:11:34 how I can sort of navigate through that
1:11:36 then you can start to say okay let me
1:11:38 take the big problems in my business and
1:11:41 figure out if there's some way I can use
1:11:43 these tools
1:11:44 to at least figure out how I would
1:11:48 attack that problem with these tools
1:11:50 right or with some combination of these
1:11:52 tools and some other tools
1:11:55 um
1:11:57 and just just start kind of learning it
1:11:59 from the inside out rather than saying
1:12:01 oh no thank you very much thank you
1:12:03 Winston rather rather than saying oh
1:12:05 there's a perfect Suite of tools out
1:12:07 there there's not why because almost all
1:12:10 of the tools right now are
1:12:13 janky pieces of [ __ ] thank you took us
1:12:17 um and they're not janky pieces of [ __ ]
1:12:19 because the companies are bad they're
1:12:20 janky pieces of [ __ ] because the
1:12:22 industry is so early
1:12:23 chat GPT is only nine months old gpt3
1:12:26 which was kind of the best model out
1:12:28 there came out in 2020. this is all this
1:12:30 is all real new stuff so you know you've
1:12:33 you've got
1:12:34 um
1:12:36 um uh copy Ai and and the other one that
1:12:40 does it's that's like copy AI I can't
1:12:42 remember it right now but doesn't matter
1:12:43 but but you have these things that were
1:12:45 built on top of gpt3 and now they're
1:12:46 built on top of gpt4 and they're all
1:12:49 these tools to help you run your
1:12:50 business but all of them are are just
1:12:54 they're just being born
1:12:56 so if you can figure out how these
1:12:59 things work and and and how to get your
1:13:02 head around them then as the tools get
1:13:05 more sophisticated and things start to
1:13:07 stratify into okay here's real
1:13:09 applications here's medium applications
1:13:11 that have a shot oh here's just janky
1:13:13 pieces of [ __ ] right now they're all
1:13:15 kind of in the janky pieces of [ __ ] and
1:13:17 some will start to emerge out of that
1:13:19 but you won't be able to recognize them
1:13:22 emerging out of the the soup unless
1:13:24 you're kind of in the soup playing so
1:13:26 that that would be my recommendation is
1:13:28 don't think that there's a really clear
1:13:29 answer out there four to two keys over
1:13:32 sabalanka
1:13:34 second set could be double Americans in
1:13:37 the finals that's crazy
1:13:39 open interpreter is super easy to load
1:13:42 been playing with it all day oh that's
1:13:44 super cool Nicholas Massey that's great
1:13:45 thank you for sharing that I'm gonna I'm
1:13:47 gonna play with it later
1:13:49 I want to build my own AI without using
1:13:52 open AI or hugging face how can I do
1:13:54 that
1:13:55 um I don't know well you're gonna have
1:13:58 to hugging face is has got lots and lots
1:14:02 of models in it I I would say it's going
1:14:05 to be llama 2 or code llama
1:14:09 um and then you know what you could do
1:14:11 is there's a project out there called
1:14:15 GPT for all
1:14:17 which is it thank you for the Llama I
1:14:20 think Winston is very clever I don't
1:14:22 know if you all have noticed this you
1:14:23 other Irregulars but Winston's got some
1:14:26 [ __ ] Emoji game going on or whatever
1:14:28 the hell those sticker things are called
1:14:30 every time I mention llama 2 a [ __ ]
1:14:33 llama comes up on screen he's just
1:14:35 sitting there waiting night after night
1:14:37 he's got the Llama like [ __ ] loaded
1:14:39 up like a like a spring
1:14:42 it's like
1:14:45 he's good he's good you other regulars
1:14:47 I'm not saying it's a competition I'm
1:14:49 just saying Winston's in the lead
1:14:52 um
1:14:54 so try GPT for all
1:14:57 um and meat
1:14:58 um
1:14:59 it's there's a there's an installer for
1:15:02 OS X there's an installer for Windows
1:15:04 and there's an installer for Linux I
1:15:05 think
1:15:06 and you just install it and then it
1:15:09 basically creates a little runtime it's
1:15:11 a little application that runs on your
1:15:13 machine and you can put different models
1:15:15 inside it and then you can build [ __ ] on
1:15:17 top of that
1:15:18 um other than that you're gonna have to
1:15:20 go to GitHub or hugging face and find
1:15:22 like little
1:15:24 you know like I would I would so what
1:15:26 what I would honestly do Amit is is
1:15:29 um
1:15:30 I'd recommend you go to Reddit and go in
1:15:33 the subreddits just search for open
1:15:35 source large open source llm
1:15:38 in Reddit and find some subreddits where
1:15:42 people are talking about all these open
1:15:43 source llms you'll find like Google
1:15:45 collab notebooks in there where people
1:15:47 have put together notebooks where you
1:15:48 can install different models on your
1:15:50 local devices
1:15:52 um so that it's it's all out there for
1:15:54 you to find it's just you got to get a
1:15:56 bit geeky GP GPT for all is good because
1:16:00 ah thank you very much Winston GPT for
1:16:03 all is good because you just install
1:16:04 this thing and then you swap out the
1:16:06 different models so
1:16:12 Madison Keys 12-3 in Majors this year
1:16:15 she's got the game face on she's ready
1:16:17 to take this thing down
1:16:19 all right I was just listening to the
1:16:21 dev and the dev for Orca chat and VC
1:16:24 about their plans for expanding oh
1:16:25 interesting oh that's super cool
1:16:28 for for what for uh
1:16:30 for for open interpreter
1:16:33 I mean honest to God if you guys haven't
1:16:35 seen code interpreter so open
1:16:37 interpreters the open version of that
1:16:40 if you haven't seen cooking and I know
1:16:42 some of you folks have seen this before
1:16:44 in here but tough not everyone has man
1:16:48 so we're gonna go to chat gbt
1:16:51 and I don't know if you all knew this
1:16:52 but uh
1:16:55 yeah you can all right
1:16:59 so code interpreter it's not called code
1:17:02 interpreter anymore you know why because
1:17:04 oh Winston Make It Rain
1:17:11 bringing cinnamon to the main stage
1:17:13 cinnamon main stage
1:17:16 never been to this strip club why would
1:17:18 you ask
1:17:19 um
1:17:21 they changed the name of code
1:17:23 interpreter to advance data analysis why
1:17:27 because in the AI learning lab we told
1:17:30 them that code interpreter was a shitty
1:17:32 name
1:17:33 we said it when it came out we said and
1:17:35 then I said it was a shitty name and
1:17:38 some geek in here was like well
1:17:40 technically Mr Shannon it's an accurate
1:17:42 name because it actually is code is in
1:17:45 the in the world of computer science so
1:17:47 there's a thing called code
1:17:48 interpretation and technically that's
1:17:50 exactly what I'm talking about
1:17:52 I said okay yeah but it's still a shitty
1:17:55 name and Sam Altman was in here that
1:17:57 night
1:17:58 and he sent me an email after I said
1:18:00 that was a shitty name it was like Kyle
1:18:02 dude man thanks can I be an irregular
1:18:05 I'm like no Sam you can't be an
1:18:07 irregular not until you change the name
1:18:09 of that shitty project that remarkable
1:18:11 project with a shitty name and Sam was
1:18:14 like uh I'll try to win back big on your
1:18:18 good graces soon tomorrow
1:18:21 I felt bad for him but you know whatever
1:18:24 you gotta you gotta tell these CEOs you
1:18:26 got to put them in their place
1:18:27 so I put Sam in his place and he changed
1:18:29 the name so now it's called Advanced
1:18:30 Data analysis but it's [ __ ]
1:18:32 remarkable okay
1:18:35 um
1:18:36 so if you haven't seen this before so
1:18:38 you know you know chat GPT you chat with
1:18:40 it right well Advanced code analysis you
1:18:44 can upload [ __ ]
1:18:46 and I'm gonna go grab a data file
1:18:50 we're gonna grab data archive five I
1:18:54 don't know what's in here
1:18:55 so so there's if you want to find data
1:18:58 files to play with to just experiment
1:19:00 with this thing there's a site called
1:19:01 kaggle
1:19:03 k-a-g-g-l-e go to kaggle
1:19:06 and then um
1:19:08 so I'm gonna say
1:19:12 please tell me what's inside this file
1:19:18 and what is interesting about the data
1:19:26 so
1:19:28 so you upload a zip file or or a CSV or
1:19:32 multiple csvs whatever it might be data
1:19:36 and then
1:19:37 it
1:19:39 I said yesterday now I've been saying
1:19:41 for about a month that it spins up a
1:19:43 virtual machine and someone told me
1:19:44 yesterday it doesn't spin up a virtual
1:19:46 machine Kyle it's just a container
1:19:49 so it doesn't spin up a virtual machine
1:19:51 but it basically spins up a virtual
1:19:53 machine that's called a container okay
1:19:56 and then it
1:19:58 within that container it's got access to
1:20:01 python libraries and it it writes its
1:20:03 own Python and it executes its own
1:20:06 python so so you just tell it what to do
1:20:10 like open that file and tell me what's
1:20:12 in it so it says this is this is crime
1:20:14 data so there's a crime CSV a crime XML
1:20:17 and an offense code CSV would you like
1:20:20 it says to understand and analyze the
1:20:22 contents I'll start by inspecting the
1:20:24 first few few rows of the CSV files
1:20:26 additionally I'll display a snippet of
1:20:28 the XML to get a sense of the structure
1:20:30 would you like me to proceed with this
1:20:32 plan yes so so I'm not having to write
1:20:35 this [ __ ] code I'm not even having to
1:20:38 open the CSV files I'm just telling it
1:20:41 to do it and it's doing it
1:20:48 it appears there's an issue due to an
1:20:50 encoding error so so this is another
1:20:52 cool thing so so it writes this code
1:20:55 to open the CSV file and then it finds
1:20:58 out that there's there's there's some
1:21:00 error it it doesn't like how it you know
1:21:03 it couldn't read it used the wrong
1:21:05 encoding Library so so now it's fixing
1:21:08 the error of its ways it's it wrote new
1:21:10 code
1:21:11 it says great the files were
1:21:13 successfully read using ISO 8859-1 so
1:21:17 like that little that little thing right
1:21:19 there where you get a data file and it
1:21:23 wasn't encoded in Ufa you utf-8
1:21:27 and so you have to go get some other
1:21:29 library to read it right if you ran into
1:21:31 that problem
1:21:33 with yourself like unless you're sort of
1:21:35 deep experienced that could be an hour
1:21:37 of you googling or even 15 minutes of
1:21:40 like oh crap what's the what's the iso
1:21:43 standard I need to get with the bet that
1:21:45 that let me go to Google and I gotta get
1:21:48 this and this what is it called an oh
1:21:50 it's ISO 58 but where's the code and how
1:21:53 did it
1:21:53 it just does it
1:21:56 so then it says okay here's all the
1:21:58 here's all the data inside there there's
1:22:00 all the offense codes and categories the
1:22:04 crimex XML
1:22:08 there are spatial boundaries
1:22:11 so it's got Vector Digital Data so so it
1:22:15 looks like it's got
1:22:17 Geo the geodetic datum is Ws
1:22:22 wgs1984 so it's got Latin longitude
1:22:25 resolution so this is basically
1:22:27 locations of crimes I assume given this
1:22:31 so the XML essentially provides metadata
1:22:34 or descriptive information about the
1:22:36 crime data set such as this format
1:22:38 constraints spatial details and more
1:22:40 interesting given this information is
1:22:43 there anything specific you'd like to
1:22:45 explore
1:22:47 I want you to determine the most
1:22:49 interesting Trends within the data and
1:22:52 visualize that
1:22:54 in a stunningly beautiful way
1:22:58 and so it's going to now
1:23:01 figure out whatever the hell I just said
1:23:03 means
1:23:05 which it like what I said meant nothing
1:23:07 like imagine if you actually knew
1:23:09 something about data analysis and we're
1:23:10 using this tool
1:23:12 um
1:23:13 certainly let's break this task into a
1:23:15 few steps so again
1:23:17 notice this is the machine prompting
1:23:20 itself on how to accomplish my goal
1:23:23 which is kind of on purpose a shitty
1:23:25 goal
1:23:27 data exploration we'll start by
1:23:28 examining the crime CSV data set feature
1:23:31 engineering based on the identified
1:23:33 trends
1:23:34 we might need to extract or transform
1:23:36 certain features to eight individual
1:23:38 visualization the third step is
1:23:40 visualization so let's start with data
1:23:42 exploration
1:23:44 first we'll look at the distribution and
1:23:45 some key columns so it looks at victim
1:23:47 count neighborhood ID first occurrence
1:23:50 date and offense category
1:23:53 and now it's doing something what's it
1:23:54 doing in here it's converting first
1:23:57 occurrence into date time so it's
1:23:59 changing the format of things so this
1:24:00 will clean clean up your data if you've
1:24:02 got shitty you know unstructured
1:24:05 horrible data it'll clean it up and fix
1:24:07 that [ __ ]
1:24:09 based on the exploration here are some
1:24:12 initial findings crime categories
1:24:17 the most common types of crime are theft
1:24:20 from a motor vehicle public disorder
1:24:23 all other crimes and larceny the least
1:24:25 common are arson and murder
1:24:28 so it's found some cool [ __ ] considering
1:24:31 the above findings we can visualize the
1:24:32 following a bar chart for the
1:24:35 distribution of crime categories a line
1:24:37 chart showing yearly trends of crimes a
1:24:39 bar chart for the top neighborhoods a
1:24:41 histogram or pie chart of distribution
1:24:44 of victim counts
1:24:47 Madison Keys serving for the match 5-4
1:24:57 wow oh look prettiness it made them
1:25:00 pretty look look at the pretty colors
1:25:03 I mean come on how cool is that
1:25:06 there's the line distribution of man
1:25:09 2021 was a shitty year for whatever that
1:25:12 crime was
1:25:14 trend of reported crimes
1:25:17 oh maybe the data ends that's why it
1:25:19 drops off so precipitously
1:25:22 this was probably a half year of data
1:25:24 and it just dropped off
1:25:28 look how pretty
1:25:30 so this is the top 10 neighborhoods with
1:25:32 most reported crimes Five Points Central
1:25:35 Park Capitol Hill
1:25:36 CBD Union Station
1:25:39 Montbello Civic Center
1:25:42 Gateway Green Valley Ranch Dia East
1:25:45 Colfax
1:25:48 cool is this
1:25:50 like it's still going
1:25:53 figuring [ __ ] out it's just it's just
1:25:56 having a good old time so now imagine
1:26:03 you know back to this idea of these
1:26:05 tools or janky pieces of [ __ ]
1:26:09 so what that's doing is
1:26:12 staggeringly remarkable if you've ever
1:26:14 done any sort of someone hands you a CSV
1:26:17 file and says figure out what's going on
1:26:19 in here like especially if you're not a
1:26:21 data analyst you're like I can do it I
1:26:25 can figure out a Excel formula the
1:26:29 Google pivot table oh this hurts
1:26:37 oh savalenka just broke her back five
1:26:39 five in the second set
1:26:42 um
1:26:44 so you look at what Advanced Data
1:26:47 analysis code interpreter is doing
1:26:51 and you're like that's very impressive
1:26:52 but also remember it's a janky piece of
1:26:55 [ __ ] right now
1:26:57 you know imagine what the sophisticated
1:26:59 version of that is where it doesn't make
1:27:02 mistakes it doesn't have memory errors
1:27:04 it knows way more than python it uses
1:27:06 whatever language it needs it's got
1:27:08 access to apis it's got approved apis it
1:27:12 might have API hooks into your personal
1:27:14 data you know on your servers
1:27:16 and so rather than uploading a few files
1:27:19 you could go into a file explorer
1:27:23 find the data that you need and start
1:27:24 running analysis on it or even better
1:27:29 hey
1:27:30 I want to be able to figure out this
1:27:32 about my business let me go look in all
1:27:35 of your data and figure out what might
1:27:36 be the data sets Ken Maynard thank you
1:27:39 very much what might be the data sets
1:27:43 um that are worth exploring and it goes
1:27:45 and finds all that [ __ ] and then it says
1:27:47 Hey to do a really sophisticated
1:27:49 analysis of this I'm going to hook over
1:27:51 to wolf from Alpha and we're going to do
1:27:54 some Advanced regression Tactical
1:27:58 spin them up spin them up I'm a jiggy
1:28:02 it's gonna do some real math over at
1:28:04 Wolfram Alpha and come back and give you
1:28:07 a prediction about how shitty your
1:28:09 business is
1:28:11 like
1:28:13 so so with like with all these tools
1:28:16 like start experimenting with them now
1:28:18 figure out what they're good at figure
1:28:19 out what they're not good at but
1:28:21 recognize like I'm I'm using that tool
1:28:25 for existing work like in my business
1:28:29 where I take you know specific pieces of
1:28:32 data
1:28:33 and rather than having to write
1:28:36 um
1:28:36 you know SQL queries or you know do
1:28:40 crazy pivot tables on on Excel sheets or
1:28:44 pay someone else to do that or pay my
1:28:45 developers to do that I can just ask it
1:28:48 questions and it does that [ __ ] for me
1:28:51 and that's you know it's kind of thing
1:28:53 where I wouldn't necessarily hire
1:28:54 someone new to do that work I'm getting
1:28:57 to do more sophisticated work in my
1:28:59 business because that tool
1:29:01 is able to do what it does
1:29:07 so
1:29:08 just nuts it's nuts it's nuts I tell you
1:29:17 I like a balloon I like the balloon
1:29:20 we're gonna keep the balloon for a while
1:29:22 we'll go back to other stuff but right
1:29:24 now I like to believe balloons is pretty
1:29:26 I like turtles
1:29:29 um Robert Frosty thank you for the donut
1:29:32 Donuts are tasty
1:29:35 guess what I bought at Costco because
1:29:38 I'm health conscious
1:29:43 like a [ __ ] 10 pound box of cinnamon
1:29:46 rolls because [ __ ] it I had a shitty day
1:29:49 yesterday got cinnamon rolls
1:29:52 [Laughter]
1:29:56 it's okay they're organic
1:29:59 all right
1:30:00 [ __ ] she said thanks for the pizza
1:30:04 you're an enabler Winston you're an
1:30:06 enabler
1:30:07 nine o'clock Kyle time OS chickens 9 42
1:30:10 now oh wow Dr J we are 42 minutes behind
1:30:14 oh I'm 99 Plus messages behind Okay so
1:30:17 if you're new here the 73 of you here if
1:30:19 you're new here
1:30:21 you might be thinking by jingus he
1:30:24 doesn't answer my questions ever what is
1:30:25 with him well
1:30:27 oh sabalinka up six five in the third
1:30:32 Madison Keys had a little meltdown
1:30:34 serving for the set
1:30:36 she's now serving to take it to a
1:30:38 tiebreaker holy crap
1:30:40 okay
1:30:42 um
1:30:44 was I gonna say oh if you so here's the
1:30:46 deal I'm not answering your questions
1:30:47 because you if you've been hanging out
1:30:50 here for longer than two minutes which
1:30:52 that's your that's on you not me
1:30:54 um you'll you'll notice that I do run
1:30:57 down some rabbit holes so what I'm going
1:30:59 to do now I'm going to jump all the way
1:31:00 to the bottom and then scroll up a
1:31:01 little bit so if if you ask questions a
1:31:04 while ago and I didn't get to them uh
1:31:06 just pop a new question in below
1:31:08 all right
1:31:10 and then let me go grab something oh my
1:31:12 God I'm a mess LinkedIn
1:31:15 uh
1:31:17 it was through Linkedin what was through
1:31:19 Linkedin oh you're having a whole
1:31:21 conversation Lonnie what's this
1:31:23 all right
1:31:25 all right this you guys are supporting
1:31:27 one another I support you supporting one
1:31:30 another
1:31:31 oh Madison Keys Falling Apart come on
1:31:35 Maddie
1:31:36 did you see that I like turtles kid 16
1:31:38 years later video on YouTube yeah I saw
1:31:40 that it's pretty funny it was a little
1:31:42 anticlimactic because he was kind of
1:31:44 normal like you know they they
1:31:48 didn't quite they didn't quite capture
1:31:51 what they should have captured with it
1:31:52 but whatever it's it's you know local
1:31:54 news
1:31:55 that you can't recreate that moment
1:31:58 the moment where the kid says I like
1:32:00 turtles and she tries to hold it
1:32:02 together there's like you can't act that
1:32:05 like you need to be an academy
1:32:07 award-winning
1:32:09 actor to
1:32:13 just kind of the subtle [ __ ] she does
1:32:15 trying not to lose it
1:32:18 let's see
1:32:24 [Laughter]
1:32:27 at the waterfront village with my friend
1:32:29 the zombie Jonathan you're looking good
1:32:32 Jonathan just got an awesome face paint
1:32:33 job what do you think I like turtles all
1:32:37 right you're great zombie and good times
1:32:40 here at the waterfront Village open for
1:32:42 the next 11 days her voice breaks she's
1:32:46 trying to string words together
1:32:49 you can't you can't act that [ __ ]
1:33:04 break point number two Madison Keys is
1:33:07 serving
1:33:08 down 30 to 40. this is break point
1:33:11 number two
1:33:13 oh yeah all right to Deuce all right
1:33:16 Madison
1:33:19 man she's
1:33:21 she's a beast she's a beast she's a
1:33:24 beast
1:33:27 precipitously grabbing dictionary like
1:33:30 precipitous there are precipitates
1:33:34 look oh I don't have any precipitates a
1:33:36 little bit precipitates on the outside
1:33:39 all of a sudden sounded a little dirty
1:33:41 didn't it
1:33:46 all right Maddie keys
1:33:49 for the set or no to take it to a tie
1:33:52 break
1:33:54 good serve
1:33:57 oh she got it
1:34:00 second set go into a tie break wow she
1:34:03 survived a double
1:34:06 break point set point
1:34:09 oh it's Vegas crime stats yeah that's
1:34:11 what that stats was
1:34:13 step into my office you're fired good
1:34:17 day to you sir good day and another
1:34:20 thing Johnson get your dick off my desk
1:34:23 that was a Playboy cartoon I remember
1:34:26 seeing when I was young never forgotten
1:34:28 that line it's [ __ ] hilarious
1:34:30 um I told you wait
1:34:32 I told you what I could do for ISO 9001
1:34:35 and older versions exactly yeah Kyle I'm
1:34:40 back do you know a great tool for
1:34:43 PowerPoint
1:34:44 um sort of
1:34:47 so
1:34:48 so Anne Rando
1:34:54 that one is actually a good one chat EPT
1:34:57 is good now if if you're wanting
1:34:59 something that actually lays out the
1:35:00 pages and stuff like that there are
1:35:01 tools that do it I'll show you I'll show
1:35:03 you one here
1:35:10 just going to chat and I'll just go to
1:35:13 3.5 I'll go to the free version
1:35:20 foreign
1:35:22 you're a Communications specialist
1:35:26 that excels at creating business
1:35:29 presentations period
1:35:32 can you give me three different
1:35:35 Frameworks
1:35:37 to present
1:35:40 new ideas within a business
1:35:46 so
1:35:49 so one of the one of the ways that you
1:35:51 can use these large language models is
1:35:53 essentially have it teach you
1:35:56 it and it might
1:35:58 it might be teaching you stuff you
1:35:59 already know
1:36:00 but like you can just know that there
1:36:03 are
1:36:05 business Frameworks out there or you
1:36:08 know storytelling Frameworks out there
1:36:09 that that you can use for
1:36:12 um
1:36:14 you know for organizing thoughts the
1:36:16 problem solution benefit framework the
1:36:18 SWOT analysis framework the impact
1:36:20 feasibility viability framework so
1:36:23 there's three of them
1:36:24 um I'm gonna say
1:36:26 could you combine two and three into a
1:36:29 novel framework
1:36:34 certainly here's the idea framework
1:36:38 to me idea uh here's the idea framework
1:36:41 impact development development
1:36:43 evaluation and adaptation it came up
1:36:46 with it with its own new [ __ ] acronym
1:36:49 and and you know ordered them so that it
1:36:52 made that acronym before it actually
1:36:53 wrote it so the impact the development
1:36:57 the evolution all right so
1:36:59 great
1:37:01 um so I'm going to say
1:37:05 let's see
1:37:09 sabalenka up 4-1 in the type we'll just
1:37:11 do it
1:37:13 so and and this is 3.5 so it does it
1:37:15 really fast right so so let's see what
1:37:17 it actually did here
1:37:19 um
1:37:20 title Tech Turtle diversifying revenue
1:37:23 streams at your company here
1:37:25 introduction to explore a new Revenue
1:37:27 stream Opportunity by selling Turtles at
1:37:29 the annual state fair agenda problem and
1:37:32 opportunity identification
1:37:35 strengths and opportunities solution and
1:37:38 feasibility weakness and threads so so
1:37:40 we've got our little framework here
1:37:43 highlight the current challenge that
1:37:44 your tech company faces opportunity
1:37:46 explain how participating in the State
1:37:48 Fair Wait So Okay so
1:37:51 [Music]
1:37:52 rather than tell me
1:37:54 how to
1:37:56 create the presentation I actually want
1:37:58 you to create the arguments and fill in
1:38:01 this outline
1:38:06 I don't want to do any work here like I
1:38:08 want to if we're gonna have the robots
1:38:10 do the work do the [ __ ] work right
1:38:11 don't just tell me how to do it you have
1:38:14 a pain in the ass
1:38:16 so problem our company faces seasonal
1:38:19 fluctuations in Revenue resulting in
1:38:20 instability and financial challenges
1:38:22 that is actually true
1:38:24 opportunity the annual state fair
1:38:26 presents an untapped Market drawing
1:38:28 thousands of attendees eager to engage
1:38:30 with unique experiences and products
1:38:33 so so one way you can attack this is to
1:38:36 do that let me see um I'm going to ask a
1:38:39 question here
1:38:41 is there a way to convert
1:38:45 the slides that you've presented here in
1:38:48 a format that I can import into
1:38:50 PowerPoint so that it automatically
1:38:53 creates the presentation
1:38:59 uh I don't know what I would say I said
1:39:01 I'm unable to do it I know it can do
1:39:03 some kind of things that you can import
1:39:04 into Excel
1:39:07 I can provide you with the content that
1:39:09 you can easily copy all right that's
1:39:10 fine
1:39:11 um so that's kind of cool so you can do
1:39:13 that
1:39:14 um code interpreter the Advanced Data
1:39:17 analysis can write files out so it can
1:39:19 write Excel files it might be able to
1:39:21 write out a file but here's we'll we'll
1:39:24 take you to another tool there's another
1:39:26 tool called tome t-o-m-e
1:39:29 dot app
1:39:34 and there's a couple of these now so if
1:39:36 you go to futurepedia.io
1:39:39 and just look for presentations
1:39:43 um
1:39:48 you'll be able to find some apps that do
1:39:50 this so
1:39:52 let's say we're going to create a new
1:39:54 thing here at Tome so it's it's kind of
1:39:57 a dark and Moody site but down at the
1:39:59 bottom here you can create a
1:40:00 presentation or a web page or a document
1:40:03 you can import a document to a
1:40:05 presentation I'm going to say create a
1:40:06 document about or presentation about
1:40:14 I want to sell Turtles at the annual
1:40:17 state fair as a way to
1:40:21 augment my revenue for the seasonality
1:40:24 in my tech business and make Arguments
1:40:27 for why this is not a bad idea
1:40:33 and so it's gonna now so it says
1:40:35 generating outline and now there's our
1:40:38 outline and I think I can go in and can
1:40:40 I edit these so it says introduction
1:40:42 seasonality in the tech business
1:40:44 benefits of selling Turtles
1:40:46 and let's say benefits of selling tiny
1:40:49 turtles
1:40:50 yeah
1:40:52 so I can go in and I can I can
1:40:55 um you know edit my outline and now I
1:40:57 say continue
1:40:59 and now shell shock that's [ __ ]
1:41:02 hilarious that's a that's a great title
1:41:04 screen
1:41:10 so the title screen shell shock making
1:41:12 the case for selling Turtles at the
1:41:14 state fair and then there's our little
1:41:17 outline and then each of the slides look
1:41:21 come on
1:41:24 so these are the like this is not my
1:41:27 style of presentation these are way too
1:41:29 [ __ ] wordy but
1:41:31 if you just need to get something
1:41:33 blocked out something just you know set
1:41:36 set up
1:41:40 and you want to do it quickly I mean you
1:41:42 know what did that take us 30 seconds
1:41:43 and you know and I've I've given it
1:41:46 ridiculous stupid [ __ ] things so
1:41:48 imagine it imagine giving it something
1:41:50 decent
1:42:00 I actually lost the second set
1:42:03 I think sabalink is going to come back
1:42:05 and clean the floor with her in the
1:42:07 third set
1:42:08 interesting
1:42:10 I didn't pay attention
1:42:16 yep
1:42:29 yep
1:42:30 seven six in the seventh Keys is when
1:42:34 the first set 6-0 in 30 minutes
1:42:39 wow all right well there you go
1:42:41 that's why you play the game
1:42:44 oh did I give you everything you need
1:42:47 yeah that's that should be that should
1:42:48 at least give you a start uh
1:42:50 just using chat GPT can probably
1:42:53 and gpt4 is better than GPT 3.5 in terms
1:42:57 of quality of what it writes
1:42:59 so I would use gpt4 that's how I would
1:43:01 write a presentation a tool like Tome
1:43:04 has just got it's just too quirky for me
1:43:06 it's they've got a specific idea about
1:43:08 how to how to do a presentation I don't
1:43:10 like it but whatever
1:43:20 good day to you sir good day to you
1:43:23 fantastic Bob tell him what is one
1:43:25 you've won a new car
1:43:30 that just doesn't work with an English
1:43:32 accent does it
1:43:35 second beverage
1:43:37 um
1:43:38 ice cold Rocky Mountain tap water
1:43:43 uh um
1:43:45 [Music]
1:43:47 alter X Alden is amazing for analysis
1:43:50 not a piece of [ __ ] oh that's cool this
1:43:51 is final
1:43:53 yeah I just I know nothing about data
1:43:56 analysis like it's something that
1:43:57 fascinates me but it just I don't have
1:43:59 the mind for it and so having a tool
1:44:01 like Advanced Data analysis is pretty
1:44:04 [ __ ] sweet I always like cinnamon
1:44:06 Amber Jade and Lexi whatever happened to
1:44:09 be Lexi to the main stage Lexi to the
1:44:12 main stage Lexi
1:44:14 let's hear it for Amber don't forget to
1:44:16 tip your waitress try the chili
1:44:20 laughs
1:44:23 oh yeah
1:44:26 only Enterprise sock 2 compliant though
1:44:30 can't input sensitive data in chat gbt
1:44:32 that's correct yeah the the Enterprise
1:44:34 version of chat CPT is going to have all
1:44:36 that [ __ ]
1:44:38 um
1:44:40 you know if if you need if if you really
1:44:43 need to do the super privacy thing
1:44:44 you're gonna have to host your own
1:44:45 models or you're gonna have to spend the
1:44:47 money to to spin up instances on Azure
1:44:49 or whatever it might be so
1:44:52 um the
1:44:54 the chat GPT Enterprise right now
1:44:58 they're they're they're communication
1:45:01 about it is if you're spending ten
1:45:04 thousand dollars a month and they don't
1:45:05 say on us but I assume it means on them
1:45:08 if you're spending ten thousand dollars
1:45:11 a month on chat gbt call us and we'll
1:45:13 set up your chat GPT Enterprise they
1:45:16 have said that they're going to have a
1:45:18 self-service version for small
1:45:20 businesses but that ain't there yet so
1:45:23 so yeah the again all of these tools
1:45:26 right now
1:45:28 let's say it again
1:45:31 they are janky pieces of [ __ ]
1:45:34 because because they're all just too new
1:45:36 and and they're gonna get better so if
1:45:39 you're if you're in an Enterprise you
1:45:42 know you're gonna have to team up with
1:45:43 Microsoft or you're gonna have to you
1:45:46 know host your own you know instances of
1:45:49 Open Source large language models or
1:45:51 you're gonna have to wait for open uh
1:45:53 open AI to get their [ __ ] together or
1:45:55 you're gonna have to wait for Google to
1:45:56 get their [ __ ] together you're gonna
1:45:57 have to use AWS you're gonna like like
1:46:00 you know if you get the budget for it
1:46:02 you're probably already dealing with
1:46:04 some of these technical issues
1:46:06 um but for these tools to get strong
1:46:09 enough for us mere mortals to use if we
1:46:12 don't have massive budgets we just have
1:46:14 to wait for it to get out of this phase
1:46:15 and it will it will
1:46:20 um
1:46:28 is he singing Hawaii 5-0
1:46:33 great job you've been live for 90
1:46:34 minutes it's 10 o'clock I do know where
1:46:36 my children are
1:46:38 all right
1:46:40 cinnamon rolls organic cinnamon rolls
1:46:42 they're organic exactly
1:46:46 so good cinnamon rolls it's it's yeah
1:46:49 it's bad it's bad situation
1:46:51 it's a bad situation in fact
1:46:53 I shouldn't right I shouldn't like champ
1:46:58 no one cares no one cares
1:47:01 chill
1:47:02 y
1:47:07 someone's existing outside his window he
1:47:09 doesn't like it
1:47:14 I mean dogs are the best but you know
1:47:17 come on just cool out dude right you
1:47:19 know Kyle Anne was looking for a
1:47:21 PowerPoint suggestion since you're here
1:47:25 um
1:47:26 I think I gave it unless she was looking
1:47:28 for something else
1:47:31 uh
1:47:33 thanks for answering on LinkedIn
1:47:37 what upcoming platform that we need to
1:47:39 keep an eye out
1:47:41 um
1:47:43 I'm gonna start at diet on Monday so I'm
1:47:45 going to need a box of cinnamon rolls
1:47:47 this weekend bubbles I fully support
1:47:49 that choice with my I will enable your
1:47:52 choice of shitty health behavior to
1:47:55 justify my choice of shitty health
1:47:56 behavior have fun this weekend just
1:47:59 Smirnoff heat it up in the microwave of
1:48:02 course and just smear it all over your
1:48:04 face right before jamming it in there I
1:48:06 I'm right there with you that's what I'm
1:48:07 gonna do after this tonight
1:48:10 um there was an actual question here
1:48:13 uh wait how many know you created some
1:48:16 digital clone that talked to you the
1:48:19 talk that no one's here I don't know
1:48:22 we'll get there
1:48:24 um wait what did I miss someone had a
1:48:25 question oh what upcoming platform do we
1:48:28 need to keep an eye out
1:48:31 um
1:48:32 the one that I'm actually most excited
1:48:34 about right now is Gemini so
1:48:38 so Google has been
1:48:41 playing catch up
1:48:44 they've been playing mustard too it's a
1:48:47 dad joke get it ketchup mustard
1:48:52 um
1:48:55 bard.google.com
1:48:58 this was laughably bad until
1:49:03 July
1:49:04 and then all of a sudden it got good
1:49:07 and not great but good like it's at the
1:49:10 Kids Table oh no it's not it graduated
1:49:12 it's at the adults table it's in the top
1:49:14 three
1:49:15 it's connected to the internet
1:49:17 um it integrates hyperlinks in a pretty
1:49:19 interesting way although it won't give
1:49:21 you a list of links which really pisses
1:49:22 me off about it I think I might knock it
1:49:24 to the Kids Table actually but
1:49:26 that's just using their Palm 2 model
1:49:29 they're working on something right now
1:49:32 called Gemini which someone within
1:49:35 Google is claiming is five times better
1:49:37 than gpt4 I don't buy that for a [ __ ]
1:49:40 second but if it's as good as gpt4 or
1:49:43 even slightly better or [ __ ] it a lot
1:49:46 better great
1:49:49 um
1:49:51 it's going to be something it's going to
1:49:53 be multimodal it will it will likely do
1:49:56 audio
1:49:58 input and output it will likely do image
1:50:01 input and output it's probably been
1:50:03 trained on all of YouTube Google owns
1:50:06 all of YouTube
1:50:08 Google owns
1:50:09 all the [ __ ] we say in our Google
1:50:12 meetings Google owns everything in our
1:50:15 doc well they don't own it they have
1:50:16 access to it and I'm sure they're not
1:50:18 using it to train all of their models
1:50:20 they've got
1:50:22 two decades of
1:50:25 all of us voluntarily giving them all of
1:50:28 our data because they give a [ __ ] for
1:50:30 free
1:50:32 when the product is free you're the
1:50:35 product
1:50:36 always remember that so Google
1:50:40 it's not necessarily the technology that
1:50:43 they're creating it's the data they've
1:50:45 got so much [ __ ] data
1:50:49 I heard I heard today
1:50:51 Google put Hardware in Dish Networks
1:50:56 um
1:50:58 uh knock their their network operations
1:51:01 center
1:51:04 so that they could take the viewing of
1:51:07 what we all watched on Dish Network and
1:51:10 take all of what we watched so that they
1:51:12 could serve up Dynamic ads to the to
1:51:14 Dish customers
1:51:16 so they've got [ __ ] everything
1:51:20 so if Gemini is trained up on even you
1:51:23 know some majority of that data it's
1:51:26 probably going to be pretty amazing
1:51:27 that's that's the one that that I would
1:51:31 I would pay close attention to if it
1:51:33 comes out and it's a janky piece of [ __ ]
1:51:36 that'll be disappointing because Google
1:51:38 [ __ ] invented the Transformer they
1:51:40 invented the thing that all of these
1:51:41 tools are are leveraging to do the magic
1:51:44 that they do
1:51:46 um and then they kind of mothballed it
1:51:48 because it threatened their business
1:51:49 model so I I just I have a feeling
1:51:51 they're going to come come out swinging
1:51:52 and then the other one to pay attention
1:51:54 to Apple's got their
1:51:56 iPhone 15 is coming out in September
1:51:59 sometime probably in a week or two
1:52:02 um
1:52:03 if they make an AI announcement in
1:52:05 September it's probably going to be
1:52:07 decently significant if they don't they
1:52:10 probably won't make an AI announcement
1:52:11 until June at the developers Conference
1:52:14 next year and it will be really
1:52:17 significant by then because they're just
1:52:19 sit Apple's just sitting back saying you
1:52:22 kids go play
1:52:24 we're gonna spend our 24 billion dollars
1:52:26 of research money on generative AI well
1:52:29 you you kids are in the sandbox there
1:52:31 and when we come out we're going to come
1:52:32 out swinging
1:52:34 so Gemini apple and then the third one
1:52:37 would be whatever Elon Musk is going to
1:52:39 do with X but you know that it might
1:52:41 take him a year or two to spin that up
1:52:43 enough to have something interesting but
1:52:46 those are the those are the three I'm
1:52:48 I'm looking at um Ola thanks for
1:52:50 answering on LinkedIn remotesque how do
1:52:52 you know that you created some digital
1:52:54 clone oh that we answered that before
1:52:56 Kyle Ann was looking for PowerPoint I
1:52:58 think I did that
1:53:00 it's an AI model training job but you
1:53:03 have to test for it and it's an hour
1:53:05 long they do pay you 50 bucks oh cool
1:53:07 that's interesting I like turtles too
1:53:09 and Rando
1:53:11 um it's not the AI learning lab without
1:53:13 without the I like turtles clip you
1:53:15 gotta have the I like turtles clip right
1:53:16 people people need the turtles
1:53:19 people need their Turtles
1:53:21 I like turtles turtles are awesome
1:53:24 I have tested and not gotten results nor
1:53:26 have I been paid yet it was just today
1:53:28 though nice did you just snort snort
1:53:31 that I like turtles I did I started the
1:53:34 the what was it a week and a half ago I
1:53:36 was drinking coffee and I made myself
1:53:38 laugh because I find myself entertaining
1:53:40 apparently I made myself laugh so hard I
1:53:43 just blew
1:53:44 um coffee straight across my face it was
1:53:47 it was fantastic
1:53:49 let's get one of my phone it was
1:53:51 everywhere
1:53:56 one one in the third set on serve
1:53:59 Madison serving
1:54:01 all right I'm using chat GPT for
1:54:03 blogging and it's crazy good yeah it's
1:54:05 yeah exactly if you write good prompts
1:54:07 so so that's the thing in Rando that
1:54:09 that the prompts are everything
1:54:12 um
1:54:12 and
1:54:14 you know it's it's
1:54:17 the the the the the big thing that has
1:54:21 has worked the best for me is you you
1:54:23 don't need to write the perfect prompt
1:54:25 to get the perfect answer very often
1:54:27 what it is is a series of smaller
1:54:29 prompts where you're giving chat GPT a
1:54:32 chance to to answer some things in
1:54:34 between the the real prompt
1:54:36 and what that's doing is it's building
1:54:38 up context so that it so that it more
1:54:41 narrowly focuses your answer right so
1:54:43 like the thing I did before where I said
1:54:45 I want you to act like a a business
1:54:47 consultant or a business writer or
1:54:49 whatever it was you know give me the the
1:54:51 storytelling Frameworks and then okay
1:54:53 combine these two and now write a thing
1:54:55 using that framework all of that kind of
1:54:57 Narrows the context so that you get more
1:55:00 and more focused answers so yeah so and
1:55:02 when you get when you get those chains
1:55:04 of prompts right like capture all that
1:55:06 [ __ ] and throw it in Trello or something
1:55:08 like that I've started using Trello for
1:55:11 my prompt Library
1:55:12 I can do different projects and then
1:55:14 each card is a set of prompts it's been
1:55:16 working pretty good
1:55:17 uh I have the full version very cool
1:55:19 comes children's zombies movie Survivor
1:55:22 uh
1:55:23 um remember when they used to tell us to
1:55:25 go look in the encyclopedia I do or the
1:55:28 library damn I'm old I know bubbles
1:55:29 listen this is the thing like in my day
1:55:32 you had to open a goddamn book
1:55:34 you had to pretend like you were reading
1:55:36 all right
1:55:37 digital Gods thank you for the lollipop
1:55:40 that looks like a cinnamon roll I think
1:55:42 I'll have a cinnamon roll after this
1:55:45 oh oh bang oh that was a nice shot oh
1:55:50 Madison Keys
1:55:53 killer drop shot
1:55:56 killer drop shot all right she's not
1:55:58 folding she's not folding
1:56:00 she's not folding like a greeting card I
1:56:03 want sabalenko to win the tie breaker
1:56:05 but I like both players just want to
1:56:07 watch a third set you got it man she won
1:56:09 that second set that was that was some
1:56:11 [ __ ] tough tennis man she was like on
1:56:14 the verge of out like gone and then she
1:56:17 [ __ ] battled back and keys collapsed
1:56:20 oh oh Keys just won the
1:56:24 second game
1:56:27 that was a nice shot
1:56:29 it's good yeah they're both good
1:56:32 and uh and Coco looked really good
1:56:34 tonight too this is going to be exciting
1:56:35 final and the the men's final is going
1:56:38 to be crazy
1:56:39 um put it in gold jewelry bubbles we
1:56:42 brought Ours from door-to-door
1:56:44 Britannica salesman the encyclopedias
1:56:46 good night everyone good night okay
1:56:47 farmer peace out tell uh Tell Tobias to
1:56:51 be careful in the hot tub we haven't
1:56:53 seen him since he dropped his phone oh
1:56:55 wait here he is
1:57:00 hang on stop
1:57:17 dog management dog management I had
1:57:21 white out I was excited when my
1:57:24 typewriter was fancy and I could correct
1:57:26 my freshman year yeah I remember those I
1:57:29 I had a fancy typewriter at one point
1:57:30 that that you could hit the whatever the
1:57:32 key was and it would go back and hit the
1:57:34 same key and white it out that was
1:57:35 pretty that was pretty slick
1:57:39 all right champy you don't remember
1:57:42 typewriters do you no because you don't
1:57:44 have language because you're a dog
1:57:46 that sucks right oh
1:57:51 all right
1:57:52 uh fingerprint
1:57:54 apocalypse that's funny
1:57:57 um which platform can design your EG
1:58:00 living room by the picture that you
1:58:02 upload of it
1:58:05 uh
1:58:11 there was a tool that did that I I think
1:58:14 it's a stable diffusion
1:58:17 I don't remember I would go to go to
1:58:20 futurepedia.io
1:58:22 go here
1:58:25 go there
1:58:28 and then look up interior design or
1:58:31 design I know there's a bunch of tools
1:58:32 that it that do interior design stuff
1:58:35 I'm pretty sure there's an app that came
1:58:38 out where you can upload a picture of a
1:58:39 room and it just it just will redesign
1:58:42 it and then I know there was there was
1:58:43 some
1:58:45 [ __ ] that did that with stable diffusion
1:58:49 um you can upload an image in Discord
1:58:52 let's see let's go
1:58:55 let's go play in Discord let me show you
1:58:58 something
1:59:03 all right we're gonna go to a mid
1:59:05 Journey channel so here's the thing I
1:59:07 was working on the other night
1:59:09 and we're gonna go slash imagine
1:59:14 um
1:59:16 let's say living
1:59:19 room uh uh 70s
1:59:25 design Style
1:59:29 photograph
1:59:37 so so what I'm in right now so I'm in
1:59:39 I'm in uh Discord so mid journey is is
1:59:43 one of the image generation tools you
1:59:45 have to make the images in Discord which
1:59:48 is a real pain in the ass at some point
1:59:50 they'll make a website they'll put a
1:59:52 website front and front end on this but
1:59:54 it's not there yet
1:59:56 so you type slash imagine
2:00:01 and
2:00:03 whatever your prompt is and then it'll
2:00:05 make your image so now it's making the
2:00:07 images
2:00:08 oh and that's interesting it didn't it
2:00:10 didn't show the uh
2:00:13 it didn't show the
2:00:16 um it building which it normally does it
2:00:18 just it just sort of presented me with
2:00:20 the images
2:00:21 so let's see we'll grab this fourth one
2:00:23 here so in mid Journey just little
2:00:26 little uh thing so it produces four
2:00:30 Images
2:00:31 u1234 means upscale one two three four
2:00:34 and then V means uh variation or vary so
2:00:37 you can do a variation of any one of
2:00:40 those images so I'm just going to grab
2:00:41 this fourth one because it's good so I'm
2:00:44 going to upscale that so it's going to
2:00:45 give us a big version of
2:00:47 of that
2:00:50 so there's our big version of that so so
2:00:52 now I'm gonna
2:00:53 I'm gonna
2:00:54 how do I do this
2:00:58 I'm gonna grab the URL here so I'm going
2:01:01 to say copy
2:01:03 copy link wait where's the copy link
2:01:06 so I'm going to copy the link to this
2:01:08 image so so this exists now on Discord
2:01:10 server somewhere so I just copied the
2:01:13 link to that so now I'm going to do
2:01:14 slash imagine
2:01:20 and for the prompt rather than typing
2:01:23 anything in right away I'm going to
2:01:25 paste in the link to that image
2:01:27 so now I'm using that image as the seed
2:01:30 for the next prompt and I'm going to say
2:01:32 modern
2:01:34 2023 living room
2:01:39 um remodel I don't know what remodel
2:01:42 will do but I'm just guessing here
2:01:45 and so now it should use this image so
2:01:49 so we should get something that's like
2:01:50 relatively in the neighborhood of a room
2:01:53 with a couch and a coffee table and
2:01:55 shelves
2:01:56 um but it should look more modern than
2:01:58 this 1970s thing with the blown out
2:02:00 carpet that carpet this this really does
2:02:03 look like you walked into a house that
2:02:05 was last remodeled in the 70s
2:02:15 oh it's coming it's coming
2:02:24 bang and so so it did kind of a cool
2:02:28 modern update here like
2:02:30 let's upscale number four again
2:02:37 and it's like it gave us a glass table
2:02:39 it gave us new carpet right so so with
2:02:43 mid-journey you can upload an image you
2:02:45 can also do this in
2:02:47 um in stable diffusion you can do it in
2:02:49 leonardo.ai that I talked about
2:02:53 um
2:02:55 the the
2:02:57 ability to sort of control things
2:03:00 there's a thing in stable diffusion
2:03:01 called control net where you can sort of
2:03:03 control certain elements the other way
2:03:05 to do something like this is to do
2:03:08 um generative fill within Photoshop
2:03:10 where you take an image like this and
2:03:12 then you can highlight the couch and you
2:03:14 could turn it into like a green velvet
2:03:15 couch instead of a red leather couch and
2:03:17 it'll just do that
2:03:19 um so there's lots of different ways to
2:03:21 do it but that's that's one way that you
2:03:23 can attack that problem so
2:03:26 hopefully that gives you some ideas some
2:03:28 some inspiration you know we want to we
2:03:31 want people to be inspired here you know
2:03:33 we don't want we don't want them like oh
2:03:35 I'm bored again with this AI stuff it's
2:03:38 just so remarkable but who cares you
2:03:40 know I just I just made myself a living
2:03:43 room and like but whatever whatever
2:03:46 like like who doesn't just generate
2:03:49 photographs out of thin air I do it's
2:03:51 dumb it's like whatever we don't want
2:03:54 you to be that person we want you to be
2:03:56 all excited like I made it I made a
2:03:58 living room it's so exciting Sally you
2:04:00 want to see my living room no I don't
2:04:01 like AI
2:04:03 oh Sally you're such a drip
2:04:05 then like a year later Sally's like
2:04:06 Could you teach me AI
2:04:09 don't don't do don't let Sally push you
2:04:13 around like that
2:04:14 Sally wasn't there for you when you were
2:04:16 learning AI and now Sally wants to know
2:04:19 about it
2:04:20 sure I sound bitter
2:04:23 but it's a competition
2:04:26 you gotta beat Sally she gave you
2:04:28 attitude it's not right
2:04:30 you thought she was your friend but no
2:04:33 she doesn't support your AI habit
2:04:36 that's why you hang out at this [ __ ]
2:04:38 weird place
2:04:42 Sally's like what do you do every night
2:04:44 I don't want to talk about it no what do
2:04:46 you do you can tell me I'm your friend I
2:04:49 watched this old guy on Tick Tock what
2:04:51 why do you do that I don't know it's
2:04:54 embarrassing don't talk to me
2:04:56 don't you aren't you going to lunch
2:05:00 you know it's got to be embarrassing I
2:05:03 understand I understand I mean I'm here
2:05:05 I'm me
2:05:06 I get it
2:05:07 it's embarrassing
2:05:11 all right fantastic Rob tell him what
2:05:13 he's won
2:05:14 can you create a web page yes let's go
2:05:17 create a web page
2:05:20 it's so good it's so good
2:05:28 oh there's a YouTube channel look it's
2:05:30 me more me yay you did it you're
2:05:35 everywhere yay
2:05:37 you're special
2:05:39 all right I'm gonna do GPT 3.5 again so
2:05:42 so let's see
2:05:44 um
2:05:45 what do you want to do just just a web
2:05:47 page okay I don't know what kind of web
2:05:49 page you want Lonnie but again here's
2:05:51 here's how I would start this I I would
2:05:53 start with the content right it's it's
2:05:55 always good to start with content
2:05:59 um
2:06:03 what's going on did Siri take over my
2:06:05 microphone
2:06:08 uh Siri please tell me you didn't take
2:06:10 over my microphone
2:06:13 oh yes she did
2:06:17 hello
2:06:19 oh dang it
2:06:24 okay so what I'm gonna do is I'll I'll
2:06:26 type sorry about this
2:06:28 um
2:06:29 let's see
2:06:31 um
2:06:32 you're that may be beyond my abilities
2:06:34 at the moment yeah that's definitely
2:06:36 Bond your abilities at the moment can I
2:06:38 have my microphone back
2:06:41 you're a business copywriter tasked with
2:06:46 writing
2:06:48 a website that describes a business that
2:06:52 sells small turtles
2:06:54 to children under 15
2:06:58 during the school day period
2:07:03 write the outline for the website
2:07:05 that will organize information in a way
2:07:09 that's intuitive for people looking for
2:07:12 Turtles for their kids
2:07:18 all right so
2:07:21 so there's your outline of your website
2:07:23 right
2:07:24 Bang
2:07:26 um so you can look through here header
2:07:30 logo and business name hero section
2:07:32 about us our Turtles blah blah blah okay
2:07:37 let's just say we like that okay so
2:07:40 we'll say
2:07:40 [Music]
2:07:45 that's good can you flesh out that
2:07:48 outline with actual content
2:07:50 of
2:07:52 the home page and the top level Pages
2:07:55 below it
2:08:01 all right so so there's our content
2:08:06 um there's lots of it there so so now
2:08:08 what you can say is okay
2:08:10 um
2:08:15 I want to make sure that this website
2:08:18 follows modern design standards and uses
2:08:23 CSS
2:08:26 and
2:08:29 modern HTML practices
2:08:32 to ensure the performance of the site
2:08:36 is as fast as a turtle
2:08:43 and then I'm going to say
2:08:50 please take the copy from the home page
2:08:52 and generate
2:08:56 the HTML for me
2:08:58 so I can see what it looks like period
2:09:01 be sure
2:09:03 to make it look pretty and intuitive to
2:09:07 read the most important information
2:09:13 and then it starts writing your code
2:09:19 things about this so so yes I can write
2:09:21 code and and the nice thing about HTML
2:09:24 or python or whatever is you can just
2:09:25 grab that code and you know pop it in
2:09:27 and see how it looks and then tell it
2:09:30 you can just say oh you picked helvetica
2:09:32 I hate helvetica I want you to use Comic
2:09:35 Sans whatever right
2:09:38 um
2:09:38 the other thing you can do is you can
2:09:40 take code from an existing site and pop
2:09:43 it in here and just have it
2:09:44 swap out the content right so so there's
2:09:47 like there's there's 18 different ways
2:09:49 to skim this cat the other thing you can
2:09:51 do is go to futurepedia.io I'm sure that
2:09:54 there are sites dedicated to making
2:09:56 websites where you you can just you know
2:09:58 type in a prompt and it'll just make the
2:10:00 whole site in fact I played with one
2:10:03 a month or so ago that doesn't do this
2:10:07 level of the site but it does the
2:10:09 the site map and the wireframes for all
2:10:13 the pages and it was literally you type
2:10:15 in a prompt and it does all of that [ __ ]
2:10:17 so I forget what that one was called
2:10:19 though but there's there's lots of tools
2:10:21 out there for this stuff so
2:10:23 wait gotta sneeze
2:10:28 winner
2:10:29 oh Madison
2:10:30 four to two
2:10:32 although sevenenko is ready to break her
2:10:35 this is gonna be a good one
2:10:38 scary man
2:10:42 all right
2:10:44 it was a good sneeze
2:10:47 um
2:10:48 web pages
2:10:50 yeah
2:10:51 um
2:10:54 so
2:10:55 so the answer to
2:10:57 the answer to the question can it blank
2:11:00 is essentially yes
2:11:04 um
2:11:05 to what degree of completeness it
2:11:08 answers that question that's that's
2:11:09 where we're in the janky piece of [ __ ]
2:11:11 category still sometimes it's really
2:11:14 good at stuff sometimes it's not so good
2:11:16 apparently gpt4 has gotten worse at
2:11:19 coding
2:11:21 um according to a Stanford thing that
2:11:23 just came out it got pretty dramatically
2:11:24 worse so I've been hearing it
2:11:27 anecdotally from people for a while
2:11:30 and apparently it's true but it's still
2:11:34 [ __ ] better than I am at HTML so so
2:11:37 you know what are you gonna do what are
2:11:40 you gonna do dear Buck huh what time is
2:11:42 it 10 24. I can't stay here all night
2:11:44 what do you do people doing with your
2:11:45 lives what are you doing like Tobias
2:11:47 just sitting in a hot tub all night
2:11:49 listening to AI learning lab weirdo then
2:11:52 he dropped his phone in the hot tub
2:11:54 it's just not right it's just not right
2:11:57 probably got a wine in one hand in AI
2:11:59 learning lab in the other Boop there we
2:12:01 go right down in the drink
2:12:03 all right
2:12:07 [Music]
2:12:12 fancy wax the keyword
2:12:15 tiny turtles that grows up to be ninjas
2:12:18 getting better thank you that's cool
2:12:20 turtle self-shoot
2:12:23 super turtle was a little pudgy what app
2:12:26 is that that app was Tome t-o-m-e dot
2:12:30 app app
2:12:32 tome.app is the one that I made the
2:12:34 presentation where all the turtles it
2:12:36 drew all the turtles
2:12:38 bite-sized Turtles what happens that
2:12:40 again that was Tome Tome a lot of people
2:12:42 asking
2:12:44 tome.app I should maybe put that on the
2:12:46 list that might be one of those to put
2:12:48 on the list have a good night and Rando
2:12:50 good night good night Anne Rando uh
2:12:52 um
2:12:53 time I think using their AI say you guys
2:12:57 tomorrow peace tome
2:13:00 what was that prompt again I do not know
2:13:03 what prompt I wrote everything with me
2:13:05 is spontaneous it just comes off the
2:13:06 brain I just I create I'm a Creator I
2:13:09 like to create
2:13:11 I don't like to be weighed down with
2:13:13 with regularity it's why we're irregular
2:13:16 we choose to be irregular some people
2:13:18 are regular other people's less regular
2:13:22 it's a choice we choose to be this way
2:13:25 it's just a creative flow it's a thing
2:13:28 it's a thing I don't I don't mean to be
2:13:31 pompous
2:13:32 I just am
2:13:34 It comes naturally to me you understand
2:13:40 all right sabalenko broke her back
2:13:42 servant servant to stay on set if she
2:13:44 can hold here although Keys is up first
2:13:46 point in the in the game for those of
2:13:49 you that don't know tennis you're
2:13:50 probably just like why does he keep
2:13:51 [ __ ] talking about this stupid game
2:13:53 hmm
2:13:56 it's add it's
2:13:59 the AI learning lab it's chat add yeah
2:14:07 um I wonder how many got that reference
2:14:09 probably none because my jokes are all
2:14:11 40 years old because I'm an old
2:14:14 [ __ ]
2:14:16 [ __ ] and old [ __ ] oh
2:14:21 [ __ ]
2:14:23 all right notice that CNN is using AI
2:14:26 generated still images with a small
2:14:28 caption that says hey I generated that
2:14:31 ain't gonna last that long that AI
2:14:32 generated [ __ ] is not gonna last that
2:14:34 long
2:14:35 although I know a lot of people are
2:14:37 saying we're going to Watermark
2:14:38 everything and this and that man
2:14:41 are you going to Watermark everything so
2:14:43 if you've got chat GPT built into
2:14:45 Microsoft Word and it's got It's just a
2:14:47 help me write button and you're writing
2:14:49 a document and they say help me write
2:14:50 and it writes this paragraph and then
2:14:52 you rewrite that paragraph and then you
2:14:54 take that and you say based on this
2:14:55 paragraph help me rewrite the next one
2:14:57 and it writes that one and then you
2:14:58 rewrite it and and it's ninety percent
2:15:00 you writing it with a little bit of
2:15:02 assist here and there where's the point
2:15:04 at which it's AI generated
2:15:09 you're using AI right now with your
2:15:10 spell checker and your grammar checker
2:15:12 is that AI generated well no because I
2:15:14 wrote it well okay where's the boundary
2:15:16 I think it's gonna be the same with
2:15:18 images I think it's gonna be the same
2:15:19 with video I think it's gonna be the
2:15:20 same with words does it it's not going
2:15:21 to [ __ ] matter I don't know why we're
2:15:23 labeling [ __ ] right now
2:15:26 oh seven cause a beast although keys
2:15:29 keys battling back man
2:15:34 holy crap oh oh drop shot was that out I
2:15:38 think that was out oh
2:15:40 see you guys tomorrow peace out
2:15:44 poor Champion I know champy
2:15:48 Laura C here changed my name but I'm the
2:15:50 same oh clear warrant hey Laura see
2:15:52 what's shaking
2:15:54 clear warned like clear warning I like
2:15:56 clear warned that's cool
2:16:00 um mama don't let the turtles grow up to
2:16:01 be ninjas Willie Nelson that's good
2:16:05 Black Mirror movie style I like it I
2:16:09 like it a lot oh by the way Irregulars
2:16:11 you're up to 33 000 likes you weirdos
2:16:14 um best clip ever you made me tired good
2:16:18 night your regular friends see I I wear
2:16:22 people out this is it you can take Xanax
2:16:25 or you could watch this [ __ ] thing
2:16:27 for a couple hours each night
2:16:29 [Laughter]
2:16:32 all right peace out bubbles uh this
2:16:35 third set is fun yeah this third set is
2:16:36 good it's good tennis Network now it's
2:16:39 just uh I don't know
2:16:41 ESPN2 I think
2:16:44 um what are you using for your prompt
2:16:45 Library you mentioned the prompts on
2:16:47 cards uh Trello so I don't know if
2:16:49 you've used Trello
2:16:51 Trello I'll show you my add in action
2:16:57 um
2:17:00 um
2:17:02 Trello Trello Trello Trello hey I prompt
2:17:08 Library so I've got a cool image and it
2:17:10 says the great Renaissance with AI
2:17:12 highlighted because I'm cool and I made
2:17:14 that image but I've got two prompts over
2:17:17 here with your social media prompts I've
2:17:19 got two areas for prompts with nothing
2:17:22 in them and then I have a client prompt
2:17:23 here so so I don't have many in it but
2:17:26 but basically you can do little columns
2:17:28 for your different sections and then
2:17:31 each little card is a prompt and I just
2:17:33 put the prompt right on there and then
2:17:35 this is create a 10 segment Twitter
2:17:37 thread from this article URL here so
2:17:40 this is one I use with with uh plugins
2:17:42 that read reads an article so I just
2:17:45 take a magazine article pop the URL
2:17:47 there copy and paste this prompt and it
2:17:49 generates a 10 12 or 15 segment Twitter
2:17:54 thread and it's [ __ ] good
2:17:57 it's so cool when you get when you get
2:17:59 these prompts dialed in
2:18:01 and they just they don't suck
2:18:05 it's really nice to be able to do that
2:18:06 stuff oh oh oh oh oh out
2:18:12 oh Keys break point this would put Keys
2:18:17 up five three in the third and deciding
2:18:19 Seth
2:18:20 [Applause]
2:18:23 um AI is going to allow us to repurpose
2:18:25 and enhance all our own company content
2:18:27 absolutely agree with that car Bruno we
2:18:29 can ship our own
2:18:30 we can ship our closets yeah it's I'm
2:18:33 telling you man the value of data
2:18:36 sitting in every single company right
2:18:38 now small ones big ones that AI is going
2:18:41 to liberate all this data it's going to
2:18:43 be mind-blowing wait sorry I gotta watch
2:18:47 this break point come on Madison
2:18:51 oh
2:18:53 did you get it nope she didn't get it
2:18:54 all right deuce
2:18:58 it's on ESPN Not tennis Network or
2:19:01 anyway I am not watching on ESPN
2:19:04 wait why
2:19:06 uh well AI aside you could just like
2:19:10 Crate and Barrel or Pottery Barn and
2:19:12 they'll design it for you free on paper
2:19:14 well you could do that but you know I
2:19:17 mean the cool thing about the AI design
2:19:19 stuff is
2:19:20 I I think I think one of the things that
2:19:23 we're going to start seeing with AI
2:19:24 design is when you when you start to see
2:19:27 interior designers embracing AI tools
2:19:31 AI is going to allow them to concept
2:19:33 just crazier and crazier designs and and
2:19:36 like the the the sort of scale of
2:19:39 creativity is going to accelerate
2:19:42 um it's gonna get really cool it's gonna
2:19:44 get it's gonna get fun you know I think
2:19:47 just the design World product Design
2:19:48 interior design architecture
2:19:52 um
2:19:52 all that sort of stuff's gonna get quite
2:19:55 amazing sculpture
2:20:01 and then sabalenko is just hammering
2:20:04 every [ __ ] ball
2:20:07 did she get that in
2:20:08 no I said like a one okay tied four four
2:20:12 in the third set
2:20:14 holy crap they are playing some crazy
2:20:16 ass tennis all right just found Irlo dot
2:20:19 AI do you have any experience with it no
2:20:20 but I want to know about it what is it
2:20:22 what's erlo
2:20:25 erlo dot a i
2:20:30 but
2:20:32 um new chat login
2:20:36 sign up
2:20:37 I hate this why do I have to sign up for
2:20:40 your stupid thing
2:20:42 I want to sign up
2:20:46 dang it
2:20:56 welcome start your first chat
2:20:59 need inspiration start here
2:21:07 count
2:21:10 log out okay
2:21:13 so I think this is just a pricing
2:21:15 strategy order fulfillment I need
2:21:17 assistance writing
2:21:19 product descriptions improve my site
2:21:21 meeting agenda
2:21:23 generate traffic card abandonment so
2:21:27 this looks like it's a business specific
2:21:28 chat gbt this look I assume this is just
2:21:32 a kind of a a fine-tuned
2:21:35 business focused thing let's see
2:21:38 um
2:21:41 um what is
2:21:45 analyze
2:21:49 storyvine.com
2:21:53 from a marketing
2:21:56 perspective
2:21:58 let's see if it's connected to the
2:22:00 internet
2:22:05 [Music]
2:22:17 I can provide high level analysis of the
2:22:19 website
2:22:29 so it's connected to the internet it
2:22:31 looked at my website and did some
2:22:32 analysis on it this looks okay I mean
2:22:35 this this would be at least worth
2:22:36 exploring I don't
2:22:38 like it's it's it's one of those sites
2:22:41 that's very very mysterious by Design
2:22:43 they don't they don't want you to
2:22:46 like it looks okay erload.ai
2:22:53 um it's you know it's a business site so
2:22:56 um it looks okay
2:22:58 it's it's probably worth experimenting
2:22:59 with I'm here's my philosophy on all
2:23:02 those things that are kind of
2:23:04 fine-tuned versions of chat gbt I'm just
2:23:06 playing in chat GPT right now because
2:23:08 those kind of sites are going to come
2:23:10 and go like like all of those sites you
2:23:14 don't know if there's a real company
2:23:15 behind it or if it's like two college
2:23:17 kids in a case of Red Bull and it's like
2:23:20 it's one or the other Madison just won
2:23:23 the fifth game it's five four Madison so
2:23:28 sabalenko will be serving to go 5-5 or
2:23:33 if Madison breaks her she wins she goes
2:23:35 to the finals it's a good game modern
2:23:38 250 AI pattern emulator is a good
2:23:41 business
2:23:44 uh okay cool web page about turtles at
2:23:49 the fair sheet metal in the Housing
2:23:51 Industry
2:23:52 I cloned my voice on resemble AI it's
2:23:56 scary good I haven't played with
2:23:57 resemble AI I did mine on 11 labs and
2:24:00 it's it's pretty good I also did mine on
2:24:03 Apple's IOS 17 and it's [ __ ] shitty
2:24:07 and it took like 20 minutes I apple has
2:24:10 a giant steaming
2:24:13 pile of
2:24:17 janky piece of [ __ ] with their with
2:24:20 their voice synthesis on iOS 17. I they
2:24:23 better get it better when they take it
2:24:25 out of beta
2:24:28 uh
2:24:30 uh do you just create useless garbage
2:24:32 with AI every day yes
2:24:35 and then I create good business stuff
2:24:39 the good business stuff takes longer
2:24:41 but this channel like just doing stupid
2:24:44 so here's the thing [ __ ] squid I know
2:24:47 that's kind of a troll you're kind of
2:24:48 doing that as a as a troll and I
2:24:50 appreciate that
2:24:51 um
2:24:52 Here's My Philosophy about about doing
2:24:55 stupid [ __ ] with AI
2:24:57 is you learn like like like doing the
2:25:01 thing like I want to start a business
2:25:03 about selling Turtles to kids during
2:25:06 school hours whatever it was
2:25:08 um you do stupid [ __ ] like that and then
2:25:10 all of a sudden you realize oh it was
2:25:12 actually really good at something in
2:25:14 there and then you a real problem comes
2:25:16 along
2:25:17 and because you were [ __ ] around with
2:25:19 something you're like I know how to
2:25:20 solve this
2:25:21 and so
2:25:23 I know you're kind of making fun of me
2:25:25 cool respect it like the game
2:25:28 um but
2:25:30 don't [ __ ] on it until you tried it like
2:25:33 play around with this stuff like
2:25:34 seriously mess around with it kick the
2:25:38 tires try to break it try to make people
2:25:40 laugh try to make yourself laugh
2:25:42 um try to do something super serious and
2:25:44 watch it fail miserably and then do
2:25:45 something stupid and watch it exceed
2:25:47 your expectations and then realize oh
2:25:50 there was something I did in that jokey
2:25:51 thing that I could apply to that thing I
2:25:53 actually want good results in let me go
2:25:56 do that that's it's like there's there's
2:26:00 some power in playing with language
2:26:03 we're language machines these things are
2:26:06 language machines
2:26:08 dick around with it
2:26:11 it's it will serve you
2:26:17 all right
2:26:19 put the typing window in blue sabalanka
2:26:22 is breaking back I this is this is a
2:26:24 really good match seven like an amped up
2:26:27 the power in her shots in the seventh
2:26:28 game I'm telling you man she's hitting
2:26:30 the ball she's hitting the ball she's
2:26:32 pissed off
2:26:35 she's she is [ __ ] smashing the ball
2:26:40 it's like it's like [ __ ] watching um
2:26:45 you know these these the young men that
2:26:48 their their balls have no Arc to them
2:26:50 they just go over the net flat
2:26:53 that's how she's hitting it
2:26:57 yeah she just blasted one past Maddie
2:27:03 she's pissed
2:27:05 I don't like it
2:27:07 you could stay here all night but none
2:27:09 of us would get any sleep that's on you
2:27:14 I'm gonna stay here till the game ends
2:27:18 all right she's serving to take it to
2:27:21 five five
2:27:23 game point Aster
2:27:25 wow damn she's good she's a beast at
2:27:29 sabalenka you had it Maddie you had it
2:27:32 in the second set four two you blew it
2:27:35 now look at you struggling
2:27:39 it's intense
2:27:40 uh this is why you play the game I'm
2:27:42 still waiting for my secret decoder ring
2:27:44 with membership uh we we do not have
2:27:47 Dakota rings at The Irregular Club but
2:27:49 we do have
2:27:51 foreign
2:27:53 we've got the button
2:27:56 all right
2:27:59 what did I miss oh Thelonious you missed
2:28:02 so much you miss me confusing you with
2:28:04 retro Punk
2:28:05 and what else
2:28:08 um Tobias dropped his phone in the hot
2:28:10 tub
2:28:13 he slipped in the hot tub and dropped
2:28:15 his phone in it
2:28:16 um and then he we haven't heard from him
2:28:18 since he's he's you know I think I think
2:28:21 Tobias is having a little wine
2:28:25 down goes the phone
2:28:29 I think the AI learning lab is at the
2:28:31 bottom of the hot tub right now
2:28:33 um it's a facade on top of chat EBT I
2:28:36 think yeah that that erlo.com was was
2:28:38 was a a rapper app I like to call them
2:28:40 have you tried Pi dot AI yeah Pi dot AI
2:28:43 is on my list
2:28:45 of there on my kids table list so
2:28:48 our nickname Road Warrior Princess our
2:28:51 nickname for pi dot AI is chatty GPT so
2:28:54 if you haven't done it yet get the PI
2:28:57 app on your iPhone and put it in voice
2:28:59 mode so it's in the lower right hand
2:29:01 corner you just turn on the little phone
2:29:03 thing and you can talk to it
2:29:05 it's good it's it's like really good
2:29:07 it's connected to the internet
2:29:09 um I have tried to use Pi like typing to
2:29:12 like do work
2:29:14 because it's connected to the internet
2:29:15 it's bad at that it's just it it's like
2:29:18 before I give you the answer to your
2:29:20 question let me just ask you another 16
2:29:22 questions
2:29:24 so it's very annoying but if you're
2:29:26 talking to it it's actually really good
2:29:29 so wait what do we Okay so Madison's up
2:29:32 40 15 serving for game number six or you
2:29:36 know
2:29:37 to go to six five
2:29:42 got it all right six five
2:29:46 wow man
2:29:47 civilenko serving to take it to a
2:29:50 tiebreaker for the winner
2:29:53 this is intense
2:29:55 damn my pal just posted a private
2:29:57 Chrissy Hine show on her birthday oh
2:30:00 posted from a private chrissyheim show
2:30:02 on her birthday damn
2:30:05 your pal is [ __ ] rocking it that's
2:30:08 cool
2:30:10 that's super cool cabrino
2:30:17 um
2:30:24 that was a great slogan what was a great
2:30:26 slogan
2:30:28 hello Nick Lang I don't know which
2:30:31 slogan you're referring to
2:30:33 it's funny that isn't it oh look this
2:30:36 kind of looks like a smile never thought
2:30:37 about that
2:30:43 code snake on python ask it the riddle
2:30:46 of how many killers in the room if there
2:30:49 are three and somebody
2:30:53 comes in and kills one
2:30:56 are a couple of tests they use on models
2:30:58 oh that's interesting yo bro have you
2:31:00 heard of the new AI bot by Mike
2:31:03 oh tuko
2:31:06 were you the one that was in here that
2:31:08 told me
2:31:10 um
2:31:11 that uh that about your cousin
2:31:14 what was your cousin's name
2:31:17 but Mike
2:31:20 Hawk that's hilarious I love birds of
2:31:24 prey I get it
2:31:26 oh tuko you're too clever
2:31:28 you know I'm Gen X right I'm not a
2:31:30 boomer I'm not a stupid [ __ ] I
2:31:32 know how to use technology right you
2:31:34 know that right
2:31:36 [ __ ] dum dum
2:31:38 foreign
2:31:40 I got I got caught up in that one good I
2:31:43 Emilio's wife do you remember what the
2:31:45 name was someone someone put the name of
2:31:48 it was a Mike [ __ ] like game like name
2:31:52 but I but I missed it I'm normally good
2:31:54 at that [ __ ] because I made those jokes
2:31:55 when I was your age I was a smartass you
2:31:57 [ __ ]
2:31:59 I was class clown were you Class Clown
2:32:02 well tuko
2:32:05 do you have a do you have a yearbook
2:32:06 from your high school days as Class
2:32:08 Clown I do
2:32:11 because I've been like this my whole
2:32:13 [ __ ] life I'll Take You Down
2:32:17 I'm looking forward to the AI Related
2:32:20 Disorders it's gonna be called ai ai
2:32:24 irregularity disorder
2:32:27 we're gonna be the poster children of it
2:32:29 oh Madison what are you doing
2:32:33 no you're gonna go into a tie break is
2:32:35 what you're doing
2:32:37 um have we talked Gavin newsman Newsome
2:32:41 announcement of CI future yeah brother
2:32:43 52. I mean we we I went and looked at
2:32:45 the article it's just they're basically
2:32:47 you know studying Ai and and you know
2:32:50 what the dangers are and things like
2:32:51 that so I think I think it's actually a
2:32:54 good thing I I want our government to
2:32:55 get educated about this [ __ ] and and
2:32:57 figure this [ __ ] out because
2:33:00 um they're gonna need to
2:33:03 you know provide some support
2:33:05 infrastructure when this stuff starts
2:33:07 making a big difference
2:33:10 40 to love
2:33:12 sabalenko serving to take it to a tie
2:33:14 break
2:33:16 and
2:33:18 keys keys lost the tie breaks seven to
2:33:22 one in the second set so
2:33:25 it's a little yikesy oh no Kyle's nearly
2:33:27 caught up must be must be the end is
2:33:29 near yeah exactly hey I've been tapping
2:33:31 away thanks to my my tummy semi-stop
2:33:33 semi-swamp
2:33:35 um I played a little today yes annoying
2:33:36 I'll get the app thanks oh Road Warrior
2:33:39 Princess for pi dot AI yeah
2:33:42 Pi dot AI as a
2:33:44 as a church EBT replacement or a bing
2:33:47 replacement or a barred replacement or a
2:33:49 clod replacement no not at all get the
2:33:52 app put it in voice mode and just start
2:33:54 talking to it it's actually really
2:33:55 useful It's like because it's connected
2:33:58 to the internet you can say hey what's
2:33:59 going on in my town tonight are there
2:34:01 any bands playing you know where are
2:34:03 they playing how much are tickets are
2:34:05 there any reviews for the show
2:34:08 when when Taylor Swift was in Denver I
2:34:10 was I was playing with pie and I was
2:34:11 asking about the show and what the
2:34:13 reviewers were saying it was it was
2:34:15 pretty slick
2:34:19 holy [ __ ]
2:34:22 let's have a lenko's just pissed during
2:34:24 the tie break
2:34:26 first mini break to sabalenko Boom
2:34:31 the answer is four Chachi PT 3.5 just
2:34:34 said two yeah well it hallucinates it
2:34:38 hallucinates fantastic Bob Telly buddies
2:34:42 won
2:34:43 I played a little today yes annoying oh
2:34:46 yeah that was pie
2:34:47 um
2:34:48 Wall Street analysts sees three
2:34:50 companies as the next three
2:34:52 wait Wall Street analyst sees three
2:34:54 companies as the next three trillion
2:34:57 dollar stocks due to AI
2:34:59 yeah I don't I I think absolutely
2:35:03 I think savalenko is going to take this
2:35:05 unless Madison turns it around this
2:35:07 point
2:35:09 she is melting down on the first two
2:35:11 points of the tie break
2:35:13 gotta get your [ __ ] together woman Gotta
2:35:15 Get It Together
2:35:16 come on woman you can do it
2:35:20 Focus
2:35:29 she's melting down I don't even know how
2:35:32 she returns the the serve savalenko
2:35:35 serve is just like
2:35:37 foreign
2:35:39 oh alphabet Amazon and Microsoft is the
2:35:42 first three trillion dollar stocks
2:35:44 alphabet Amazon and Microsoft
2:35:48 yeah
2:35:50 I could see that I don't
2:35:51 Amazon's
2:35:53 um
2:35:58 for love
2:36:01 sabalenko she's gonna [ __ ] take it
2:36:04 what a match
2:36:07 what a match
2:36:11 you said dick around with AI it will
2:36:13 serve you I did
2:36:16 I'm not sure it's a message they sent
2:36:18 you after class
2:36:21 well Nick Lang it's true dick around
2:36:24 with it it'll serve you it will I'm
2:36:27 telling you man have fun with it because
2:36:30 this shit's wild okay she won a point
2:36:32 Madison did
2:36:34 all right now it's thoroughly confused
2:36:37 you say Gen X but you're defensive
2:36:40 defensiveness is giving Boomer ah yeah
2:36:42 JD you say Gen X but you're
2:36:46 language skills are saying gen z
2:36:49 um AI darling stock plunge highlights
2:36:51 Wall Streets wait wait what AI darling
2:36:55 stock plunge highlights
2:36:57 wall Street's increased scrutiny over
2:37:00 technology oh I see
2:37:03 the stock plunge highlights increased
2:37:06 scrutiny over technology
2:37:10 I think I sort of get that Thelonious
2:37:12 it's late you're that is not quite a
2:37:15 that sentence almost parses
2:37:20 the cousin learning chat GPT
2:37:23 um the c3.ai tumbles on earnings as
2:37:26 analysts
2:37:28 no meaningful AI Tailwinds
2:37:32 c3.ai what is that
2:37:35 c3.ai
2:37:41 this is Enterprise AI generative AI for
2:37:45 the Enterprise AI software
2:38:06 yeah so this this looks like this this
2:38:08 looks like a machine Learning Company
2:38:10 that's been around for a while right
2:38:15 so there's so there's a big difference
2:38:18 between there's a big difference between
2:38:20 Enterprise machine learning and AI which
2:38:22 has been around for decades in
2:38:23 generative AI
2:38:26 this looks like the former and it's like
2:38:29 like
2:38:31 ai's been around for a while but the the
2:38:33 mass adoption of AI is going to be
2:38:36 because of generative AI That's just
2:38:38 starting
2:38:39 so
2:38:41 I predictions of the demise of this [ __ ]
2:38:44 are going to be very similar to the
2:38:45 predictions of the demise of the
2:38:47 internet when when the.com Bubble Burst
2:38:51 um it's it
2:38:54 yeah whatever
2:38:57 it's it's an easy story to write
2:39:00 it's all it's all a waste of time this
2:39:03 this AI is not going to do anything or
2:39:05 anyone that's that's it's just a waste
2:39:08 of time don't spend your time on it
2:39:09 don't invest in it okay
2:39:12 don't
2:39:13 oh
2:39:15 Madison five three
2:39:17 she's battling back
2:39:21 if someone comes into the room okay that
2:39:23 we did that one pretty sure terribly
2:39:26 written headline yeah it was a rough it
2:39:28 was a rough headline felonious
2:39:32 if someone okay we did that pretty sure
2:39:35 it's just labia shutting the door behind
2:39:37 them yeah exactly the company with the
2:39:39 ticker AI doesn't appear to be
2:39:41 benefiting from an AI boom Bank of
2:39:43 America says what I yeah just whatever
2:39:45 whatever right now I think sabalinka has
2:39:48 it I think so too but Maddie's serving
2:39:50 three five
2:39:54 second serve oh geez sabalank is just a
2:39:58 beast holy crap
2:40:00 six three she's got it she's serving for
2:40:04 the match triple match point
2:40:07 triple match point savalenka
2:40:20 whatever happens this was a great match
2:40:21 yep
2:40:23 terribly written headline not my byline
2:40:25 felonious not taking credit for the
2:40:28 crappy thing he posted in here I like it
2:40:30 I support your avoidance of
2:40:33 responsibility for that headline
2:40:36 I really seriously Kyle I just copied
2:40:39 and pasted it don't put that on me
2:40:42 laughs
2:40:49 that's it
2:40:51 sabalenka go into the finals [ __ ]
2:40:54 Maddie keys
2:40:55 what a match holy crap
2:40:59 oh no
2:41:03 I thought it was to seven it's the
2:41:04 finals so what's it go to nine ten
2:41:08 they have to play too
2:41:10 or did she win it
2:41:15 oh no they're still playing
2:41:17 what do they go to in a in a finals do
2:41:19 they go to nine or ten
2:41:22 oh first to ten points to buy two
2:41:25 okay so so this isn't necessarily over
2:41:27 but
2:41:29 the fat lady is warming up
2:41:33 okay
2:41:36 did she just ace her no okay
2:41:43 pretty sure it's just lobbyist shutting
2:41:45 the door behind them
2:41:50 old oh that was a good shot by Maddie
2:41:54 that was weird that was bad okay
2:41:58 one of the domains on my list never
2:42:00 heard of c3ai had to look them up they
2:42:02 grossed maybe 190 Million last year yeah
2:42:05 they've been around for a while it like
2:42:07 because they're selling into the defense
2:42:09 department and [ __ ] like that they've
2:42:11 they're doing so so you know again
2:42:15 this channel is about kind of post chat
2:42:17 GPT AI not about the AI that's been
2:42:21 you know
2:42:22 sitting in our Enterprise servers for
2:42:24 the past decade
2:42:26 um c3ai is one of those
2:42:28 you can never put the genie back in the
2:42:30 bottle that's correct this [ __ ] isn't
2:42:32 getting uninvented correction our list
2:42:36 all right what's going on oh that was a
2:42:39 good shot oh
2:42:42 sabalanka 9-3
2:42:47 fat ladies going la la la la la la la
2:42:53 [Laughter]
2:42:59 changing sides
2:43:01 Maddie serve
2:43:04 she has to win seven points in a row no
2:43:07 eight points in a row to Win It
2:43:11 foreign
2:43:15 it ain't over till it's over that's what
2:43:17 they say
2:43:20 oh
2:43:22 sabalanka just whiffed one one down
2:43:27 what can one do to access AI from a
2:43:29 mobile phone
2:43:31 um
2:43:32 just go to your browser and go to
2:43:34 chat.openai.com
2:43:37 go
2:43:39 wait wait for it go there
2:43:43 or there
2:43:45 we're there Claude day all three of
2:43:48 these URLs work on a phone just make
2:43:50 them a shortcut on your home screen
2:43:52 um
2:43:53 chat.openai
2:43:55 chatgpt has their own app she just won
2:43:57 two points so it's nine five
2:44:01 so she's still got to win four points
2:44:02 just to tie it
2:44:10 uh it's like one-on-one baseball without
2:44:12 the hot dogs
2:44:16 oh
2:44:20 did she win it savalenka she won it did
2:44:24 she win it yeah she won it
2:44:26 Maddie's out damn it damn it Madison
2:44:32 whoa seven link is two in the world
2:44:35 Maddie was what 11 or
2:44:37 13 or something that was a great match
2:44:39 that was a really great match she was 17
2:44:41 in the world something like it was two
2:44:47 wow good match good match all right
2:44:49 everybody listen
2:44:51 we're out of questions anyway
2:44:53 oh it says it's not available in my
2:44:54 country I live in Cameroon Africa um
2:44:57 you might need to use a VPN I don't know
2:44:59 what's available the the thing that you
2:45:02 can play with what's your name
2:45:04 um
2:45:06 uh Sheikh Mohammed
2:45:12 um
2:45:15 check so go to Reddit and look for the
2:45:19 um open source large language models
2:45:21 try
2:45:23 labs.perplexity.ai
2:45:33 tried this one
2:45:35 and you could try Bing I'm sure they've
2:45:38 got Bing blocked
2:45:40 labs.perplexity.ai they might not have
2:45:42 blocked although if they've got all AI
2:45:44 blocked they might but this is
2:45:47 perplexity hosting meta's llama 2 models
2:45:50 which are open source models so you can
2:45:52 try that and then if that doesn't work
2:45:55 go to
2:45:57 um Reddit and then open source large
2:45:59 language models and I know that there
2:46:02 are some projects
2:46:03 where they're optimizing the large
2:46:05 language models to be able to install on
2:46:07 phones
2:46:08 and so you might be able to do it that
2:46:10 way yeah all the sort of which countries
2:46:13 are blocking what I have no idea wow you
2:46:16 can you can pronounce my name so not
2:46:19 quite I was close
2:46:21 but but I tried peace out all right
2:46:24 listen peace out everybody let's do a
2:46:26 little Romper Room thing here I see
2:46:27 Thelonious and Dr J I see this is not an
2:46:30 exit as I make my exit
2:46:33 hair growth geek thank you for the rose
2:46:35 thanks everybody for hanging out tonight
2:46:37 ion China cabrono Dr J Emilio's wife
2:46:43 share bear peace out peace out how high
2:46:46 the moon so high good night Tobias I
2:46:50 hope your phone's okay hope you're okay
2:46:52 feel better Irregulars have a good night
2:46:55 everybody
2:46:56 see you tomorrow