ChatGPT-a-thon Part 3: 9 AM – aiLL Nov 30, 2023 – The Past, Present and Future of ChatGPT

Description

Chapters

Transcript

0:22 all right Kevin I'm here are you
0:25 ready Kevin will probably join here
0:28 shortly okay I'm back Eric Sanner good
0:31 to see you man the 24-hour chat GP Aon
0:36 continues
0:38 um I am about to
0:41 go into a zoom meeting to talk about
0:44 chat gbt Kyle have you slept at all I
0:46 did I took a nap between 11: and
0:49 midnight I took a tap a nap between 4:00
0:53 a.m. and 6:00 a.m. and now I'm at work
0:57 so I'll be I'll be turning the live on
0:59 and off but I'm I'm doing a zoom call
1:01 right now um and uh you're going to get
1:06 to hang out on the zoom call good
1:07 morning Lauren what's happening welcome
1:10 welcome welcome so feel free to share
1:13 this but I I am going to be essentially
1:16 ignoring you but I will have the people
1:19 that I'm talking to I'm going to move
1:20 the zoom window over in front of the
1:22 where chat GPT is right now happy
1:24 birthday to chat GPT um share the live
1:27 if you'd be so kind we just had a bunch
1:29 of people in here and then I canceled
1:31 and went out and it looks like a bunch
1:33 of people are joining again welcome
1:34 everybody my name is Kyle Shannon this
1:36 is the AI learning
1:39 lab let me see if Kevin Clark is
1:47 around on
1:50 when are ready and Tik Tok is
1:59 live
2:03 okay this is going to be weird I'm going
2:04 to move the people I'm talking to over
2:07 in front of this
2:08 camera but I'm going to be talking to my
2:10 camera up there so it's just there's no
2:14 other way I could arrange
2:18 it oh
2:20 man chat GPT birthday it is chat gpt's
2:23 birthday thank you Lauren happy birthday
2:26 happy chat GPT
2:29 birthday
2:30 oh man man alive do we have
2:34 uh it's it's been newsy and I it just
2:37 looks like uh stability AI just made a
2:41 big announcement on Twitter that they're
2:43 moving to some sort
2:45 of bigger subscription model that if
2:47 you're going to develop on their [ __ ]
2:49 you have to pay them which I think is
2:53 interesting interesting interesting
2:58 interesting da
3:01 [Music]
3:07 haha you love your wife I
3:12 do hello again Edson how are you wow we
3:16 got 70 people in here uh welcome so my
3:18 name is Kyle Shannon this is the AI
3:20 learning lab I'm about to go live to
3:22 talk about the oneyear anniversary of
3:26 chat gbt on a group that I belong to
3:29 Cindy who's in here dang didn't get a
3:32 notice oh Luna dick dang it Kyle are you
3:35 addicted to live streaming it's okay if
3:37 you are no I'm doing a I'm doing a
3:40 24hour I'm doing a 24hour chat GPT Aon
3:44 but it's okay if you are well thank you
3:48 I appreciate that no no I'm just doing a
3:51 uh because I make wise life decisions I
3:54 I'm I'm uh going live as much as I can
3:57 in 24 hours I've got a work to do I'm
3:59 running a company so I I do have work to
4:03 do all right let me see here's
4:07 Kevin hello Kevin Hey I'm going to put
4:12 you in a little window off to the side
4:14 here so I'll keep talking to you all
4:17 right it's fine but I'm I'm putting the
4:19 zoom
4:20 Window Live on the Tik
4:23 Tock that's fine so we're we're getting
4:26 some uh some extended reach here for the
4:29 content evolution new La extendo reach
4:31 extendo reach and there's Cindy [ __ ] um
4:35 hi friends hey and what was I gonna say
4:38 meeting to be here today you what oh you
4:41 skipped the meeting I CH the meeting no
4:43 no I didn't skip I messaged somebody and
4:45 said I have a very important meeting I
4:46 have to be at this time could we please
4:49 reschedule oh that's awesome that's
4:53 awesome go sounds good I'm going to I'm
4:57 going
4:59 to I'm GNA mute myself for
5:06 something hello Tim hey Tim morning
5:09 morning Kevin how are you I'm
5:12 good that's a good post on
5:14 Forbes thank
5:16 you yeah it's a I think it's an
5:19 interesting perspective when you think
5:21 about um the whole concept of as I
5:25 called it AI for every man or AI for the
5:29 masses
5:32 yeah that
5:34 way I was just talking to somebody I
5:37 said the um as A's you know three rules
5:40 of Robotics when you read you know the
5:42 novels it's always you know one robot
5:46 and one human and a dilemma you know
5:48 that you have in the conflict and I said
5:50 the conversation is simply shifting and
5:54 it's the same problem but it's shifting
5:56 from I to we it's not I Robot it's we
5:59 robot
6:01 robot all right it's the uh it's the
6:04 scale of augmentation interaction anyway
6:09 it'll be
6:10 fun got to
6:13 start
6:14 yeah where's
6:18 Mike I hear him I don't see him yeah
6:21 he's coming there he is hey Mike Hey
6:24 Kevin yo Mike hey Andrea hey everybody
6:26 hey Mike hey Cindy Hey Hey Kevin hey J
6:31 hey I'm going to get going here I'm
6:32 going to start recording stand
6:35 by recording in progress greetings
6:39 everyone it is uh the birthday of Chad
6:42 GPT the 30th of November and uh we're
6:47 delighted to have you with us Mike uh my
6:49 co-host is here on the West Coast how
6:51 are you I'm well I'm well how's how's
6:53 everybody doing
6:55 today doing great uh so I'm glad that we
6:59 can come together we have a special
7:01 guest you know one of our uh own uh team
7:06 members Kyle Shannon who is our chief
7:08 generative officer for Content Evolution
7:11 and he's the CEO of storyvine delighted
7:13 to have you with us Kyle and you have
7:15 some other people who are um
7:19 participating via Tik Tok so explain the
7:22 bridge I do so so so here's what's going
7:25 on I um so I have I have my Tik Tok
7:28 Channel as every good 58 year old
7:30 influencer should
7:33 have and so um because I make good life
7:38 decisions I decided to do a 24-hour chat
7:41 gpta Aon um to go live on Tik Tok as
7:44 much as I could for 24 hours so it
7:46 started last night and so off to my
7:49 right like right now I'm looking at the
7:52 at the Tik Tok there are 75 people on
7:54 the Tik Tok so we've got an extended
7:57 reach and S Cindy's on the Tik Tok as
7:58 well so let me get rid of the the the
8:01 notice there so wave wave to everybody
8:03 you're waving to 75 people out on AI
8:05 learning lab land uh and uh so I'll be
8:08 talking to you here but they'll be
8:10 watching you over here so if it looks
8:13 like I'm looking weird occasionally it's
8:15 because I'm I'm managing that um but
8:18 anyway no weirder than normal yeah
8:20 exactly you you you have been multimodal
8:22 for decades what can I say exactly so
8:25 this is this is just durger for Kyle
8:28 Shannon so glad to have all of you um
8:33 Kyle why don't you just give us and and
8:35 then Mike and I'll Chim in we're going
8:37 to go to the larger group is give us a
8:40 point of view about what has happened in
8:43 the last year
8:45 because uh reality has changed a little
8:49 bit and I value your perspective because
8:53 you really done a deep dive I've been
8:55 looking more at this at a where's the
8:58 training data sets coming
9:00 and the Govern issues but you've been in
9:03 the trenches actually using the
9:04 technology really by the way I feel like
9:06 I'm a pretty good prom engineer now yeah
9:08 you're not you're not bad oh yeah well
9:10 you're welcome and you know I I think
9:12 anyone who's who's good with words uh
9:15 you know is at an advantage with with
9:17 all this stuff um okay so for the for
9:21 the past year so so for me the the
9:24 significance of chat GPT and the reason
9:28 that you know it
9:31 the reason it's worth in my opinion
9:33 celebrating the the the birthday of chat
9:36 GPT is that I I think that it actually
9:38 marks the inflection point of
9:42 um of a shift from machine learning and
9:45 AI uh for researchers and scientists to
9:50 you know generative AI being available
9:52 to the rest of us basically to to
9:54 business and it's a very similar shift
9:56 that the the worldwide we web did to the
10:00 internet where the internet had been
10:01 around for decades uh and then all of a
10:04 sudden the worldwide web comes along and
10:05 it makes it simple enough to just you
10:07 know click click on a hyperlink and you
10:09 know anyone can make a website and it
10:12 democratized that powerful tool so chat
10:14 GPT for me marks that inflection point
10:18 and I find it remarkable that a year
10:21 into chat GPT having this you know
10:25 incredibly large
10:27 lead that no real competitors have
10:31 showed up yet you've got um anthropics
10:35 Claude is close but it's kind of like a
10:37 little bit better than chat GPT
10:39 3.5 um you've got Google threatening to
10:43 release Gemini which they haven't but
10:45 their Bard thing is almost as good as
10:47 chat GPT
10:48 3.5 um there there are eight or nine
10:52 major players that have some version
10:54 Amazon just announced Q which is
10:56 apparently almost as good as chat GPT 3.
10:59 .5 grock is apparently launching this
11:02 week I wouldn't be surprised if Elon
11:04 Musk launches it today um just to mess
11:07 with Sam Alman but we'll see um but
11:10 apparently it's almost as good as uh
11:12 chat GPT 3.5 so so one thing I find
11:15 astounding is that they still have this
11:17 this seemingly insurmountable lead
11:19 someone will catch them at some point so
11:22 then they have all this and then you
11:24 know a week ago out of nowhere um this
11:27 company that's flying so high
11:29 just has absolute flipping chaos they
11:32 fire Sam mman the CEO and then they try
11:37 to unire him and then they bring in you
11:40 know uh uh EMT Shear um to to be the the
11:46 new CEO and and now Sam's going to go to
11:48 Mig there was all that drama that
11:50 happened it now looks like they've kind
11:52 of wred the ship and open AI is going to
11:54 survive I think it was quite clear that
11:57 open AI would not have survived survived
12:00 that transition and it would have become
12:03 a wholly-owned
12:04 Microsoft effort basically um so so they
12:09 figured out a way to salvage the open AI
12:12 with the nonprofit Mission and the
12:13 profit Mission um which is
12:16 fascinating um so so that that's just
12:20 just on sort of the the the business
12:23 like the the the tech business side of
12:25 things it's just fascinating Dynamics
12:28 all around then on the what the tool
12:31 actually is side
12:34 um the capability of chat GPT from when
12:37 it launched a year ago so if you go to
12:39 chat GPT right now if you go to chat.
12:42 open.com and put it in chat GPT 3.5 mode
12:46 that's essentially the chat GPT that
12:48 launched last year and if you've got
12:51 access to chat GPT 4
12:53 mode comparing what chat GPT 4 can do to
12:57 what chat GPT 3.5 can do they're just
12:59 like night and day the quality of what
13:01 it generates is different um it can do
13:04 Advanced Data analysis it can read
13:07 images it can see images it can create
13:10 images with dolly3 um the the mobile
13:15 application of chat GPT you can put it
13:18 in voice mode so you can actually now
13:19 have a conversation with chat GPT and
13:23 then they also just launched these
13:25 things called
13:26 gpts which I think is brilliant branding
13:28 by the way that didn't call them Bots
13:30 they called them gpts so they're going
13:31 to own that term like Xerox owns copers
13:34 and Kleenex owns tissues um they're
13:37 going to own
13:39 GPT um and what those are is basically
13:42 the democratization of very
13:44 sophisticated application development
13:46 where anyone can make uh a GPT it sort
13:50 of goes beyond no code into the realm of
13:53 you can just talk in English what you
13:55 want your application to do and it will
13:58 do all the prompting for you you and it
13:59 will build this thing for you and you
14:02 can share those gpts with other people
14:05 um and at some point here shortly
14:07 they're going to be launching a GPT
14:09 store which is likely the equivalent of
14:12 the Apple App Store and lots and lots of
14:15 businesses were built on the Apple App
14:16 Store um so I I think this idea of the
14:20 democratization the sort of um
14:23 casualization of application development
14:26 is is probably one of the bigger things
14:29 that happened in chat GPT in the past
14:31 year and it's like 3 weeks old right so
14:34 still very early still very janky um but
14:37 but that's what's been going on and just
14:39 the the other thing that that
14:41 I'm I guess taking it down to the human
14:44 level one of the things that I'm seeing
14:46 on my on my Tik Tok Channel um is
14:51 there's a there's a a Vibrant Community
14:54 forming around this idea of generative
14:56 Ai and what I'm seeing consistently
14:59 are people who were retired kind of
15:06 unretires software Engineers turning
15:08 into artists I there there's I just see
15:12 um for the people that are embracing
15:15 this technology it's it's augmenting who
15:18 they are and what they enjoy in
15:20 unexpected ways and there's kind of a
15:23 joy happening for people that I'm
15:26 working with and interacting with on a
15:28 reg
15:30 basis as they're embracing this thing
15:32 it's just like things are coming to life
15:34 for them and and that I find I I think
15:38 that's unexpected for me because I I
15:40 think where my head was was these tools
15:43 are going to you
15:46 know ruin people's lives and and you
15:49 know they're going to lose their jobs
15:51 and it's going to disrupt everything and
15:52 there's like there's a lot of press
15:53 about the negative impact of this but
15:56 there's this subtle reinvention thing
15:59 happening where I think there's there's
16:00 so much opportunity for people to to
16:03 reinvent themselves their businesses
16:05 things like that that that there's a I
16:07 don't know there's there's a there's an
16:08 energy to this that I haven't
16:10 experienced since the mid 90s with the
16:12 early days of the worldwide web so with
16:14 that I'll shut up okay well Kyle that
16:17 was a that was really good U couple of
16:20 observations back to gpts for for a
16:23 moment we were
16:26 working um in the cont
16:29 Evolution uh
16:32 collab on you know developing
16:35 capabilities
16:37 and the uh it would take you know a week
16:41 or so maybe a couple of weeks and we
16:43 would you know understand what we want
16:45 to do and we'd use zapier we'd use a
16:48 Google form and we would kind of
16:49 automate a process and so on and so
16:51 forth um and it was you know was pretty
16:55 cool what we were doing but like you
16:57 said all the things that we were doing
17:00 there that were already kind of amazing
17:02 because you were using gen of
17:04 AI can manifest now in like 10 minutes
17:08 yeah right uh well and and can manifest
17:12 in 10 minutes in in an even more
17:13 powerful way so one of the things we
17:16 were trying to do in collab was use the
17:18 data that content Evolution had produced
17:21 as a knowledge base and in order to do
17:23 that we we had to set up Lang chain and
17:26 things like that and we had an outside
17:27 technical consultant helping is trying
17:29 to set that up well now you can just do
17:31 that in one of these gpts so not only
17:33 can you produce this app that does
17:35 amazing things you can produce this app
17:38 that is trained on your data right which
17:40 is one of the things people have asked
17:41 me about for the past year if they don't
17:43 know about chat GPT almost always the
17:46 first thing they ask me is well can I
17:47 put my data in there it's the natural
17:50 instinct is I want this thing to
17:52 understand who I am or you know what I'm
17:54 about and and that's now just instant
17:56 and it's a part of it and and and free
17:59 so well 20 bucks a month the the the
18:02 pace of
18:03 change that we were experiencing you
18:07 know from pre- internet by internet 50
18:11 years old next year right so that's an
18:13 interesting you know thing to yeah
18:15 notice um and take note of so 50 years
18:20 there web manifests and it becomes more
18:23 accessible to people so first thank you
18:26 vincer then thank you Tim burners Le
18:28 everybody else who contributed Y and
18:30 that's 30 years ago you know that was a
18:33 lot of log rithmic right um change this
18:38 really does live up to the phrase
18:40 exponential doesn't it yeah I think it
18:43 does and I think it
18:47 um I I was talking about this on the
18:49 live last night one of the things that
18:51 feels like it's the the biggest shift to
18:53 me is that how we've used computers
18:57 essentially since all all of us have
18:59 been alive
19:01 is we have
19:04 to we have been trained by technologists
19:08 that we have to change how we interact
19:10 with computers based on how they want to
19:12 interact right if we want to make a
19:14 computer do what we want it to do we
19:17 have to learn the computer language that
19:20 that computer understands and then we
19:21 have to talk in this very particular way
19:24 and we have to think about logic like if
19:27 someone does this then what happens well
19:29 then that happens or then that happens
19:31 and and we've had to interact with
19:33 computers in in a in a in a way that
19:36 we've had to adapt to them and what
19:39 generative AI does is flips that
19:41 completely around and and now we can
19:44 just say hey computer I want you to do
19:47 this and it does it right now yeah but
19:50 can I just inter yeah please there
19:53 something was said early by Kevin that's
19:55 I think
19:56 important how you actually
20:00 create or write a prompt m is a language
20:04 question in my mind it
20:07 is because you just said you know you
20:10 can just State you know a question which
20:13 is basically true but to get the maximum
20:16 out of your ai ai experience you really
20:21 have to ask it in a way that it
20:23 understands it and I just wanted to
20:25 interject that one particular point
20:27 because what what Kevin talked about and
20:30 made that term as a prompt engineer
20:32 that's actually become a whole
20:34 new uh category for uh a career for
20:39 people yep it absolutely has we we are
20:42 we are quickly moving though there are
20:45 hints of what's coming in chat GPT 4
20:48 right now so if you use its data
20:51 analysis tool code interpreter um you
20:54 basically just give it a goal and then
20:56 it starts writing and executing it own
20:58 python code and checking its results and
21:01 if it doesn't get a good result it tries
21:03 again and it's sort of talking to you
21:05 the whole way through it so it's
21:06 prompting itself um on on your behalf
21:10 and so so I think there's sort of two
21:12 modes here I think right now the current
21:14 tools require prompt engineering to get
21:16 the most out of them I absolutely agree
21:18 with that um I think if you want to be
21:21 in development in software development
21:23 you this will still be a key skill
21:25 moving forward but I think for the for
21:27 the masses for the rest of us these
21:29 tools are going to get increasingly
21:30 autonomous where we just give them goals
21:33 and they will go out and figure out how
21:35 to do all the prompting on their own
21:36 absolutely yeah absolutely okay Mike
21:40 jump in yeah I as Tim sparked this
21:43 thought I was thinking you know that as
21:46 Loosely as the term writer or part of my
21:48 job is
21:49 writing um what struck me is with the
21:52 first couple hours I played with four is
21:54 I'm not a writer
21:55 anymore I I used to think that was cool
21:58 important part of me um with tools like
22:01 this I'm an editor yeah right you're a
22:04 refining editor this is very much I got
22:06 to go back because it's what I know you
22:09 know you would often be working with
22:10 multiple editors as a kid journalist
22:13 right but there was usually one or
22:15 two who weren't just this is how you
22:18 structure The Story You're missing this
22:20 fact it would be that's nice you're
22:22 still too passive in this section mhm
22:25 right and so it's actually going to turn
22:28 is that said I think the task of writing
22:29 outside of truly original
22:32 fiction um and original reporting I
22:36 guess um we're editors now and I think
22:39 that's Tim what you were talking about
22:40 is not it's I don't have to learn a
22:43 specific um set of constructed things
22:46 like I did when I was trying to figure
22:47 out basic programming sure with these
22:50 very narrow Concepts but I do have to
22:53 have enough insight and intelligence to
22:56 say that's close that's not EX actly my
22:59 voice but I'm more concerned about
23:00 change these two things and add this in
23:03 so again it's it's an iterative process
23:06 yeah I think that's Kyle I think what
23:08 where a lot of people are still kind of
23:09 flx was oh my God it responded back to
23:13 me in natural language like a human
23:15 being and I think that was the first
23:17 eight months yeah right was every like
23:19 oh wow okay I can generate and this is
23:21 where I was along for the last several
23:22 months I've been okay what's the
23:24 implication for this for marketing or
23:26 for sales it's like oh my God they're
23:28 just G it's going to get worse or better
23:30 they're just going to be doing it faster
23:32 [ __ ] excuse my language faster all the
23:34 time yeah right and that's I'm still
23:37 kind of stuck there right because I'm
23:40 not sure what Mark benof thinks Einstein
23:42 is what did he call it recently not the
23:45 safe did he call it the safest no he
23:47 didn't it was something and you know as
23:50 far as I know it's been a year and a
23:52 half two years since I talked to him I
23:54 don't know how many people in the you
23:56 know Salesforce marketing Cloud customer
23:58 based are actively using Einstein on a
24:01 daily basis yeah I don't I yeah I don't
24:03 I don't know and it's and it's you know
24:07 the right now our primary primary
24:10 interaction with these tools is is with
24:13 text and now there's sort of going
24:14 multimodal they're not really doing
24:16 video yet but they are doing voice so
24:18 you can now have a conversation with
24:20 chat gbt and and haen just released
24:23 their their new Avatar tool where you
24:26 can upload three minutes of video of
24:27 yourself and it turns it into an avatar
24:30 that synthesized your voice and the
24:32 video that you put up there so you can
24:33 now just type you know type in a script
24:36 and it will be you talking on video and
24:39 and so the distance between that and
24:41 being able to interact with that Avatar
24:43 in real time is it we're you know we're
24:47 within months of that bizarre
24:50 interaction that you're having just as
24:52 chat GPT feeling like a remote employee
24:55 or like a really well educated intern
24:58 that that you're interacting with
24:59 remotely it it's going to shift very
25:01 very quickly to you're going to be
25:03 interacting with these things in real
25:05 time that that are just you know and I
25:08 think you know you I don't if everybody
25:10 saw the link that uh Kevin sent out last
25:14 week uh uh uh my former employer one of
25:18 their Keynotes about Ai and a couple
25:20 others and it struck me as we talk about
25:23 this iterative process and now we have
25:25 to be editors as we're using these tools
25:27 I think pretty much in any role we
25:30 have where do we see I the first
25:33 bottleneck I see is
25:35 management and not because they're
25:36 scared about losing their job it's like
25:38 I'm going to be generating so much stuff
25:40 so quickly and it's going to be pretty
25:42 good yep if we are we now managers are
25:46 going to have to come up with their own
25:47 chat gpts you know to say okay edit it
25:50 with this cap the my approach to editing
25:52 or reviewing these doc yeah yeah proc
25:55 process it for me yeah process is going
25:57 to kill it
25:58 even more uh uh important than that I
26:03 just had a
26:04 conversation with a financial
26:06 institution
26:07 recently and uh Kyle I have kind of an
26:11 open invitation to you know give them a
26:13 briefing right in the first quarter I'm
26:15 going to invite you great
26:17 but they haven't touched generative AI
26:21 because of the finra and other
26:24 regulations around their computer
26:26 systems and they have no clue what's
26:28 going on um you know they they can't
26:32 bring anything that they do at home you
26:34 know to to work and it's kind of like
26:37 you guys are in
26:38 trouble right uh and what it turns out
26:42 is I did some quiet inquiries the
26:45 industry is in trouble yeah right um
26:48 because there is you know no way for
26:52 them to do an
26:54 implementation inside all right and feel
26:57 confident you know relative to their
26:59 Securities and Exchange Commission their
27:01 finra all the regul they're highly
27:03 regulated right and so they put a force
27:06 field around this and say you can't use
27:08 this technology right oops that's a
27:11 problem yeah um Cindy I uh brought you
27:17 bringing you into the conversation
27:18 acknowledge that Cindy is
27:20 another uh chief officer she is our
27:24 chief media officer for Content
27:26 Evolution and she is is a faithful uh
27:30 participant in Kyle's generative
27:32 sessions in collab Cindy join us so I
27:36 have three quick things I have to say
27:38 before I get to my point based on things
27:40 that were just said first of all Mike
27:42 I'm a writer too and in this process I
27:44 haven't considered myself an editor and
27:48 instead I've been considering myself a
27:51 curator so I say that as a distinction
27:55 because I'm not exclusive and I don't
27:58 have loyalty to one Ai and I have to
28:01 curate across all of the different AI to
28:05 determine what's the best service for
28:08 what I'm trying to make happen number
28:10 one and then when it comes I'm not
28:12 really editing so much yet at this point
28:15 a little I but honestly it's more it's
28:18 it's like I'm working with an artist and
28:21 I say yes last time we had a
28:24 conversation we talked about this whole
28:26 thing being having more blue
28:28 good job but what I really meant was
28:31 more ocean blue and so as also a trained
28:35 artist I really see what I'm doing as
28:38 I'm
28:39 curating the artist that I'm partnering
28:43 with which is whichever AI so I just
28:45 want to throw that out partnering I get
28:47 I think the one thing because words
28:49 matter um
28:52 is with the notion of being a
28:55 curator there's a certain distance
28:58 you're putting between yourself and the
29:00 output all right sometimes sometimes
29:02 depends on how you
29:04 collaborate hear me out um and I think
29:08 this is why you know
29:11 um Clarity and
29:13 transparency um I still haven't come up
29:16 with what I want to put on my substack
29:17 about what I'm using AI which so far
29:19 most the posts I put up there were all
29:22 me yeah right I didn't you know the
29:24 early version was kind of clever for
29:26 doing an outline I don't consider that
29:28 that's anybody can do an outline it did
29:30 a good job give me give me right but
29:32 what I'm saying with the sense of when
29:33 you say curator you are putting a
29:36 certain distance between the I I just
29:38 collected this stuff and put it out
29:40 there or you know my favorite I'm just
29:41 asking questions here's the question um
29:45 and I think and I'm not saying putting
29:47 it down so much I think we need to think
29:48 about that right because that notion of
29:51 curation can free us to a degree to
29:56 say uh geez it was kind a cool output um
29:59 I was just curating this and I this get
30:01 down to that point that point
30:03 identifying transparency identify works
30:06 I do I think it's your background and my
30:08 background so as somebody trained in the
30:11 creative field and someone who's done a
30:12 lot of curation in my life I think it
30:15 involves way more collaboration than you
30:17 imagine and I don't think there's any
30:18 curator at any space anywhere who would
30:21 say they aren't responsible for the
30:23 output so I would say it's far more
30:26 collaborative than editing to me feels
30:29 like I'm doing it by myself and I
30:31 actually don't feel like I'm doing it by
30:33 myself collaborative you know absolutely
30:36 and I got to I'll rethink this because
30:37 editing isn't right because it was in my
30:40 experience when I was working with an
30:42 editor or another reporter who had more
30:43 experience than I did it was a
30:46 collaborative process and it was a
30:49 personal decision at which point did I
30:51 say you know Tim really helped me out
30:53 with that 10-minute conversation about X
30:55 and Y before I put this together
30:58 jeez hey Tim you want a you know want a
31:00 cooby line on this you know or you just
31:02 want a contributor line at the bottom
31:05 right that I'm old school language there
31:06 apologies but that's and this is where I
31:08 think we have to I think it is important
31:11 what we put on our things that we put
31:13 out there where we use the tool not just
31:15 hey I used I used some of this and some
31:17 of that right a little more detail right
31:19 because you can see that in the way we
31:21 as humans want to see an attestation of
31:24 the Providence of something right you've
31:26 seen if you look at the way bios now are
31:28 on most
31:29 Publications they're way more expansive
31:32 than they used to be right and there's
31:34 way more information in there about what
31:36 this reporter at this place at the what
31:38 New York Times has been is doing and has
31:41 been doing so I I've done no I just
31:44 want I think you're right we have to
31:46 come there's something there I think
31:48 overall yeah we're
31:50 curating wrent large when we put our
31:53 name on it then I think we have to say
31:55 hm what do we what will our language be
31:57 yeah I so a couple other things just I
32:00 just wanted to say Kevin to your point
32:02 about the financial in 2027 I with the
32:06 applied fatures lab and the threat
32:08 casting Lab at ASU and um we put out for
32:12 um the Secret Service a threet casting
32:14 report on the future of um uh digital
32:18 Financial crimes and what happens when
32:21 we no longer have paper currency and the
32:25 alignment with everything you just said
32:28 I just thought instantly oh my God I got
32:30 to share that report with this group
32:32 because it's like a
32:33 foreshadowing in not really because we
32:36 all knew this but you know it's great to
32:38 sort of see it's like
32:40 yep um and then I want to get to the
32:43 point of why I even wanted a question
32:45 which was to share with those who aren't
32:49 as familiar with how GPT in particular
32:53 is working and how you can make your own
32:55 gpts right now um I just I wanted to
32:59 emphasize that even more especially for
33:01 the CE folks who are trying to figure
33:04 out how to you know efficiency in
33:05 business all the way around if you have
33:08 not built yourself a a GPT for your
33:13 company which if you're not in four
33:15 already please go get on the wait list
33:17 they're going to open it again but get
33:19 yourself on the list it's closed right
33:21 now because of the chaos what was it
33:24 what was it again what chat GPT
33:28 get over open AI get over get yourself
33:30 on the list so that when they're
33:32 starting to issue the $20 a month chat
33:35 yeah yeah because it's closed right now
33:38 it's closed but just get on the waight
33:39 list so here's I just want to express
33:42 and this is why you guys got to go file
33:44 file sorry follow Kyle all the time
33:49 so because he's teaching on these
33:53 nightly Tik toks and it has twice a
33:55 month salons how to do all these things
33:58 so you don't have to go figure this out
33:59 there's a giant community that has come
34:01 together just come meet up with all of
34:04 us we're already doing all this you
34:05 don't have to figure this out on your
34:06 own here's what I wanted to say though
34:09 for everyone in
34:10 C we don't know how this is going to go
34:13 with building your own GTS so I cannot
34:17 encourage everyone enough if this ends
34:19 up going like the App Store you want
34:23 your name now yep get over there and
34:26 build your yourself a GPT as soon as you
34:29 gain access with your business name so
34:33 as an example I've built about 10 gpts
34:36 to this point and I'm training them with
34:38 my own content uh a couple of them are
34:41 public I I've shared them out and I can
34:43 share them out further you can go and
34:45 play with my
34:46 gpts they are the applied futurists and
34:50 the threat casting analyst and they're
34:53 being loaded with my reports my books my
34:56 content I'm training them I'm working
34:58 with them to get them to work with me
35:00 first but will that ever be out in the
35:02 public to everyone to go and use mine I
35:06 don't know but am I going to be ready if
35:10 that does happen Hells to the yes y so I
35:14 really really want to impress upon this
35:16 entire Community go get your name build
35:19 a GPT so you own it that's number one
35:23 and then second Kyle talks about this a
35:25 lot we're not going to have time to get
35:26 into it today but if you're not in GPT
35:30 and once you get in there make yourself
35:32 a note today when you gain access one of
35:36 the first things after you just play in
35:39 basic get into the custom instructions
35:43 go watch a video or ask chat itself how
35:47 do I use custom instructions actually
35:48 it'll just tell you how to do it there's
35:50 a sidebar you can fill why is that so
35:54 important because you can load it with
35:56 the things that that benefit your
35:58 business quick example when you go to a
36:01 my applied futurist GPT it's focused on
36:05 future casting so if you try to ask it
36:08 something about dystopian Futures it'll
36:11 go oh come on this topas for third
36:14 graders that's easy we're GNA stay
36:17 focused on
36:20 protopia but but if you're really into
36:24 like threat Futures here's a link to the
36:28 threat casting book y go get more for
36:32 yourself so I'm not losing those
36:34 potential people I'm just sending to
36:36 another my business so I just and and I
36:40 do that throughout so my I have custom
36:42 instructions and then I have back end on
36:44 my gpts that are deeper that do things
36:48 like if somebody asks a question I make
36:50 sure they always reference me my team my
36:54 partners as the experts as the go go to
36:57 go to their books go to their websites
36:59 so it's loading completely so that if
37:03 this ever does go super
37:06 public we're the leaders Y in this yep
37:10 and that's what I wanted to express let
37:12 let me just jump in really quick someone
37:13 on Tik Tok asked where do I get on the
37:15 the wait list so if you go to chat.
37:17 open.com that's chat GPT that's the
37:20 official chat gbt website just go there
37:22 and start playing and just sign up for
37:24 the wait list there for for access to
37:26 GPT chat GPT plus is what you want want
37:29 to get on the weit list for so yeah that
37:32 that's that's awesome Cindy I I mean I
37:35 would agree and you know I think that
37:38 creating these gpts I think using it as
37:41 a marketing tool is really really
37:43 brilliant because you know there there's
37:45 a lot of talk always in marketing about
37:47 deliver value first right deliver value
37:50 first and then sell your stuff after
37:52 with these gpts you can absolutely do
37:55 that you can put into your potential
37:57 customers hands a tool that they can
37:59 interact with and ask questions of and
38:01 and you know kick ideas around future
38:04 casting with that tool and then you know
38:07 the ability to just say now if you want
38:08 to talk more reach out to me I I think
38:11 that's brilliant and and I think even if
38:14 this even if these gpts don't end up
38:17 going mainstream and it doesn't become
38:19 the App Store just the process of
38:22 experiencing what it's like to be able
38:24 to
38:25 informally create an incredibly powerful
38:29 application is just it's just a
38:31 different experience than you know I'm
38:33 not a programmer but I know enough about
38:35 programming to appreciate what it takes
38:38 to make even a mediocre application
38:42 right and and with this tool you can
38:44 essentially just talk into existence you
38:47 know a deeply powerful tool that is
38:50 specialized to the thing that you want
38:52 to talk about so yeah fully fully uh
38:56 reinforce or you were saying
38:58 Sydney fantastic so I'm uh Doug hullen
39:02 has been with us and Doug I'll uh
39:06 apologize you had a pretty interesting
39:08 panel yourself yesterday that I was
39:10 unable to attend I hope there's a
39:11 playback for it yeah yeah there will be
39:14 it was on um AI human hallucination and
39:18 AI hallucination and we had Dr Harvey
39:21 Castro who is a medical doctor and
39:24 written books on AI and medicine and uh
39:27 and then some um doctors of
39:30 Neuroscience um please don't hallucinate
39:33 today be lucd for us today all
39:37 right I'm getting old and confused I
39:39 thought that was what was happening
39:41 today so I love to see the recording
39:44 yeah yeah and I I'll send it out once
39:47 once that's available but uh yeah we it
39:50 was a great talk and um you know just
39:53 exploring what Hallucination is and you
39:55 know both from a human perspective um in
39:57 fact I put it to the chat um I was also
39:59 on another call um talking about media
40:02 and AI you know and there was a a person
40:04 from the BBC looking at Ai and H you
40:08 know a using AI um for the news
40:11 reporting and there's a media
40:12 organization that uh has put out a a
40:15 10-step guide on integrating AI um into
40:19 you know how should media people and
40:22 authors think about how to use uh AI U
40:25 as a tool right and and this challenge
40:28 of
40:29 hallucination um you know and uh you
40:32 know but this this ability to just
40:34 create a vast amount of content and um I
40:37 you know I I one of my favorite books is
40:40 uh the information a history a theory of
40:42 flood by James Glick and you know all
40:45 this vast information out there and he
40:48 wrote this about 10 years ago and now
40:50 there's you know what we thought was Big
40:53 10 years ago you know it's just just
40:55 amazing and though the question is is
40:57 how can we have accurate data right act
40:59 accurate actual accurate information um
41:03 and this will be the challenge and you
41:05 know like the different uh you know like
41:07 large language models like Claude 2.1
41:10 came out and they said there there's
41:11 twice as good at reducing hallucinations
41:14 and getting actual information right um
41:17 you know so this is this race to the
41:19 like how do you make sure that what
41:20 you're creating is is factual is good
41:24 information um Dave C um and I'll put
41:27 the link in there too he has a company
41:30 um that tries that has an and going back
41:32 to this agent um where it acts as a
41:36 researcher and it gives some really good
41:37 information so like like for instance
41:39 you know like looking at the term
41:41 hallucination um in
41:43 2023 um Boston uh or cambrid dictionary
41:48 um is the AI Hallucination is actually
41:51 hallucination the word of the year in
41:53 2023 and so I asked Google bar
41:57 um well when did AI hallucination become
41:59 something and uh you know they said well
42:02 it was first coined in 2022 but then I
42:05 used Dave's tool and it was actually
42:07 coined in
42:09 2002 um by a research and uh and I had
42:13 done I tried to look at it and and do so
42:15 and I don't know if you use barard but
42:17 what you can do is click on it and say
42:18 okay tell me the things that you think
42:19 are more factual and less factual and
42:22 you get green if it thinks it's more
42:23 factual you know more more PR precise
42:26 and uh you know it was green saying it
42:29 it was coined in in 2022 but actually it
42:31 was coined in 2002 yeah yeah this is
42:34 going to be an issue I'll tell you just
42:36 a f a funny thing um Miriam Webster's
42:39 word of the year is
42:41 authentic um so so if if one of them is
42:45 uh is hallucination and the other one is
42:47 authentic there's uh there's something
42:49 there I think authenticity is going to
42:51 be big moving forward I'm going I if I
42:53 can just tell it I'm wondering how much
42:56 this matters
42:57 because because the the here's a here's
43:00 a joke that I heard you know the Obama
43:02 and and and Trump get into a foot race
43:05 and it's going to be you know and Trump
43:07 is talking about how great it's going to
43:08 be and how he's going to you know do so
43:11 well and then they have the foot race
43:13 and you know what happens obviously and
43:15 and so uh OB uh Trump posts afterwards
43:19 he said man this was a great race I came
43:21 in second and Obama came in next to
43:25 last
43:27 it's all about shaping The Narrative
43:28 right well that but both of those are
43:31 true statements right yeah so so the
43:33 idea of when when we're talking about
43:36 truth and and what is real and what is
43:39 is is hallucination it so much depends
43:42 on point of view and I I don't subscribe
43:44 to the information overload and all this
43:46 kind of stuff I subscribe to it matters
43:49 in the context of whatever your context
43:52 is and that that you know if you if you
43:54 want to protect the US certain things
43:56 matter in certain ways if you want to
43:58 protect China certain things matter in
44:00 certain ways um or don't etc etc and and
44:04 it's very we we get hung up on what is
44:08 what is true and real and not real it
44:11 doesn't matter so much yeah critical
44:13 thinking is absolutely critical you know
44:16 that that's probably the number one
44:17 thing you need to teach them you think
44:19 that's a totaly right
44:23 there so the um
44:27 you know
44:28 we're I'm interested uh in your
44:33 perspective Doug
44:35 about
44:37 um how how is the infrastructure part
44:41 helping or hurting in terms of access
44:45 right um because if
44:49 we Tim uh bearon who was with us earlier
44:53 on the call he's still on but uh wrote
44:56 something
44:58 about Ai and know you know being very
45:02 helpful in in terms of making us you
45:05 know augmented and more
45:07 superhuman but only works if you're
45:09 connected it only works if you have
45:11 access right so where are we right now
45:14 you're you're kind of a you know
45:15 infrastructure yeah the good news is
45:17 there in 2022 we reached five billion
45:21 people that had a smartphone wow and
45:23 like in 2002 we had a billion people
45:26 that had just a cell phone so um you
45:29 know we we reached that and like I have
45:31 this blog going is when do we reach a
45:33 billion of something so next year I
45:35 think a billion people will using AI at
45:38 least 30 minutes a day and saving at
45:40 least 30 minutes a day you know that
45:42 that they can say get an agent to say I
45:44 want to do X and you do that and I come
45:47 back and I fact check you I critical you
45:49 know but it's it's it gives me value of
45:51 at least 30 minutes but if I'm looking
45:54 at the population in general is it fair
45:57 to say that three fifths of the people
46:00 on the planet have access and two fifths
46:02 are you know disenfranchised right now
46:05 um well since five million PE billion
46:07 people have a
46:08 smartphone uh you know the ability to
46:11 have ai um like Pi for instance is free
46:15 um you know and it's it's a voice
46:17 activated right now the question five
46:19 five billion out of what 8.5 eight
46:22 billion actually fun fact is in in
46:25 November of 20 22 we reached 8 billion
46:28 people on the planet at the same time uh
46:30 generative AI chat GPT was launched okay
46:34 in the month so so my my fraction is
46:36 slightly off but you know there
46:39 are
46:41 people you know take out a billion take
46:44 out a billion kids who were under the
46:46 age of whatever three or five you know
46:49 say that's a billion and and then that
46:52 that that's a huge percentage that's
46:53 like five five out of out of seven
46:56 now there gets to be a big percentage of
46:59 of people who have access yeah well a
47:01 billion people don't have access to
47:03 electricity in fact more people have a a
47:05 cell phone than they do have
47:08 electricity yeah I'm just pointing out
47:10 that you know the uh you know talking
47:13 about you know gaps this you know
47:16 because it's exponential we we just said
47:18 this is exponential technology the
47:21 ability to leave a percentage of
47:24 humanity in the dust of okay uh is is
47:28 problematic okay um you know we're we're
47:32 we're we should be talking uh about you
47:35 know some of this uh in the human rights
47:39 territory all right as opposed to just
47:42 you know it would be interesting if we
47:43 could eventually get get to them right
47:46 because people who are augmented versus
47:48 those who aren't okay is going to be a
47:51 very bright dividing line going forward
47:53 but I I'll stay on the positive side all
47:56 right
47:56 the purposes of birthday
47:58 [Laughter]
48:00 celebration okay um but Kevin can I just
48:04 add to that come it's it's not just that
48:08 people uh that we're leaving them dust
48:10 because they don't have the stuff
48:11 because many of the people who don't
48:13 have it still know about it and are
48:15 aspirational and will eventually get it
48:18 the the interesting problem is that
48:20 people are choosing to be LIF left in
48:22 the dust they're choosing to be
48:24 non-enhanced and they're thinking that
48:27 somehow it's better to be an unenhanced
48:30 human than to be an enhanced human and I
48:33 I see these is you know I see this as a
48:35 positive growth and that this is It's
48:38 Evolution and there's nothing we can do
48:40 about it because we're all going to be
48:41 that way eventually it's a new body part
48:44 but but a lot of people don't see it
48:45 that way and that to me is a bigger
48:47 issue but that's that's Mark that's
48:49 that's every every technological
48:51 advancement in the history of man is is
48:55 met with the exact same response which
48:57 is fear and trep trepidation and
49:00 resistance you know we we don't like
49:02 change and it this is just another one
49:04 of those but this one's more intense
49:06 because the capability of it is is so
49:09 profound and the speed that it's moving
49:12 is it's it's not like anything I've ever
49:13 seen so um so I think the the resistance
49:16 is normal um but but the uh for those
49:20 people resisting the the amount of uh
49:23 sort of impact it's going to have on
49:25 them very quickly I think is if you
49:28 wanted to be a l though all right there
49:29 was a there was actually a factory to go
49:32 you know mess up right you know take an
49:35 axe to yeah uh there is no axe all right
49:38 to take to this Tech right there there's
49:42 no um you know Central Point you know to
49:46 do anything and the government will
49:48 regulate
49:49 it totally but my perspective I totally
49:53 agree with with with Kyle just said but
49:55 what what's happening is that because
49:57 this is moving so fast faster than
49:59 anything ever has the it's needs to get
50:03 to the young people earlier and sooner
50:06 and that's that's the barrier that I see
50:09 as happening uh being prevented
50:12 by often adults by by schools by
50:15 families by cultures by all these kinds
50:17 of things and so I really want to see
50:20 how we can do that and and I'm open to
50:22 suggest I think that would be a great
50:24 project for all of us to say how do we
50:27 provide how do we create things or or is
50:30 there something that we can create that
50:33 that would really make kids say hell yes
50:36 I want to do this and I don't know of
50:38 any but there may be uh that there's you
50:40 know some site you can go on and
50:42 suddenly you're a magician suddenly
50:44 you're 100 times more powerful uh that's
50:46 not you know that's specifically geared
50:48 to to you um I don't know if we could do
50:51 that but that'd be great okay so I'm
50:55 going to invite uh I see a comment here
50:59 from Ala Roy ala you want to join us for
51:02 a second about your perspective on AI
51:06 [Music]
51:07 hallucination you're
51:09 there maybe you're not can you hear me I
51:12 can hear you okay great
51:16 um yeah I I think it's just such a
51:20 interesting time in in many ways as
51:23 you're discuss discussing this and I put
51:25 a question for this team as well um
51:28 because there seems to be just this real
51:32 um urgency for adoption um I feel that
51:36 you know a group of people are almost
51:38 working for like um as a chat GPT or
51:42 open AI agents like oh it's the best
51:44 thing you know you must adopt it I think
51:46 it's great to play with I think it's
51:48 wonderful to use and you know but I
51:50 think it's
51:52 also important for us to bring as
51:56 someone else mentioned earlier the
51:59 critical thinking like looking at it
52:02 looking at the the game behind the game
52:05 and the layers behind it to understand
52:07 what's happening just like you would do
52:09 with any marketing machine right so I
52:11 think of Chad gpt's writing I think Mike
52:14 or somebody was talking about you know
52:16 editor or someone was talking about
52:17 being a curator
52:20 as being a really you know generating
52:24 writing that's
52:26 the the marketing speak the pr speak
52:28 that we see which means you have to
52:31 really see what is real out of it what
52:33 is useful out of it what makes sense in
52:36 here and the term AI
52:39 hallucination I offered an article it's
52:42 a little older but it's really
52:44 interesting I think you should just take
52:46 a read on it is that
52:49 um I was I I don't think that we should
52:53 have allowed the term AI hallucination
52:56 to be used because I think it
52:59 legitimizes the problems that are
53:01 underlying in this technology um calling
53:06 it hallucination gives it some sort of a
53:09 u you know it sounds uh
53:12 cool it gets it legitimacy of some sort
53:15 uh whereas it's really
53:18 uh incorrect they are errors that that's
53:21 a problem you know it's like calling
53:23 someone's lies uh politician talk you
53:26 know someone else gave an example of oh
53:28 that's political talk but they're lies
53:30 you know lies are to be called lies um
53:33 so I think that's that was my kind of
53:37 comment on on
53:38 hallucination I think will tell you that
53:41 you
53:42 know becomes mimetic right is that you
53:45 simply have to launch a sufficiently
53:48 competitive meme to replace it um I've
53:52 got one you're not going to you're not
53:54 going to succeed in in outlawing
53:56 language right it doesn't work right
53:58 there
54:00 um you can certainly raise your you know
54:03 you can certainly raise your hand I see
54:05 um Andrea Goldberg has some uh thoughts
54:10 that she wants to share Andrea join us
54:13 let let me just jump in let me just let
54:15 me just jump in for a quick Andre let me
54:17 just jump in for a quick sec so the the
54:19 the potential um counter to
54:22 hallucinations my nickname for chat GPT
54:24 is mansplain as a service uh it will
54:27 confidently answer you even if it's
54:29 wrong so you're you're
54:32 welcome so thanks for joining us yes
54:36 please so one of my concerns is that you
54:39 know the ne those of us who are First
54:41 Learning chat TP learned the basics so
54:43 like if we're marketeers we learned how
54:45 to write marketing plans wearn learned
54:48 how to do the things and now we have the
54:50 ability when we get something back from
54:52 cha GPT to say this is garbage curate
54:55 edit whatever term you want to use my my
54:58 biggest concern is to the extent the
54:59 next generation of workers comes in and
55:02 everything's done for them you just
55:04 speak are they going to have the ability
55:07 to fully understand what they're getting
55:09 back and what's quality and and what's
55:12 not quality and I saw this personally
55:15 when we first started you know
55:17 automating things and people started you
55:19 know writing you know having things done
55:21 for them and I saw our analysts
55:24 misinterpret data because they didn't
55:26 realize the program had lost 90% of the
55:29 cases or you know looked at the wrong
55:31 cohort and they they took it and I ran
55:34 with it I just I think it's important I
55:37 don't want to say we shouldn't um use
55:39 the technology but I think we have to
55:41 think about what skill sets do we want
55:43 to make sure that we don't lose along
55:46 the way say you met people mention
55:49 critical thinking but I I I I what
55:51 enables you to do the critical thinking
55:54 is kind of experience doing some of
55:56 these basic um executions so I wanted to
55:59 you know bring that up if you if you do
56:01 it uh and use the technology often
56:04 enough and are met and the kpis are
56:07 telling me you know that you're failing
56:10 okay it doesn't work right you're going
56:13 to lose your job right and there's going
56:15 to be a consequence all right to you
56:17 know using the technology poorly as
56:20 opposed to enabling something that's
56:22 spectacular Mike you know this is your
56:25 territory all right you were you know in
56:28 this world all right and you can't be a
56:32 good analyst or a good you know
56:35 marketing person using this Tech if
56:37 you're using it all right in an
56:39 unsophisticated way you're you're just
56:40 going to you're going to crash and burn
56:43 I AG well I I I see Andre's Point Andre
56:47 I agree with you I think here again what
56:49 kept struck me was you know what's going
56:51 to protect that new worker or what's
56:53 going to help that new worker get that
56:55 experience getting his or her ass kicked
56:58 by their boss with that output yep right
57:02 that's how and now back to this thing
57:04 the more I think about this is really
57:07 interesting
57:09 managers
57:10 who as we I think many of us have
57:13 experienced in the last 10 or 15 years
57:16 have moved from let's face it more
57:19 direct handson and not because of the
57:21 pandemic this was happening before that
57:24 I don't need to talk to Andrea that much
57:27 because I have this console over here
57:30 that has all these kpis and metrics that
57:32 we have with with Andrea so I can just
57:35 look at that hey everything's great
57:37 something's a little bit yellow I guess
57:39 I better talk to her this week right
57:41 with the ability to create and generate
57:44 so much so
57:46 fast this is going to actually I think
57:49 take management back to what they were
57:51 doing more in the you know when I
57:52 started in the 70s ' 80s and '90s and in
57:56 that point which was I'm gonna have to
57:58 actually talk to Andrew to make sure she
58:00 understands exactly what that output did
58:02 yep and it's going to be and it's going
58:04 to be no screens Andrea I want you to
58:06 take your phone and turn it upside down
58:08 on the desk we're put both of our phones
58:10 in the middle of the desk and we're
58:11 going to talk because that's how it's no
58:14 matter what you know that's still in the
58:16 end I have to feel as your manager I
58:19 have to feel like hey it's cool she
58:21 didn't know this tool that much and
58:22 she's new in this industry but boy
58:25 everything is [ __ ] looking pretty good
58:27 great more like that oh wait I you
58:31 misinterpreted your input your
58:34 prop showed a misunderstanding of what
58:36 that the new version of the product
58:38 actually does let's talk about that a
58:39 little bit more yeah right actually it's
58:42 G to more add more time I think this is
58:44 something I think you heard steuart nois
58:47 I a few months ago said no is it does
58:49 stuff fast but it's actually taking me
58:51 longer because I got a vet to try to
58:53 figure out and triangulate where did
58:55 this information come from where that
58:57 yeah I'll be I agree with you and the
58:59 car L is when social media first came
59:02 and people were assigning 25 year olds
59:04 to to do all the social media all the
59:07 marketing All The Branding for their you
59:10 know their company and they discovered
59:12 those 25 year olds didn't understand
59:14 their brands and didn't understand how
59:16 to phrase things and it was a disaster
59:18 they understood the technology but they
59:21 didn't have the basic understanding of
59:24 marketing communic
59:26 and they got into some embarrassing
59:28 situations so I think we have to make
59:30 sure that people understand the broader
59:32 concept here that to your point Mike
59:35 it's going to create a different type of
59:36 management well I'll so let me throw in
59:39 a counter to let me throw in a counter
59:41 to this Andrea which is I my my nickname
59:44 for for one of the things going on right
59:46 now is Revenge of the Gen xers which is
59:50 um I like I feel like if I were hiring
59:53 someone right now I would I would rather
59:55 hire someone like yourself someone
59:56 that's got 20 or 30 years experience or
59:58 someone like Mike that has enough World
1:00:01 experience and business experience that
1:00:04 even if you don't have the skills with
1:00:05 AI you're going to have the skills to be
1:00:07 able to look at the output of AI and say
1:00:09 oh that's good that's bad and you're
1:00:11 going to be able to know it like that so
1:00:13 I think that there's a superpower for
1:00:15 older workers that might even be
1:00:17 Reinventing themselves or stepping back
1:00:19 into the workforce that that you know
1:00:22 they might not be experienced on the
1:00:24 skills side but that experience I think
1:00:26 is going to be incredibly
1:00:28 valuable and I would add to that if if I
1:00:32 may that the the making those pairings
1:00:35 of somebody like that with somebody
1:00:37 young um and facilitating those pairings
1:00:40 is a is a huge thing just like uh you
1:00:43 know Jack Welch did a GE when he said we
1:00:46 need we that the young people will
1:00:48 mentor and executive but but what what
1:00:51 I'm interested in Andrea is is your
1:00:54 assumption that we did this well before
1:00:56 that we were critical thinkers before
1:00:58 and and it isn't true it isn't true some
1:01:01 of us understand some things in some
1:01:03 ways where there's no consensus about
1:01:05 how we understand things obviously we'd
1:01:08 all be one thing but and so I I think
1:01:11 it's much more important to to
1:01:13 experiment and there'll be a lot of
1:01:15 upheaval as people don't do the same
1:01:18 understand things in the way they did in
1:01:20 the past and do the same things they way
1:01:22 but I think it's going to come out as
1:01:23 just as Kevin says in the wash in terms
1:01:26 of do does it work do people accept this
1:01:29 stuff who accepts it who are who at the
1:01:32 level you know it doesn't matter whether
1:01:34 I don't think somebody's critically
1:01:36 thinking in my in my definition of it um
1:01:41 if if it was all working in their world
1:01:44 and doing this kind of stuff it doesn't
1:01:46 matter that my I taught my son I tried
1:01:49 to teach my son look people in the eye
1:01:52 when you meet them right and that's a
1:01:54 big part of life and he wouldn't do it
1:01:56 he still doesn't do it he he he looks
1:01:58 people at the screen and so that but but
1:02:01 that matter okay they uh add on to uh uh
1:02:08 Mac right uh who has uh joined us uh M
1:02:11 you have a question not really a
1:02:14 question but a comment and follow up
1:02:18 on take a system Dynamic thinking
1:02:21 approach to this we have a generation of
1:02:24 EXA
1:02:25 who are not seeing upline growth so they
1:02:28 have to make their numbers by cutting
1:02:31 expenses and frequently it is the more
1:02:34 more senior and higher paid employee
1:02:38 that gets cut yep so in the situations
1:02:42 that we're talking about that manager
1:02:45 that is meant to be coach and help make
1:02:48 sure the information is correct they're
1:02:50 not going to
1:02:53 exist and then that problem gets set up
1:02:57 where you have very inexpensive and
1:03:00 perhaps younger talent that can't
1:03:04 differentiate between what is accurate
1:03:07 and what is
1:03:08 competent and I think that's the
1:03:10 scenario we are setting up and we're
1:03:12 going to have to watch it
1:03:16 carefully okay very
1:03:20 good all right um we're coming up to the
1:03:22 top of the hour so um G to ask uh Mike
1:03:26 uh do you have any concluding comments
1:03:28 and then I'm going to move it to uh Kyle
1:03:31 to uh give us his final thoughts no what
1:03:34 a what a great discussion and thanks to
1:03:37 everybody on Kyle's Tik Tok Channel and
1:03:39 everybody else who joined um you know
1:03:42 the more as we went through that that is
1:03:44 something I think I'm going to be
1:03:45 thinking a little bit about the next
1:03:46 couple of days is you know what does
1:03:48 this mean the management process from a
1:03:50 small to a large Enterprise based on
1:03:52 some of the things we discussed uh but
1:03:55 gosh what a great discussion Kyle over
1:03:57 to you yeah so yeah thanks like f
1:04:01 fascinating discussion and and I for for
1:04:04 me the critical the critical thinking
1:04:07 piece that was referenced in in a number
1:04:09 of different context in the
1:04:12 conversation that to me feels like that
1:04:15 is the skill of the future and to some
1:04:18 degree that that our ability to sort of
1:04:21 look across disciplines and know that
1:04:25 the the the machines are going to
1:04:27 increasingly execute the Tactical pieces
1:04:30 and what's required is okay what do I
1:04:33 actually want to accomplish what do I
1:04:34 actually want to do um the the other
1:04:37 thing I would say is
1:04:38 this the the all of the conversations
1:04:41 we're having today about the limitations
1:04:43 of AI is a snapshot you know where we
1:04:46 are today this stuff is moving so fast
1:04:49 that what what I would encourage you to
1:04:50 do is get curious about these tools um
1:04:55 interact with them and even if you even
1:04:57 if you're of the mind like well it's not
1:04:59 really ready for prime time now check
1:05:01 back in 3 months because the the the
1:05:04 amount of features that get added to
1:05:06 these tools and just improvements that
1:05:08 get added to these tools kind of the fly
1:05:10 under the radar are dramatic so if
1:05:12 you're not actively engaging with them
1:05:15 you may miss that so there might be
1:05:16 something that actually solves your
1:05:18 problem that comes along very very
1:05:19 quickly uh and you and you sort of let
1:05:21 it pass by for a year or two there's
1:05:23 going to be a lot of missed opportunity
1:05:25 so I would just say get curious you
1:05:27 don't have to love it but I would say
1:05:28 get educated about uh what these tools
1:05:31 can do where the current boundaries are
1:05:34 and and you know what they're good at
1:05:35 what they're not good at and then use
1:05:36 your critical thinking to say okay
1:05:39 here's the 80% I can use them for today
1:05:41 and the 20% I'll do the way I did it
1:05:43 before and just pay attention to that
1:05:45 balance over time so yeah yeah may I say
1:05:48 one more thing uh briefly okay critical
1:05:53 thinking I think is is is is a wrong
1:05:54 term I think it's just it's thinking we
1:05:57 have we had slow thinking which is
1:05:59 analytical thinking and what some people
1:06:01 calling critical thinking we have fast
1:06:03 thinking which is heuristic that Conan
1:06:05 pointed out now we have algorithmic
1:06:07 thinking it's all thinking it all has to
1:06:10 be integrated by all of us and if you're
1:06:13 not using one of the ways of thinking I
1:06:15 think that you're missing it okay and my
1:06:20 you know thought uh as we you know you
1:06:23 know blow out the CLE on you know that
1:06:26 single candle on top of the GP chat GPT
1:06:29 birthday cake is when you first us this
1:06:32 it's actually fun yeah right I mean
1:06:35 we're talking about you know uh now
1:06:38 taking it you know how are you going to
1:06:39 use it you know how is
1:06:41 this I would just suggest to you that um
1:06:45 anyone who is you know mildly you know
1:06:49 put off by this it's just given an
1:06:52 opportunity to use the one that you've
1:06:55 already signed on to for a moment say
1:06:57 ask it to do something right because the
1:07:01 initial experience is oh that's
1:07:04 remarkable that's cool right um do you
1:07:09 also get a little tingle that oh you
1:07:11 know this is
1:07:14 uh this is going to be you know big yeah
1:07:17 yeah it is right but you should go ahead
1:07:20 and try to expose people to the fun part
1:07:23 initially and if they can at least have
1:07:26 that experience as a craft right the Art
1:07:29 and Science comes later so I would you
1:07:33 know just give somebody a I did that
1:07:35 with with with Heidi the other day
1:07:37 Heidi's my wife uh and she
1:07:40 goes I didn't know you could do that and
1:07:43 it was the chat gbt 4 now connected to
1:07:46 Dolly that you can ask it to make a
1:07:48 picture right I mean it's really cool
1:07:51 it's amazing can I close this out with
1:07:53 the Jerry bir see a quote Kevin about
1:07:55 exactly what you link to what you said
1:07:58 uh he was one of the guitar first guitar
1:08:01 players that could regular was using
1:08:03 digital effects built into his guitar
1:08:05 not a boatload of pedals these are the
1:08:08 things you guys see stomping on the
1:08:10 stage and one time uh one of the
1:08:13 songwriters John Perry Barlo came into
1:08:15 the rehearsal studio and he was Jerry
1:08:18 was practicing and he had a new midi set
1:08:20 up where it was a he's playing the
1:08:22 guitar but it sounds like a
1:08:25 mhm right that's what it is and the guy
1:08:27 says
1:08:28 wow man that's you wow you could be you
1:08:32 could have been a really good trumpet
1:08:34 player and Jerry kept play and looked at
1:08:35 him and said I'm a great [ __ ] trumpet
1:08:39 player right now and that's exactly what
1:08:41 where he's able to say to your point I
1:08:44 can
1:08:45 now I can make the sound like a trumpet
1:08:48 with some of these effects with my
1:08:49 Strings On My Guitar I'm a terrible
1:08:52 trumpet player cuz I'm not a very good
1:08:54 gu guitar player sure and so I mean that
1:08:56 transference I think and adapting to
1:08:59 that new thing to where Jerry Garcia is
1:09:01 a trumpet player even though he never
1:09:03 played it y That's the kind of I think
1:09:06 the joy you're talking about the fun
1:09:08 right let find that at some point fun is
1:09:11 always going to be like dessert so maybe
1:09:13 we then get some fun and we learn hey
1:09:15 wait a minute because I just did
1:09:16 something with that I sent to Kyle that
1:09:18 I didn't think it could do yep exactly
1:09:21 right which is once I programmed it I
1:09:23 said wait a minute that doesn't sound
1:09:25 like me and it went back and did another
1:09:28 take yeah and it was pretty good so
1:09:30 sorry for the bad
1:09:33 language thanks everybody for being with
1:09:36 us if you were a guest today and you you
1:09:39 like the the format of this particular
1:09:41 call uh let Kyle know or you can write
1:09:44 to C content evolution. net and we'll
1:09:47 add you to the distribution list y so
1:09:49 everyone thanks so much for being with
1:09:51 us uh for this special edition of
1:09:53 content evolution new world and in
1:09:55 everything you do be intentional Kyle
1:09:57 thanks for being with us today thanks
1:10:00 cheers everybody happy birthday thank
1:10:01 you happy happy birthday chat
1:10:05 gbt recording
1:10:09 stopped all right well there you have
1:10:13 it that was fun that was fun all right
1:10:18 let's see how do
1:10:22 I all right I can do a few more minutes
1:10:25 here very cool thanks for sharing thanks
1:10:26 for thanks for hanging out there's a
1:10:28 bunch of people in here that hung out
1:10:30 watching watching a bunch of people
1:10:33 pontificate about the future of AI so
1:10:37 thank you for hanging out um let me go
1:10:39 get
1:10:44 my let's put chat GPT back front and
1:10:48 center happy birthday chat gbt um let me
1:10:51 flip this here
1:10:55 and do I have pictures open no hang on
1:10:58 uh photos let's see
1:11:03 photos hang on one
1:11:18 sec all
1:11:20 right there we go that's a good picture
1:11:26 we'll throw that picture there that's a
1:11:29 little happy birthday chat GPT
1:11:36 picture lovely lovely lovely lovely all
1:11:39 right so I didn't have a chance to
1:11:41 really I saw a couple of comments fly
1:11:43 through but I was obviously having to
1:11:45 pay attention to
1:11:48 uh to what was going on there so um if
1:11:51 anyone has any questions pop them in the
1:11:53 comments I know we've got some mods in
1:11:54 here that can pop them up um this is
1:11:59 um I'm doing 24 hours of a chat GP Aon
1:12:05 um I have to I'll have to leave
1:12:06 occasionally to go do my day job running
1:12:09 Story vine is it possible to access the
1:12:11 recording of This Zoom meeting yeah so
1:12:14 if you join the AI
1:12:17 Salon which is go to AIS salon. MN as in
1:12:21 Mighty networks. Co uh if you're not a
1:12:24 member of the AI Salon join the salon
1:12:27 and then um the mods posted the links to
1:12:31 the new world recordings in uh I don't
1:12:34 know where you posted
1:12:36 them where did you post them uh probably
1:12:40 in news is it difficult to learn prompt
1:12:43 engineering no so especially with chat
1:12:47 GPT well actually all the large language
1:12:49 models
1:12:51 it I I hate the term prompt engineering
1:12:54 I prefer the term prompt
1:12:56 crafting um it's really just talking the
1:13:01 the thing the thing that I think will
1:13:02 improve your prompt engineering the most
1:13:04 you said to post to news okay great so
1:13:07 if you go to the AI Salon in the news
1:13:10 channel um there's a link to the the
1:13:13 content evolution new world calls so
1:13:16 that Zoom call that we were just on the
1:13:18 recording of that will probably be up
1:13:19 tomorrow or the next day um so so the
1:13:23 link to it's in the Salon though uh
1:13:25 prompt articulation prompt
1:13:27 crafting here's the thing that that
1:13:29 shifted the the biggest for me is that
1:13:33 if you think about how you prompt Google
1:13:36 it it's sort of this one prompt in one
1:13:39 prompt out uh thanks Eric AIS salon.
1:13:44 mn.
1:13:47 um you you kind of have to get your
1:13:49 prompt right to get back a good search
1:13:51 result what happens with large Lang
1:13:53 anguage models is it's much more like a
1:13:55 conversation so you can either get
1:13:57 really good at prompt engineering right
1:13:59 and figure out here's all the components
1:14:01 of the prompt I need to make to give me
1:14:03 the exact right output and d da d da da
1:14:05 and give it nested and formatted and all
1:14:07 that sort of [ __ ] or you can just start
1:14:10 a conversation with chat GPT hey I'm
1:14:13 trying to figure this out can you help
1:14:15 sure blah blah blah blah blah and then
1:14:17 you you go back and forth with it and
1:14:19 give it more and more context so I would
1:14:22 say the thing that will improve prove
1:14:24 you're prompting the most is just do it
1:14:27 don't feel like you have to learn
1:14:29 anything don't feel
1:14:31 like you need to become a prompt
1:14:35 engineer if you're a human being with
1:14:39 language then you can interact with
1:14:41 these things now might people with an
1:14:45 engineering mindset do it better they
1:14:48 might I've got a a creative writing
1:14:52 mindset and I I do okay with prompting
1:14:55 do I follow all of the rules and I tell
1:14:58 it to act in this role and do this and
1:15:00 give me the output like this like prompt
1:15:02 Engineers do no I don't because I got
1:15:05 add and I don't give a [ __ ] I feel like
1:15:08 I feel like any specific set of rules I
1:15:10 learned today are going to be obsolete 3
1:15:12 months from now anyway so why would I do
1:15:15 that so for me it's just like just start
1:15:19 talking with these things start yeah
1:15:22 yeah Broody H digital chicken said you
1:15:24 learn after a while how do you learn go
1:15:29 play don't try to solve your biggest
1:15:32 problem right out of the gate like do
1:15:34 stupid
1:15:36 [ __ ] what what meal can I make with the
1:15:39 crap in my refrigerator and with chat
1:15:42 gp4 you can take a picture of what's in
1:15:44 your refrigerator and it will analyze it
1:15:46 and then write you a recipe based on
1:15:49 what it sees so start playing just start
1:15:53 playing
1:15:54 chat GPT can now use um can now use
1:15:59 images as prompts so so a good prompt
1:16:03 might be the Whiteboard session that you
1:16:05 just
1:16:07 had um let me just make sure I'm not
1:16:10 missing anything at work there's one
1:16:12 thing I need to
1:16:17 do
1:16:19 um all right so wait what time is it
1:16:23 I've got uh let me do another I'll do
1:16:25 another 20 minutes here I sent a card to
1:16:28 Sam Alman with close
1:16:32 enough that's
1:16:34 good that's very good hopefully close
1:16:37 enough was
1:16:39 misspelled or I didn't see it was Tiny
1:16:42 text so you might have actually made
1:16:43 that
1:16:46 joke
1:16:48 um they will become better at natural
1:16:50 language you mean computers Pate so Pate
1:16:53 works at Google
1:16:55 so the T-shirt rack just sent my request
1:16:59 to join AI Salon okay great let me go
1:17:03 look over there I don't know if anyone
1:17:05 is letting people in because it's invite
1:17:09 only so let me go
1:17:18 [Music]
1:17:22 look
1:17:30 all right so people that were waiting to
1:17:32 be let in you are in all right fantastic
1:17:37 fantastic Bob tell him what he's want he
1:17:40 you want a new car you want a new car
1:17:42 that's for sure mhm yeah B jingus it's a
1:17:46 uh it's a Mercury Topaz it's a 1997
1:17:50 Mercury Topaz that's free and clear we
1:17:53 we've we've paid for all of the fees
1:17:56 we've we've paid for them all and I'll
1:17:58 tell you what we're going to give you a
1:18:00 $50 gift card to fill up your tank of
1:18:02 gas with that Mercury Topaz that's yours
1:18:05 that's free and clear good to go you're
1:18:07 going to have to pay the taxes though
1:18:09 okay all right fantastic we picked that
1:18:11 up down at the auto trader and uh it's a
1:18:14 it's a gem of a car it still runs good
1:18:17 it still runs it's a lot of miles it's a
1:18:19 lot of miles on it there's a smell of a
1:18:22 it's a it's some sort of feline it might
1:18:24 be a feline kind of situation uh or
1:18:27 squirrels uh but boy it's a gem of a car
1:18:30 so you enjoy that you just take it you
1:18:32 drive it wherever you want for it's
1:18:34 probably good for about 50 miles before
1:18:36 before it overheats but uh that Mercury
1:18:38 Topaz it'll just keep running and
1:18:40 running all right do you know how to use
1:18:42 chat GPT and Dolly wait do you know how
1:18:45 using chat GPT for and Dolly
1:18:49 works with using those images online for
1:18:53 for your company yes use
1:18:57 them use them I mean if you generate an
1:19:01 image with a Coke logo on it I wouldn't
1:19:03 use that you know you need to follow
1:19:06 basic basic uh things but there there's
1:19:08 no um no one's going to come after you
1:19:11 for using a generated image image thank
1:19:14 you so much for letting me in Kyle
1:19:16 seriously you're welcome Edson welcome
1:19:18 I've seen you you've been here for the
1:19:19 past couple of nights I think so welcome
1:19:21 welcome welcome and it's not night by
1:19:23 the way it's morning where where I
1:19:26 am uh booking tons of travel today used
1:19:30 a new
1:19:31 GPT I made that I'll post once I'm done
1:19:34 perfecting it oh that's super cool
1:19:35 Silver Fox I these gpts I
1:19:38 mean the the thing I think I'm most
1:19:41 excited about is people who are not
1:19:44 historically software developers are
1:19:47 going to develop stuff that software
1:19:49 developers would never have developed
1:19:51 right or just they'll just do it in a
1:19:52 different way
1:19:53 cuz they don't know any better right
1:19:56 they don't know the limitations so
1:19:57 they're going to figure some [ __ ] out
1:19:58 that we're going to be like oh I didn't
1:20:00 know you could do that that's a unique
1:20:03 way of thinking about it show us that
1:20:05 whiteboard what do I have on
1:20:08 it oh this is actually you
1:20:14 know let me see
1:20:22 something
1:20:33 um these are uh interface design
1:20:36 screens
1:20:38 for
1:20:40 um okay so my current company Story vine
1:20:44 which
1:20:45 is there's my little there's my little
1:20:47 trade show Banner we don't go to trade
1:20:49 shows anymore so I just I just have it
1:20:51 set up in my office
1:20:55 um so it's it's an automated uh video
1:20:59 storytelling platform so people use our
1:21:01 app to answer some questions in the app
1:21:03 and a few minutes later it makes a video
1:21:05 we've been around for 11 and a half
1:21:06 years it does not use AI to do that
1:21:09 automation
1:21:12 um but about a year ago we incorporated
1:21:15 open ai's whisper transcription into the
1:21:18 platform so for every video in the
1:21:20 system we now know what it says so what
1:21:22 these screens are
1:21:23 is we're building a system where you can
1:21:26 either take a Story vine video or you
1:21:27 can upload a video to it and within
1:21:30 about 10 or 15 seconds it'll
1:21:32 automatically transcribe the video it
1:21:35 then chapter IES the video so it'll
1:21:38 break it into its chapters its component
1:21:40 parts with time code it then writes a
1:21:43 title of the video a description of the
1:21:45 video and then there's a whole set of
1:21:47 buttons that you can push to generate
1:21:49 tweets a blog post about the video so so
1:21:53 um so that's what that is what else do
1:21:56 we have up here idea interview
1:22:01 people to let's see the surrender
1:22:05 experiment all right that's just
1:22:07 self-help [ __ ] everything here is ai ai
1:22:12 future road map [ __ ] for my company
1:22:16 yay um you did it chat GPT you're the
1:22:21 best all right it's his plans to take
1:22:24 over the world you're damn right maybe
1:22:26 do more lunch and learns while we're not
1:22:29 falling
1:22:30 asleep maybe I can I can maybe do that
1:22:33 I've got a I've got a fair amount of
1:22:38 um I got a fair amount
1:22:42 of weekly meetings but you know one
1:22:45 thing I could do I do a LinkedIn live on
1:22:50 Fridays called AI office hours so now
1:22:53 that I've sort of figured out how to do
1:22:55 this thing where I can be on Tik Tok and
1:22:57 in a in in another meeting at the same
1:23:00 time maybe I'll maybe I'll stream those
1:23:04 Tik Tok things and I'll figure out a way
1:23:06 to deal with the comments better CU I
1:23:08 didn't I didn't I just ignored you all
1:23:10 for the last
1:23:12 hour that's a lot of stuff okay thanks
1:23:19 guys you wrote chat PT on your
1:23:23 poster uh no chat GPT wrote chat PT on
1:23:28 the poster so one of the things image
1:23:30 generation in in all these image
1:23:32 generation tools does really poorly is
1:23:34 spelling so it got birthday right and it
1:23:39 got happy right so this's just a g
1:23:42 missing you can Photoshop that
1:23:51 in
1:24:01 [Music]
1:24:05 all right
1:24:13 uh I actually need to use claw right now
1:24:18 to do a data analysis report because of
1:24:23 gen generative AI Peeps are searching
1:24:26 Google differently they know they now
1:24:28 ask how and why and expect a gen AI
1:24:32 experience
1:24:33 interesting I don't know how Google
1:24:36 deals with this quite
1:24:38 honestly see CU
1:24:41 Microsoft Microsoft can lean in heavy
1:24:44 into this stuff because their primary
1:24:46 revenue is not from
1:24:48 advertising um Kyle I noticed that the
1:24:51 viewer count has been in the upper 300s
1:24:53 that that pin just went away do you pay
1:24:56 for any AI subscriptions which ones are
1:24:59 worth it for hobbyists like image
1:25:01 generating um so it depends what you
1:25:03 want to do for for me chat gp4 is if I
1:25:08 if I could only have one subscription
1:25:09 that would be the subscription I'd pay
1:25:11 for um if you want to do sort of
1:25:13 higher-end art then mid journey is worth
1:25:17 paying for I think you can do a 10 Buck
1:25:18 a month thing that'll get you a decent
1:25:21 amount of of render
1:25:25 um as exciting as all the video stuff is
1:25:28 the video things are all quite expensive
1:25:30 just in terms of like per per minute of
1:25:33 video produced they're pretty
1:25:36 expensive um hey Jen and Runway ML and
1:25:41 pabs and all that um and then same with
1:25:45 the auto audio synthesis stuff they're
1:25:48 kind of expensive I would say chat GPT
1:25:50 is the big one that I would pay for for
1:25:53 and then if you really want to do you
1:25:54 know specialized higher-end art um
1:25:57 invest in a mid Journey subscription as
1:26:00 well I think those two are the
1:26:02 biggies um we give you a geni response
1:26:05 in Google
1:26:06 searches might be a beta feature that
1:26:09 you enable yeah P but one of the
1:26:12 problems is this would actually be an
1:26:14 interesting Insight from you I I
1:26:17 understand why it's there but it's it's
1:26:19 frustrating if I pay Google
1:26:23 if I pay Google I don't get the advanced
1:26:27 features the advanced features are not
1:26:29 available to work work group or whatever
1:26:32 the hell the Google product is for
1:26:34 companies that pay for Google all that
1:26:37 advanced [ __ ] we can't use it's really
1:26:39 annoying um Kyle I noticed that the
1:26:41 viewer count has been in the upper 300s
1:26:43 during this live during this day live
1:26:46 and looks like your followers count is
1:26:49 up awesome thank you very much Tobias
1:26:51 for for checking in on that yeah this
1:26:53 was amazing this this uh live where we
1:26:55 just we did the the zoom call um started
1:26:59 at 75 people it kind of peaked at about
1:27:01 150 and came back down we're we're right
1:27:03 back where we started at 80 right now so
1:27:06 it's amazing
1:27:08 um it will be smart enough to sift
1:27:11 through them and Spotlight which ones
1:27:12 are most interesting to you does Kyle
1:27:14 look tired to anyone yes I've been I've
1:27:18 been so last night I went live at 8 uh
1:27:21 I'm also shooting up it's a it's an up
1:27:23 angle so my chin looks particularly
1:27:25 large which chin the the third one looks
1:27:28 big
1:27:30 um um I went live at 800 last night till
1:27:34 10:30 took a nap went live at midnight
1:27:37 till 3:30 in the morning yeah there
1:27:39 there were more than 400 this morning I
1:27:41 know it was crazy um then I got up at 6
1:27:45 I did a a live from 6:00 to 810 or
1:27:48 something like that then I just did this
1:27:50 one now I got to go do some work for
1:27:52 store Story vine then I'm going to come
1:27:53 back in the afternoon then I'm going to
1:27:55 go home and have dinner then I'm going
1:27:56 to come back so I'm going to keep doing
1:27:59 you will have to deal with me looking
1:28:00 tired but that's okay how and why
1:28:03 searches in Google went up 38% that's
1:28:08 amazing amazing uh mid Journey feels
1:28:11 hard to use so there's a technical
1:28:14 reason for that it's because it's hard
1:28:16 to
1:28:19 use
1:28:21 um if if you want ease of use um Dolly 3
1:28:27 is the easiest if you have a
1:28:31 subscription to gp4 it's built into it
1:28:34 you can just say make me an image of
1:28:36 this or you can be in the context of a
1:28:38 conversation talking about a marketing
1:28:40 strategy and say hey I'd like some
1:28:42 images for that marketing strategy and
1:28:44 it will generate them for you so the
1:28:46 nice thing about do3 is it does the
1:28:49 prompting for you you kind of make a
1:28:51 request it does the actual ual prompt
1:28:53 behind the scenes so you can do that
1:28:55 either at um chat.
1:28:58 open.com the paid version
1:29:02 or co-pilot
1:29:05 microsoft.com is Microsoft's version of
1:29:08 chat GPT and it's got Dolly 3 built into
1:29:11 it so you can use that for
1:29:13 free
1:29:17 um yep I already I already answered that
1:29:21 all right fantastic I've been using
1:29:22 Dolly and it's great what about
1:29:24 Photoshop so photoshop's good
1:29:26 photoshop's got a thing called
1:29:27 generative fill within it here's how
1:29:29 I've been using
1:29:30 Photoshop the Adobe
1:29:34 Firefly base generative stuff is not
1:29:39 really up to par with the other things
1:29:40 although they did just come out with
1:29:42 firefly 2 which is
1:29:45 better um but what I've been using
1:29:48 Photoshop for is I'll generate something
1:29:51 say in do and then maybe do adds some
1:29:55 words in the upper right hand corner and
1:29:57 I don't want those words there or I I
1:29:59 did it in the wrong aspect ratio but I
1:30:01 love the image I'll then go into
1:30:03 Photoshop and do generative fill or I'll
1:30:06 select certain areas and and and use
1:30:09 generative fill to fill in certain areas
1:30:11 of an
1:30:12 image um I wonder if I've
1:30:17 got no I don't have it up um so anyway
1:30:20 all this generate I didn't that that
1:30:23 went away too quick before I could read
1:30:24 it clap is impressive where it takes a
1:30:27 30 minute video and creates shorts for
1:30:29 posting yeah there's a bunch of those
1:30:31 Kevin there's one that we're using
1:30:34 called VI does does that there there's a
1:30:37 there's a pile of them out there that
1:30:39 will take long form videos and turn them
1:30:41 into shorts all this generated all all
1:30:44 this generates so much anxiety I feel
1:30:46 lost already okay Nicholas super
1:30:50 important um
1:30:54 that's exactly the appropriate feeling I
1:30:58 have been wait you aren't home right now
1:31:01 no I'm at I'm I'm at the office
1:31:04 um that feeling of
1:31:07 overwhelm is is not something that goes
1:31:10 away I hate to break it to you
1:31:14 but if the Panic that you're feeling is
1:31:17 oh [ __ ] I've missed it or oh [ __ ] I've
1:31:19 behind or oh [ __ ] I'm not smart enough
1:31:21 just let all that [ __ ] go just start
1:31:24 playing with it if you're on this
1:31:26 channel right now if you're like wanting
1:31:29 to be
1:31:30 curious you're way ahead of the game way
1:31:33 ahead of the game we are very early as
1:31:37 as sophisticated as these tools are
1:31:40 they're they're babies this is the
1:31:42 oneyear anniversary of this tool that
1:31:44 kind of changes the game one year in um
1:31:48 so yes there's a lot to catch up with
1:31:50 but you don't have to catch up with all
1:31:52 of it if if there's one thing I would
1:31:54 say to do is chat
1:31:56 GPT it is by far it it's farther ahead
1:32:00 in terms of its capabilities than any
1:32:03 other tool right now and so rather than
1:32:05 feeling like you got to learn all that
1:32:07 other [ __ ] just pick a single tool and
1:32:10 just play with it don't feel like you
1:32:12 have to you
1:32:15 know do anything other than have fun
1:32:18 because in having fun you'll discover
1:32:21 something where you go oh I know what I
1:32:24 could do with this and then you'll go do
1:32:25 that and then you'll solve a problem
1:32:27 over there you're like okay that wasn't
1:32:29 so bad and then just go back to having
1:32:31 fun all right this is fun I like making
1:32:33 pictures what if I made a card game oh
1:32:35 it it gave me the rules of the card game
1:32:38 well wait what if I turned that into an
1:32:39 action adventure text based game oh it
1:32:41 could do that too wait could it
1:32:42 visualize the images oh there's the IM
1:32:45 just
1:32:46 play and then someone will come to you
1:32:48 and they'll be like you know what I need
1:32:49 to do I need to figure out how to make a
1:32:51 action Adventure game for my pizza
1:32:53 parlor how do I don't how about I even
1:32:55 begin to do I can do that for you I just
1:32:57 did
1:32:59 that charge him 500 bucks and make him a
1:33:03 game is there a mentor list that someone
1:33:06 can get on yes it's called the AI Salon
1:33:09 hang
1:33:19 [Music]
1:33:21 on
1:33:22 hang
1:33:23 on my presentation's up in the cloud I
1:33:26 got a I got a downloader we got a
1:33:29 downloader from the cloud if you know
1:33:30 what I mean you put your you put your
1:33:32 files up in the cloud there and then uh
1:33:35 and then you need internet access to go
1:33:36 to the cloud and get them it's it's a
1:33:38 cloud it's it's a technical thing I
1:33:40 don't want to be too intimidating but
1:33:42 I'm you know I'm technically gifted I
1:33:43 know terms like
1:33:45 Cloud all
1:33:48 [Music]
1:33:51 right
1:33:52 okay
1:33:54 this this is what we
1:33:58 want in this community the AI salon so
1:34:01 this is a group that I
1:34:03 started
1:34:07 um wait I've noticed that in the last
1:34:10 month everyone is starting to catch up
1:34:12 they are
1:34:14 except they're catching up to like chat
1:34:16 GPT 3.5 no one's anywhere close to GPT 4
1:34:21 um the AI Salon if you go to AIS salon.
1:34:25 mn. this is a community I started a year
1:34:28 ago next Tuesday is our Oney year
1:34:30 anniversary we started this group the
1:34:32 week after chat GPT launched because
1:34:35 when chat GPT launched I lost my [ __ ]
1:34:39 mind and uh and this group is is about
1:34:44 people that are opting in learning about
1:34:46 this stuff so so what this ends up being
1:34:49 is kind of like a support group for for
1:34:51 for people that are trying to learn this
1:34:53 AI stuff so if you're feeling a little
1:34:55 drifting in the wind and you want to
1:34:56 hang out with people that are up to
1:34:59 things that's what this group is we meet
1:35:01 every other Tuesday we meet in person
1:35:03 here in Denver and we also meet remotely
1:35:06 this coming Tuesday is our Oney year
1:35:08 anniversary December 5th um the
1:35:10 information of that is is there there's
1:35:13 a show int tell Channel there's guilds
1:35:15 so you can actually join a guild whether
1:35:17 it's writing or art or AI 101 if you're
1:35:20 a beginner um Peter Kaminsky runs the a
1:35:24 AI 101 Guild he's super he's amazing guy
1:35:27 really patient really softspoken he'll
1:35:29 walk you through stuff he makes it very
1:35:31 not scary so anyway yeah that's your uh
1:35:36 that's your Mentor group as far as I'm
1:35:40 concerned all right all right people
1:35:43 listen I got to I got to go do some work
1:35:45 so support group and secret cyborg
1:35:47 confessionals
1:35:49 exactly the uh AI confessionals are the
1:35:51 best best Sherry D thank you so much I
1:35:53 appreciate that um thanks for hanging
1:35:56 out with me I'm going to go do some some
1:35:58 of my day job work like go build some of
1:36:01 this [ __ ] and then um um I'll be back
1:36:07 let me look at my calendar I'll be
1:36:11 [Music]
1:36:14 back I will be back probably around
1:36:16 12:30 Mountain maybe 1:00 Mountain so
1:36:19 sometime in that Arena Be not Afraid
1:36:22 thank you so much for those gifts that
1:36:24 means a lot I appreciate it um all right
1:36:27 peace out everybody hope you enjoyed
1:36:29 that thanks for hanging out y'all are
1:36:31 awesome I will see you later tell your
1:36:34 friends tell the people in your life
1:36:36 that you think this will bug that that
1:36:39 you're
1:36:39 watching an AI learning Lab live it'll
1:36:43 piss them off send them all right see
1:36:46 you later