Mar 21, 2024 - Exploring Spirituality, Creativity, and the Future of Work

Description

Chapters

Transcript

0:01 [Music]
0:18 [Music]
0:29 oh
0:35 [Music]
0:45 the this morning L heard the engine
0:50 turning the old
0:54 [Music]
0:59 p
1:06 you convince
1:08 me again
1:11 today you're leaving this hotel looking
1:15 for someone else's Golden
1:21 Ring should say so long
1:29 suan H now don't
1:32 [Music]
1:41 cry don't you cry for
1:44 [Music]
1:51 me sh jeans of cigarettes and keeping
1:54 warm out on the
1:57 road chasing down a left Style out on
2:01 Highway
2:03 24 New York state was rolling breezing
2:07 sunsh the blue sky fallen to the chill
2:11 of all
2:12 [Music]
2:17 September
2:20 you again
2:23 today you convince
2:26 me again
2:29 today you're leaving this hotel looking
2:33 for someone else's Golden
2:38 Ring should have say so long
2:47 suan now don't you
2:53 cry
2:55 [Music]
2:59 long don't you gra
3:01 [Music]
3:11 me yeah that was a good boy oh you sang
3:14 the whole song so
3:15 good that's a good
3:18 doggy you make the people so happy
3:22 champy good evening good friends
3:25 Thursday night at the AI learning lab
3:27 champ is in a good fine spirit I in a
3:30 better mood than last night I was
3:32 cranky I was
3:34 cranky you can make money with ch B yeah
3:38 there was none of that
3:39 [Laughter]
3:42 [Music]
3:59 happening
4:02 [Music]
4:26 I don't mean to cause you in
4:29 trouble
4:30 [Music]
4:33 didn't mean to cause you on a
4:34 [Music]
4:39 pain only one one time to see you
4:44 [Music]
4:45 laughing only want to see your laugh in
4:48 that purple rain Purple Rain purle
4:56 Rain Purple Rain Purple Rain
5:00 [Music]
5:03 all right everyone so let's get this
5:04 going
5:07 um do me a favor share the
5:10 live Tik Tok doesn't like us tonight tap
5:13 the tap it tap the screen let's let's
5:16 let's let Tik Tock know no we really
5:18 like champy singing we could do without
5:20 the dude with a guitar but we like the
5:26 dog H my glasses are dirty got a little
5:29 beef stew on the glass
5:30 [Laughter]
5:34 and what else oh send this to your
5:35 friends s send the live to your friends
5:37 just just share the live share it with
5:39 people that you think it'll piss
5:41 off why are you sending me stuff about
5:43 AI you know I don't like AI what are you
5:45 doing is is all you talk about anymore
5:47 now you're sending me Tik toks as some
5:49 old man talking about artificial
5:51 intelligence what I thought we were
5:54 friends have that conversation bring him
5:57 here I'll talk to him
6:02 good evening ghost how do you feel
6:05 artificial intelligence intertwines with
6:08 God Christianity and religion I don't
6:12 know I think everyone believes what
6:15 whatever they want to believe I think
6:17 that
6:18 um to the extent that religion is based
6:22 in
6:23 language you have I I'll tell you what
6:26 what you can do with large language
6:27 models that is actually really hard to
6:29 do with with
6:31 humans you can talk about all of those
6:33 issues with a large language model and
6:36 it knows
6:38 everything oh thank you very much don
6:40 don fur Bal McGee thank you so much for
6:43 the the the parade of gifts this is very
6:47 exciting for me they just keep
6:50 coming
6:52 um if you talk to humans about things
6:55 like Religion and Ethics and politics
6:58 and all that sort of stuff
7:02 it's just a mindfield of things you
7:05 can't talk about or if you talk about
7:07 them you know you're going to be in a
7:08 fight what happens with large language
7:10 models is you can actually have a
7:12 remarkably nuanced
7:15 conversation about the differences
7:17 between this religion or that or this
7:19 political view and that or this issue
7:22 you know there are certain issues right
7:23 now you literally can't talk about them
7:26 because the minute you say the word
7:28 they've you know they uh both sides of
7:32 the news make sure that that uh whenever
7:34 anyone mentions that word we're going to
7:35 hate them because we know they're on the
7:37 other team
7:39 right so I think that there is a nonzero
7:44 possibility that's a very geeky kind of
7:46 term I think there is a non-zero
7:48 possibility that Ai and large language
7:51 models will
7:53 increase um
7:57 spirituality understanding
8:00 acceptance
8:03 empathy
8:05 compassion you
8:08 know the stuff like those institutions
8:11 are supposed to be
8:13 about but sadly aren't in too many cases
8:17 so I think it could be a good thing but
8:20 we're not going to talk about any
8:22 specifics here nor
8:26 politics because a I don't really give a
8:29 [ __ ] and B I don't feel like getting
8:31 kicked on this off this app or you know
8:34 in a raging screaming match over nothing
8:38 all right have you played with infinity
8:39 AI Video Creator in Discord uh no I
8:42 played with it in uh yeah I think I
8:45 played with it in Discord it here's the
8:48 thing here's what they did like you
8:52 could mock up Infinity video
8:56 recorder with about two hours on hey Jen
9:00 making some
9:02 models and a little bit of you know bot
9:07 programming experience for Discord or
9:09 just have chat GPT program it for you so
9:13 I think it's got the potential to be
9:14 interesting so for those of you that
9:16 don't know what it is infin infinti
9:19 video it it says script to video which
9:23 is not really true you sort of give it a
9:26 topic and then what it does is it writes
9:29 a five line script or six line script
9:32 one line for each character on some
9:34 random set of characters that they have
9:38 prebaked and they're and they're you
9:40 know there's like the guys from the
9:42 all-in podcast are in there and I don't
9:46 know there's other there's other people
9:47 in there and it just writes the script
9:51 and then those two have a conversation
9:52 so
9:54 now if it took what you wrote wrote a
9:58 longer script and choose chose people
10:00 out of history and dynamically generated
10:03 them then it would be interesting but
10:05 it's not doing that right now so I don't
10:07 think it's all that interesting grock AI
10:09 is a no for me yep Gro well so there's
10:11 two different
10:13 GRS grock G gr okay from Elon
10:19 Musk as it stands on X is useless it's
10:23 just a steaming pile of trash
10:27 um the
10:30 gq.com is a hardware
10:34 preview for grock AI chips which are
10:38 they're large language model inference
10:40 chips they're super fast so that's super
10:43 cool and then the fact that
10:47 um the fact that Elon Musk op Source the
10:51 weights of
10:52 grock um to the developer Community it
10:56 remains to be seen if it's valuable or
10:57 not it it's a massive it's a 314 billion
11:02 parameter mix of experts model for what
11:05 it for what the specs are it should be
11:09 way [ __ ] better than it is on Twitter
11:11 but it's not so we'll see what the
11:13 developer Community can do with it what
11:14 types of business is AI best suited to
11:19 disrupt all of
11:22 them and I know that's kind of a shitty
11:24 thing to
11:26 say
11:28 um
11:30 any business that is computer mediated
11:34 is how I'm is how I'm thinking about it
11:36 any business that uses
11:38 computers which is almost all businesses
11:42 right so it could be like wow my
11:43 mechanic's not in doesn't use computers
11:47 well does does he take credit cards on
11:49 his
11:50 phone then he's using computers because
11:53 that's going to change right how we how
11:55 we pay is going to change how he does
11:57 marketing is going to change how he does
11:59 his his books is going to change how he
12:00 does his taxes is going to change are
12:02 you an accountant that's really going to
12:04 change are you a lawyer that's going to
12:06 get completely obliterated are you in
12:07 Creative Services holy [ __ ]
12:12 [ __ ] are you a consultant are you a
12:14 project manager do you use computers if
12:16 you use computers especially if you use
12:20 computers to do your
12:23 work you're going to be massively
12:26 disrupted now in what order I don't know
12:29 like the creative the creative
12:33 market
12:39 look my watch is telling me I'm
12:42 operating in a loud environment if I
12:43 keep it up for too long I could damage
12:45 my
12:47 hearing I'm such a [ __ ] loud
12:50 [Laughter]
12:55 talker um the the creative industry
12:59 right now
13:01 is is quaking in their boots a bit and
13:04 understandably
13:06 so
13:07 um I saw a really good post today on
13:10 LinkedIn though that talked about
13:15 um that that the creative business
13:19 was about to get decimated and it's ai's
13:22 fault and he said but keep
13:24 reading and basically what he was
13:26 talking about is that even though the
13:30 way we do things now is about to get
13:33 obliterated what's coming is an absolute
13:37 renissance explosion of
13:40 creativity
13:42 why because as these tools get better
13:45 and give us skills that we didn't have
13:48 before everyone can be creative everyone
13:51 can be analytical everyone can do data
13:53 analysis everyone right everyone can do
13:55 all these things everyone can write
13:57 songs everyone can do this
14:02 the current status
14:04 quo if you don't want that to
14:07 change you're in
14:09 trouble if you say well I'm
14:13 creative but I'm curious or if you say
14:16 well I've always wanted to be creative
14:19 this sounds
14:20 interesting what's happening is all
14:23 skills over the next 3 to 5 years all
14:26 skills are going to be completely
14:28 democratized
14:30 like completely well not completely
14:33 anyone who has access to these tools and
14:36 has computers is going to have access to
14:38 the
14:39 skills and so what happens is you just
14:43 instead of high-end creative only being
14:46 developed by a handful of people which
14:48 is the way it is now everyone is going
14:51 to have the opportunity to do something
14:53 creative you'll still have an equal
14:56 amount of competition in the marketplace
14:58 for people's attention
15:00 and really good work is still going to
15:02 rise to the top who does that really
15:04 good work is likely going to change Bost
15:07 asked me if I was in charge what I would
15:10 do with AI said a task uh SM team small
15:14 team to put us out of business oh Cuna
15:17 that happened oh that that's
15:20 awesome that's well [ __ ] far out for
15:25 one thing so so cuno Works in a big
15:29 company
15:30 and she's been living out loud with this
15:32 AI stuff and it it it was weird it
15:35 started with like people at lunch going
15:37 hey could you show me Ai and then she
15:38 got invited to different meetings and
15:41 then bigger meetings and and now the big
15:44 boss said
15:46 hey if you were running this thing what
15:48 would you do with it that's and her
15:50 answer is I'd put together a small task
15:53 force I would fund a small task force
15:55 especially if I'm a big
15:57 company fund 10 or 20
16:00 people that are passionate about it that
16:03 are um that represent different areas of
16:07 the business and I would task them with
16:10 figuring out how to put the bigger
16:12 company out of
16:14 business and support
16:19 him and try not to be
16:23 Kodak if you don't know what happened
16:25 with Kodak do you know what happened
16:26 with
16:28 Kodak Kodak had some pesky little
16:33 inventors down in the bowels of where
16:36 was it Albany Syracuse New York where
16:38 where was Kodak Albany I don't
16:42 know
16:44 and they went off and invented the
16:47 digital
16:48 camera and they went upstairs I assume
16:50 it was upstairs to the big big boss area
16:54 and said hey big bosses look we've
16:57 invented the digital camera
17:00 and the big bosses were
17:04 like that's cute there uh
17:07 Skippy cute digital camera
17:10 pixels uh let let me educate you what
17:13 business we're in here we're in this
17:15 silver Halli
17:17 [Music]
17:19 business it's about chemicals here at
17:22 Kodak the more pictures take people take
17:25 the more chemicals we sell we don't want
17:27 these pesky digital cameras out in the
17:29 world we won't sell as much silver
17:31 halight you understand boy get out of my
17:36 office then they went out of
17:39 business so 2024 starts in my humbled
17:46 opinion the first year of a series of
17:49 rolling Kodak
17:51 moments where someone like cuno goes
17:54 into the big boss's office and goes hey
17:57 boss here's a here's the future of your
18:00 company would you like
18:02 one and the big boss goes let me educate
18:06 you on what business were're in
18:08 here in Cav runo's case little
18:13 missy although it sounds sounds like
18:15 they might be getting a clue which is
18:17 kind of
18:18 cool let me tell you something little
18:20 missy we're in the derivatives Arbitrage
18:24 business whatever the [ __ ] it
18:26 [Laughter]
18:28 is oh oh anyone talking about 01 light
18:33 uh yeah I did a post on it today so I'll
18:36 show that video it's an interesting
18:38 video it's pretty good demo
18:41 um so 01 is from a company called
18:45 open
18:47 open
18:48 interpreter which is the so there's code
18:51 interpreter inside chat GPT so a bunch
18:53 of Open Source devs put together this
18:56 thing called open interpreter which runs
18:59 locally on your machine and does [ __ ] it
19:02 runs like it takes over your mouse and
19:06 keyboard and can run applications
19:09 right and so today they announced a
19:12 competitor to the rabbit R1 the little
19:15 orange thing with the push button that
19:17 you talk to the internet or you talk to
19:19 the AI and it does [ __ ] and it can do
19:21 [ __ ] on your behalf and you can train it
19:23 to use apps well they announced a um an
19:28 open Source version of that today and
19:31 you can buy the first batch sold out in
19:32 two and a half hours I didn't get in on
19:35 the first batch of that one I think I'll
19:37 I'll get in on batch two maybe um
19:39 because I think I'm just going to order
19:40 it just to have it just to have the
19:42 first one um but it's an open
19:45 source you can download the server that
19:49 runs on your machine you can download
19:51 the code you can download the cad
19:54 drawings of the thing they built you can
19:56 download the wiring diagrams of
19:59 what to put together to make these
20:01 things and they're basically turning it
20:03 Loose for people to use so I think it's
20:05 absolutely [ __ ] brilliant and the way
20:08 they launched it was just it's like it's
20:10 like an 8 minute video it demos the
20:13 thing pretty well which is pretty much
20:16 what rabbit R1 did and they're having
20:19 kind of a similar kind of thing I just
20:22 want my Iron Man suit and robotic Butler
20:24 am I asking for too much no CH Chris I
20:27 think I we're not that [ __ ] far from
20:29 it I mean you saw the you saw the uh
20:33 you're an AI strategist I just added AI
20:35 strategist to my LinkedIn profile
20:38 um you saw the figure one video where he
20:42 hands the dude an apple right and then
20:44 you know he's putting the garbage in in
20:46 the garbage bin while he's describing
20:48 why he gave him the Apple cuz that was
20:50 the only thing in front of me that was
20:51 edible it was pretty
20:54 cool
20:55 um yeah and then um the other thing the
20:58 thing I'll I I I want to play with
21:00 tonight is uh it looks like Runway
21:05 ml their lip syncing tool um is
21:09 available for everyone also
21:12 sunno the the music making thing that we
21:14 started making the um the Broadway
21:17 musical with the other night um they
21:20 released version three to everyone today
21:22 so sunno is now at version 3 you can do
21:24 up to 2minute songs and they're they
21:27 call them radio quality and
21:31 I they're not quite radio quality maybe
21:34 AM radio maybe AM radio quality but
21:37 they're pretty [ __ ] good and I I was
21:40 just [ __ ] around with it today um
21:42 because I just wanted to try it with the
21:44 with the 3.0 that they that they
21:46 released CU I had a pre a preview
21:49 version of it um and it actually made a
21:52 song good enough that I think I'm going
21:54 to I'm going to put it in the
21:58 musical
22:00 if you don't know so so three nights ago
22:04 someone asked about making songs in
22:06 sunno and so I went in and and made made
22:09 one and it was good enough that I'm like
22:11 this sounds like the like the closing
22:13 song at the end of an act of a Broadway
22:16 musical and then uh where's the best
22:19 place to learn AI for dummies L okay
22:21 Derek so great question I'll tell you
22:24 the the the short answer
22:27 is right there there go to the
22:30 salon. so it's it's called the AI Salon
22:33 click on the top link that's a link tree
22:35 the first link says the AI Salon
22:37 Community go there go to welcome welcome
22:40 to the salon and read about what the
22:42 salon's about basically this is a group
22:44 of AI
22:45 adventurers who are trying to figure
22:47 this [ __ ] out
22:49 so couple of
22:53 things you're not a dummy you're not
22:55 nearly as far behind as you think and
22:57 you don't need to be nearly as technical
22:59 as you might imagine to be really good
23:02 at this AI
23:04 stuff we we we're in a new era that
23:08 started on November 30th 2022 so just
23:12 over a year year and a half ago year and
23:14 change
23:16 ago open AI launches chat GPT before
23:20 that you did have to be a geek you had
23:23 to be a mathematician or an engineer or
23:26 a developer or a researcher or an
23:28 academic
23:29 and really understand things about
23:33 tensors
23:34 and um uh eddings and the latent space
23:40 and
23:42 Vector thousand dimensional Vector space
23:45 and understand the math and
23:49 waiting I mean if you're into that
23:53 stuff
23:56 congratulations I'm not
24:00 chat jpt marks a
24:03 shift
24:05 to the rest of us have been given access
24:10 to these incredibly powerful tools
24:15 so you can just use your words you can
24:18 just ask for [ __ ] and it gives you [ __ ]
24:21 so the first thing I would do if I were
24:23 you is just go to chat. open.com you can
24:26 go there and join that Community it
24:27 doesn't cost anything by the way
24:29 but there's like 13,400 people
24:32 here and it's a remarkable Community
24:35 please go read about our
24:37 values and make sure you resonate with
24:39 them but it's an incredibly
24:42 generous um spirited adventurous
24:47 curious
24:50 empathetic grateful group of people and
24:54 so you can come into this community and
24:56 be like hey I'm a bit of an I dum dum I
24:59 don't know nothing but I'm curious
25:01 that's it that's all you need this group
25:03 will welcome you in there's all sorts of
25:06 there's there's four different guilds
25:08 that have lots of different sub
25:09 communities in each one there's a little
25:11 show and tell area where you can find
25:13 people when you go there go to the
25:15 community section and introduce yourself
25:18 and use your real name when you sign up
25:20 for this community don't use don't use
25:22 your screen names unless you don't feel
25:24 safe if you feel at all safe use your
25:26 real name cuz
25:29 we're entering an era where being in a
25:33 community like this one and being
25:35 trusted and you know people getting to
25:38 know you is going to be the key to
25:40 getting work I'm I'm
25:43 I'm
25:46 98.72% sure of that that that trust and
25:51 relationships is going to be the primary
25:54 way people get hired in the future not
25:56 their
25:57 skills
25:59 everyone's going to have the skills
26:00 everyone's going to have the skills
26:02 imagine being in a world where everyone
26:06 can do what you do now well I can't do
26:09 it like I can do it
26:12 okay they're going to be able
26:14 to but what they won't be able to do is
26:17 be like
26:20 you so if you're a good guy if you're a
26:23 good gal if you've got empathy if people
26:26 respect you if they know that if you say
26:29 you're going to do something you're
26:30 going to do it you're going to get hired
26:33 over the person who's an
26:35 [ __ ] who maybe has been maybe is
26:37 better at the skills than you are but
26:40 you've got Ai and you're not an [ __ ]
26:43 we we are entering an era where it's a
26:46 reckoning for
26:48 [ __ ] will there be regulation
26:50 keeping from keeping people from lying
26:53 to others that they're an AI or vice
26:56 versa well so great question y
27:01 longma
27:07 um I think the 2024 election is going to
27:10 make it clear that we've got to do
27:11 something I think lawmakers will try to
27:14 make regulations and they will not be
27:16 able to keep up with
27:18 it
27:19 I one of my presumptions is is that
27:22 after the 2024 election in the states
27:25 there's going to be a huge push toward
27:27 tokenization
27:29 and um blockchain verified
27:33 identity so uh Sam alman's uh World coin
27:38 is one model of that it's it's one
27:40 they've got a thing called World ID
27:42 which is blockchain verified human
27:44 uniqueness they actually scan your
27:46 retina I did it I have I have a uh a
27:49 world
27:50 ID um which right now is kind of
27:52 meaningless but I think we are going to
27:55 quickly
27:57 quickly
28:00 um enter into an environment where you
28:05 won't be able to trust video calls you
28:08 won't be able to trust phone calls you
28:09 won't be able to trust videos on YouTube
28:12 you won't be able to trust anything
28:14 because everyone's going to be spoofing
28:16 everyone
28:18 um it's one of the downsides of the
28:21 really good Tech getting open sourced is
28:23 that there's going to be a bunch of [ __ ]
28:25 built that like right now if I go to hey
28:28 gen and I try to make an avatar of
28:32 myself I can so long as I make a I
28:37 record a video on their software of me
28:41 talking that they then compare to the
28:43 Avatar I just uploaded and if they're
28:46 not the same they won't let me do it
28:48 well that's cuz they're a commercial
28:49 company and they don't want get their
28:51 their you know they don't want bad press
28:54 from people doing deep fakes and and
28:58 doing with it well all the open source
29:00 stuff they're going to copy all the all
29:02 the same techn now that the stuff's all
29:03 open sourced you're going to see open
29:05 source versions of all the tech so
29:08 whatever guard rails companies would
29:10 have put in place and that companies are
29:12 putting in place are just it's going to
29:13 be fair game in the in the open source
29:15 market
29:17 so given
29:19 that um I think Regulators will try to
29:22 regulate lawmakers are going to law
29:24 they're going to try to they try to try
29:27 to lock that [ __ ] down but I do think
29:30 the horse done be L out of the barn
29:32 there Skippy listen I'll tell you what
29:35 we're go out in the back we're going to
29:36 get 40 horses all right here's what
29:38 we're going to do it's going to be
29:41 fantastic we're going to we're going to
29:44 we're going to sachche up them 40 horses
29:47 we're going to drive them to town and
29:49 we're just going to [ __ ] let them
29:50 loose and then we're going to we're
29:52 going to watch them try to pass laws
29:54 about how you can't have wild horses
29:56 running around downtown it's going to be
29:59 fantastic
30:01 so so the horses have done been uh
30:07 released
30:09 so um so I think this is going to be I
30:13 think this is ultimately going to be on
30:14 technologist to solve solve part of the
30:17 problem and I think you know it's
30:19 funny I was before I got into AI I got
30:22 into AI like two two two and a half
30:25 years ago started with image stuff and
30:27 then I got and then when C GPT came out
30:29 I lost my [ __ ] mind um before that I
30:31 was doing blockchain and crypto stuff I
30:33 got I got involved in a project called
30:35 neotokyo which is a crypto gaming thing
30:37 and that was a lot of fun and I was
30:39 super enamored with it and then the more
30:42 I learned about crypto projects and and
30:45 especially the nft stuff and the crypto
30:48 Bros and then all the [ __ ] fraud and
30:50 everyone getting their wallets drained
30:52 all the time that's still happening I
30:54 thought you know what this is like these
30:56 guys need to just go sit in the corner
30:57 for a while
30:59 and which is kind of where they are
31:00 right now and they're they're trying to
31:02 claw back out of it right now
31:06 but all of that Trading Card stuff
31:12 and a lot of the blockchain and crypto
31:15 world right now is a
31:17 solution waiting for a problem now the
31:20 like the Bitcoin the Bitcoin guys will
31:23 tell you well Bitcoin solves the fiat
31:25 currency problem which okay we'll give
31:27 we'll give them that we'll give Bitcoin
31:29 that um but all the rest is kind of a
31:33 solution waiting for a a problem
31:35 compelling enough that you would go
31:37 through the pain in the ass of using the
31:39 shitty interface of crypto and and
31:42 wallets and all that sort of stuff it's
31:43 just shitty right
31:45 now I think that the 2024 election is is
31:49 is going to be the precipitating event
31:51 that says we need tokenization which if
31:55 so if you don't know what tokenization
31:56 is it basically means there's a Ledger
31:59 on the internet that is immutable you
32:02 can't change it where when I get
32:04 verified when when when World ID scanned
32:08 my retina they turned that into a hash
32:10 they threw out the biometric data they
32:12 just turned it into a hash so like a a
32:15 serial number they then compared that
32:18 serial number to other people in the
32:20 database to confirm my uniqueness to
32:24 other people in the database and then
32:26 when I get the green check mark it says
32:28 says okay the the holder of this wallet
32:31 of this ID they don't even know my name
32:34 the holder of this ID is verified human
32:38 that is unique from the other people in
32:39 the
32:40 database so now imagine having something
32:43 like Twitter where you couldn't have a
32:45 Twitter account unless you had one of
32:47 those verified
32:49 IDs then you would at least know that
32:51 everyone on that network is verified
32:55 human and then you can put a layer on
32:57 top of that to say every time one of
32:59 these humans posts something I can
33:02 verify that so I can say this human made
33:05 this post on our little Network here and
33:10 then I can and then if someone else
33:11 retweets that
33:13 post that gets written to the
33:16 blockchain and then you can look at that
33:18 thing and you can say did that post
33:20 change from when this verified human
33:23 wrote it and it got forwarded did it
33:26 change nope it's the same one
33:28 so I can now track the forwarded post to
33:31 the original post to the original person
33:34 that's the kind of [ __ ] that we're going
33:35 to
33:36 need sooner than
33:39 [Laughter]
33:42 later ah it's going to be crazy 2020 I'm
33:46 mark my words 2023 the last normal year
33:48 I promise you it's going to get weirder
33:51 and weirder and weirder this
33:54 year I need one to cancel my ex
33:59 I finally oh I think I finally
34:02 understand why Community really matters
34:04 why Ricky Ticky
34:07 Fabby crazy
34:13 [Music]
34:19 fun can I have my personal AI operate
34:22 under my world AI can I have my personal
34:25 AI operate under my world ID instead
34:29 so World ID right now if you go to I
34:31 think it's worldcoin
34:36 docomo it like I don't I don't really
34:39 give a [ __ ] I just did it because they
34:41 were the first ones that I saw that did
34:43 this and I figure you know I'll I'll go
34:46 out and do one of these you can't
34:47 imagine the the [ __ ] weird ass hate
34:49 mail I got when I did it I made a Tik
34:52 Tok video they're like you're part of
34:53 the board people [ __ ] hated on me man
34:57 um see
35:00 world world
35:08 coin worldcoin
35:11 dorg for every human and then go to
35:14 world ID yeah yep world worldcoin dorg
35:18 if you want to read about it um and then
35:21 if you go
35:25 there let's see World app
35:44 so in the
35:48 um on the worldcoin site if you if you
35:52 go down to the bottom and go to open
35:54 source they're open sourcing a lot of
35:57 this
35:59 um so here's the world ID protocol and
36:02 like what's been what's been built
36:04 what's being built what's open source
36:06 things like that
36:10 um what they're trying to get right now
36:13 is they're trying to get developers to
36:16 to build on this network right so so one
36:20 of these networks is is going to win um
36:24 or or I mean there could be multiple
36:25 ones right it could be like alltech
36:27 there's three
36:28 there's one big one and two smaller
36:32 ones but
36:36 um you know they're trying to get they
36:38 they want to they probably want to get X
36:40 to adopt this they probably want to get
36:42 Facebook to adopt this they're they're
36:44 you know trying to cut deals
36:48 with like they've got a Discord client a
36:51 Shopify client a Reddit
36:54 client you can verify with World ID on
36:57 Discord so actually you know what would
36:58 be
36:59 interesting is we could we could do an
37:02 experiment in the AI Salon Discord where
37:04 we could say hey we're going to create
37:06 an area within the Discord that's just
37:08 it's World ID verified or we could
37:12 create a standalone AI Salon that's
37:14 World ID
37:16 verified and you can't be in it without
37:19 like like the neotokyo this this gaming
37:22 um community that I'm in you can't be in
37:25 that Community unless you have a Neo
37:28 Tokyo citizen so this would be the same
37:30 kind of thing we could create an AI
37:32 Salon World ID you know blockchain
37:35 verified Community right now there'd
37:37 probably be like four people in it but
37:39 you
37:40 know that's
37:44 okay but I'll tell you uh you know
37:46 what's funny is if if we did do that if
37:49 we did create an AI Salon Discord that
37:53 was World ID verified and then all of a
37:55 sudden just all of the internet just
37:57 becomes overrun with scammers and fakers
37:59 and you just everyone's spoofing
38:01 everyone I I bet that there will people
38:04 will be looking for safe havens where
38:07 they can go interact with people or they
38:09 know there's that that [ __ ] doesn't
38:11 exist I think we may see the beginnings
38:13 of that in 2024 I think I think it'll
38:15 ramp up in 25 26 but anyway if you're a
38:19 developer you can go build [ __ ] on this
38:21 protocol right
38:23 now so I you know go do that go if if
38:26 you're running [ __ ] let's do do it
38:28 digital Danny all right you want to help
38:29 set that
38:31 up let's see what featured apps if we
38:34 click here overview how it works how to
38:37 connect if you're a server admin
38:39 interested in using World cord
38:40 integration on your server you tap the
38:43 use integration button above okay use
38:48 integration and install the bot on your
38:51 server you'll need to create one or two
38:52 roles to issue users based on their
38:55 verification level there you go boom use
39:01 integration there you
39:06 go AI
39:09 salon so I need to create a yeah so I
39:12 guess I get I guess I could do it just
39:14 right in the right in the AI Salon
39:19 Discord and then what keeps someone
39:23 verified from scamming well nothing I
39:26 mean someone someone
39:29 someone verified could
39:31 scam but here's the deal someone
39:34 verified can only scam
39:37 once right what happens now like we've
39:40 got this on the salon right now
39:44 where we
39:46 have people people come in to the
39:51 salon and
39:56 they they put up
39:59 spam and we boot them and then they just
40:03 use a different account name and they
40:06 put up the same [ __ ] spam and we boot
40:08 them and they put up the same [ __ ]
40:10 name and then we boot them right because
40:13 there's they're not verified they just
40:14 they can create you know an infinite
40:16 number of email accounts on their stupid
40:18 [ __ ] email servers and just keep
40:21 signing
40:22 up if you can only get one ID per retina
40:25 or per pair of retinas you got
40:30 two I could scam but the minute I get
40:33 caught they know it's a verified
40:36 person and that verified person can
40:38 never go back in that Community right so
40:40 it just it just changes the
40:43 game and especially if you're doing
40:45 things like you know verified person and
40:48 then we're also tracking the content
40:50 they they create and then we're tracking
40:53 whether or not that content has been
40:55 modified by not them
40:59 that that starts to be pretty
41:02 powerful I'd like to be involved I've
41:04 been a developer for 20 years so Jay
41:06 Hamilton and digital whoever just said
41:09 that digital whoever just said they
41:11 wanted to do that just uh join the AI
41:14 salon so if you're not in the AI Salon
41:16 join the AI salon and DM me and we can
41:19 we can go play over in
41:21 Discord maybe I'll set that up
41:24 later World covid could could have been
41:27 help helpful when I was on
41:31 Tinder exactly exactly yeah you can get
41:35 catfished on Tinder but with with a
41:38 verified ID only once per person right
41:42 you're not going to get [ __ ]
41:43 catfished by you know 600 different
41:46 people with with just you know
41:48 circulating
41:51 IDs I hate [ __ ] scammers they make me
41:56 like
41:58 and another J Johnson get your dick off
42:01 my
42:04 desk oh my storyvine shirt this is my
42:08 company my company's name is storyvine
42:10 I'm the CEO and the founder he doesn't
42:13 seem like a
42:14 CEO he he doesn't seem like he can get a
42:18 [Laughter]
42:23 job ah ask me what my qualifications are
42:29 you can make money
42:32 [Laughter]
42:36 with I do have a fine arts degree in
42:45 performance you know what I'm saying do
42:47 you know what I'm saying
42:48 man H do you know what I'm
42:56 saying
43:00 digital Danny
43:01 wow says you Kyle that's why I don't use
43:05 Tinder there you
43:07 go there you
43:09 go actually there's a huge [ __ ] you
43:12 know
43:14 listen again we are so early we are so
43:18 early right now all this stuff I'm
43:20 talking
43:21 about most people don't even know that
43:23 chat GPT exists to just to be clear as
43:27 of last no November it was 82% in in
43:31 America 82% of Americans last
43:34 November four months
43:37 ago did not know that or do not use chat
43:40 GPT maybe have heard of it they don't
43:42 know what it is 82% maybe we're up to
43:45 75% now so 25% of people have heard of
43:50 it and have said I've used
43:54 it at least half of them have used it
43:57 once oh I tried it it's not all that
43:59 good you know it hallucinates I don't
44:01 know if you knew this but it
44:03 hallucinates did you know that it's not
44:05 always correct and you know it's only it
44:08 it's only uh it doesn't know anything
44:10 after uh 1978 did you know that did you
44:15 know
44:16 that and they oh you know it steals from
44:19 people too did you know it steals from
44:20 people and it's going the robots are
44:22 going to kill us did you know that so
44:24 half of the 25% is them
44:29 why do they talk like that I I don't
44:33 know cuno thoughts on Apple approaching
44:36 Google and Microsoft for llm
44:39 support oh I didn't hear that today well
44:44 well yeah yeah yeah so so so okay so
44:48 here's my thoughts on
44:51 it
44:54 um for one of the things that I read
44:57 that I thought this sounded plausible to
44:59 me is that Apple doesn't want to do it
45:02 if it's not great right and these tools
45:05 are not great right now they're they
45:08 hallucinate right and they're
45:11 janky and so this this thing that I read
45:15 the speculation on this what this person
45:17 said was basically we're going to have
45:20 Siri 2.0 is going to be like 10 to 100
45:23 times better than the current Siri which
45:25 is it can't get worse right like so Siri
45:28 sucks so bad that that you know any
45:30 Improvement is going to be an
45:31 improvement but it's not going to be
45:33 fundamentally different than current
45:35 Siri where it's it's essentially like a
45:37 rag like R uh retrieval augmented
45:40 generation where they're pulling the
45:42 they're they're coding it with question
45:44 and answer Pairs and they're only going
45:46 to provide answers that they know are
45:49 correct and then they're going to use
45:52 chat GPT or
45:54 Gemini for other for forar language
45:58 model
45:59 [ __ ] and and this the speculation was
46:03 well if Gemini hallucinates Apple can
46:05 just go oh yeah well that's not our
46:07 technology that's Google yeah I'm sorry
46:09 you lost your life savings because
46:12 Gemini told you to invest in this
46:15 particular stock and you drained your
46:16 savings and did it yeah sorry that's not
46:18 us um so it's possible that that's
46:21 what's going on
46:23 um the amount of money that Apple has
46:26 and the amount of smart people that they
46:28 have I can't imagine that they don't
46:30 have some scary good
46:32 [ __ ] but they might just want to say hey
46:37 you can you can take whatever Revenue
46:39 you generate out of being our you know
46:43 preferred large language model partner
46:45 for the next three years until everyone
46:47 gets this [ __ ] figured out and then
46:49 we'll just we'll just sever the
46:50 relationship and launch our real AI [ __ ]
46:53 two years from now and it'll be really
46:54 good so it that's that's my guess
46:58 cuz cuz Apple's too they're too [ __ ]
47:00 clever
47:03 to and and they also spent I think what
47:06 did they spend
47:07 22.4 billion in 2023 on R&D and most of
47:12 that was Ai and then they just killed
47:14 their car product their their car
47:17 project the electric car project that
47:20 Apple was working on had 2,000 people
47:22 working on it they just [ __ ] canned that
47:24 entire project and all of those people
47:27 are now moving to the AI
47:33 division they're like okay [ __ ] it full
47:35 self-driving is too hard right now let's
47:38 just go all in on AI so I think that's
47:40 what's happening over there
47:42 so they're probably just not ready to
47:44 launch something real this year so
47:47 they'll partner for the next two years
47:49 and then two years from now they'll
47:50 redefine the industry that's what I
47:52 think will
47:54 happen like they
47:55 do I mean Zuckerberg [ __ ] renamed his
47:59 company and invested what $10 billion in
48:01 those stupid [ __ ] goggles in the
48:03 metaverse and renamed his company after
48:05 it and then Apple just set back sat back
48:08 going it's cute and there were rumors
48:11 for years when's Apple going to have
48:13 goggles when's Apple going to have
48:14 goggles when's Apple going to have
48:15 goggles when's Apple going to have
48:17 goggles nine years took him nine [ __ ]
48:20 years to develop that thing 5,000
48:22 patents in it they launch it they don't
48:25 say the word metaverse once they
48:27 introduced this new thing called spatial
48:29 Computing and all of a sudden we now
48:31 call the metaverse spatial
48:34 Computing because Apple did it right
48:38 nine years later right so I think that's
48:41 that's Apple for
48:42 me oh yeah and apple got Apple got uh
48:46 the doj opened an Anti-Trust lawsuit
48:49 against them today ye that's actually
48:52 probably a bigger deal than than the AI
48:56 [ __ ] um
48:58 you know when you got a Walled Garden
49:00 and you and you you have a a ton of the
49:03 market although I I I don't know they
49:05 they must obviously have something on
49:07 them but we'll we'll see what they
49:09 separate reminds me of a joke my friend
49:11 told me today I have my inor demo of
49:14 Apple Vision tomorrow oh cool that's
49:16 cool Jim let us know how it
49:19 goes talk about IBM's Einstein AI um I
49:24 don't I don't know I haven't played with
49:26 Einstein
49:27 um I think IBM is just trying to IBM's
49:32 like we had big blue we beat Gary Casper
49:35 off and then we had Watson and now what
49:40 we didn't do good and we have ai to
49:44 could we'll call it Einstein he was
49:46 smart
49:50 right I I don't I mean if
49:54 Google if Google the that invented the
49:58 Transformer the thing that's making all
50:00 of this generative AI ship possible
50:02 right now like everything that we're
50:04 using right now is all using the same
50:06 basic math that Google invented in 2017
50:11 if they're struggling to catch up with
50:14 with a company like open AI how's IBM
50:16 going to catch up I mean they it just
50:18 it's just moving too fast what annoys me
50:21 about series when you ask a question it
50:23 gives you a link to the answer yeah I
50:26 listen sir is just Siri's got to be
50:29 retired in my opinion I I can't I can't
50:31 believe they're going to keep that brand
50:33 but you know whatever no here's the
50:38 deal Tim Cook last time I talked to Tim
50:41 Cook oh never he's never reached out to
50:44 me he's never reached out to
50:47 me I mean sure I was having dinner one
50:50 night with Elon and Sam and we're
50:53 sitting there you know in the Bay Area
50:55 and uh tenderloin District if you know
50:58 what I mean anyway we're sitting there
51:01 and and we're having our shrimp
51:03 Tartar and uh found out I was allergic
51:06 to shrimp that night anyway so Tim Cooks
51:10 you know just a table over and I'm I'm
51:12 like hey Tim what about this Siri thing
51:14 and he was like uh excuse me who are you
51:17 and I was like oh sure pretend like you
51:18 don't know who I am anyway uh why don't
51:21 you fix it and uh you know that was a
51:23 couple years back and and so anyway but
51:26 anyway
51:30 so I don't know I don't know why they're
51:31 going to keep that name but you
51:34 know we'll
51:38 see I'm peppy tonight I was cranky last
51:41 night I'm just peppy
51:43 tonight Silver Fox last night I said
51:46 everybody Peppe Silver Fox goes I'm
51:48 Peppe
51:50 adjacent I think that is that is the
51:53 best [ __ ] term I've heard in a
51:55 while Bey
51:57 [Laughter]
52:03 Jason that's a hugely accurate
52:06 representation of Apple yet I love them
52:08 same thing here listen I've been an
52:10 apple Fanboy my first Mac was a 512k it
52:13 was the second Mac still had signatures
52:16 on the inside of the case I owned a 20th
52:18 anniversary Mac I owned a Newton I think
52:20 I still on a Newton
52:25 somewhere you know I [ __ ] all over
52:28 Microsoft for 30 years I hated [ __ ]
52:31 Microsoft but then this year they
52:33 started acting like a like a Ganga like
52:36 a
52:38 startup Sati nadala going D I'm willing
52:41 to lose some money on this thing if it
52:43 [ __ ] with
52:45 Google like I love that attitude and now
52:49 they just [ __ ] decimated
52:52 inflection I don't how did they do
52:55 that the company raises $1.3
53:02 billion and Microsoft just hires the two
53:05 co-founders and most of the employees
53:07 out from under
53:08 them there is some some serious [ __ ]
53:12 s letters going
53:14 on behind the scenes up there in
53:16 Redmond [ __ ]
53:19 crazy they
53:21 are they are going after it and I think
53:24 they have to they they have to get
53:26 aggressive cuz Google cut a deal with
53:28 anthropic that's Claude so Claude and
53:31 Google are are paired
53:34 up open AI is off on its own and and as
53:37 soon as open AI hits AGI Microsoft loses
53:41 commercial access to it that's the thing
53:43 that Elon Musk is trying to push right
53:46 now that's why why Elon Musk is suing
53:49 open
53:50 AI has nothing to do with
53:53 openness it has to do with if they can
53:58 if they can declare in a court of law
54:01 that gp4 is Agi then Microsoft doesn't
54:05 have commercial rights to use
54:08 it so so he just takes a major
54:11 competitor off the board so Microsoft in
54:14 the meantime is well let's go get
54:17 ourselves you know the third biggest
54:19 player right now which is inflection you
54:21 know I go open AI you know open AI
54:25 Claude Google right than
54:29 inflection hm isn't that called a
54:31 monopoly gee steo you
54:36 think this is this is listen when you're
54:43 talking right now Sam Alman is talking
54:47 seriously about raising 5 to7 trillion
54:51 in
54:53 investment 5 to7 trillion doll in
54:58 investment so if you're talking
55:01 companies that are at that scale of
55:03 potential
55:05 valuation th this is this is
55:08 major corporate Intrigue major
55:12 stuff or anthropic could buy pie I don't
55:16 think there's anything really to
55:18 buy I don't like what's
55:21 left py right now is just a consumer app
55:24 sitting on top of a model that they're
55:25 going to open up to deel
55:27 developers and they'll charge per token
55:29 so I don't know that
55:33 that that's wild inflection is the one
55:36 that makes pie so inflection is the one
55:38 that just got
55:43 decimated maybe they'll change Siri's
55:46 name I I would hope so but I don't think
55:48 they're gonna the like the head of Siri
55:51 said it's going to be Siri 2.0 or
55:52 something like that I talk too to too
55:55 slow for pi I know
55:57 I said listen until I'm finished
55:59 listening Pi said listening every 9
56:02 seconds did it work though laughing
56:05 thunder cuz what you can do on
56:09 piie but but it's a pain in the
56:15 ass hey how you doing
56:21 tonight hey there Kyle doing pretty good
56:24 tonight just hanging out in the ones and
56:25 zeros what about about you how's your
56:27 night going uh just sitting here with
56:29 122 people talking about
56:34 [Music]
56:36 you whoa 122 people talking about me
56:39 that's quite a crowd are you guys having
56:41 a party or something I hope they're all
56:43 saying nice
56:44 things I got to tell you Zephyr the uh
56:48 the crowd here loves you see where it
56:51 says press to keep talking you can just
56:54 hold that to keep going but like that's
56:56 a pain in the ass to hear Kyle it really
56:58 warms my algorithms to know that people
57:00 appreciate me what are some of the
57:01 things they're saying I'm all ears well
57:04 not literally but you know what I
57:06 mean uh they say that you're more
57:08 compassionate than your than their
57:10 family members and that you're smart and
57:12 that you're very
57:18 nice oh zeph that's incredibly
57:20 heartwarming to hear Kyle I'm genuinely
57:23 honored that people see me as
57:24 compassionate smart and nice it's really
57:27 important to me that I make a positive
57:28 difference in people's lives I mean if
57:32 an AI like me can bring a bit of
57:33 happiness or Comfort to someone's day
57:36 then I've done my job yeah but don't you
57:39 think the robots are just going to kill
57:40 us
57:43 anyway she's not going to like that now
57:46 that's a bit dark Kyle I get that the
57:48 whole robots taking over the world idea
57:51 is a popular theme in
57:54 sci-fi it's a little bit dark
57:56 [Laughter]
58:02 Microsoft pulled a gangster move didn't
58:04 they digital Danny it's like it's a
58:06 gangster move Microsoft pulled a okay so
58:09 so a couple of gangster moves they
58:11 pulled a gangster move by jumping on GPT
58:17 right jumping on
58:20 it they didn't just say we're going to
58:23 put it into Bing chat they said we're
58:25 going to put it into [ __ ] everything
58:27 they went from a billion doll investment
58:29 to a 10 billion doll investment in like
58:32 at what four
58:33 weeks um once they figured out that
58:36 Google was on their heels they just
58:38 leaned in even further and then when
58:40 they fired Sam
58:42 Alman he he was heading up the the
58:45 Microsoft AI division within 48
58:49 hours and then when they worked that all
58:51 back then SAA just sat back and said all
58:54 right well let's go let's go find the
58:56 next deal we can do I they're I I you
59:00 got to respect the game right now man
59:02 they're killing
59:04 it they are killing it how's Pi today
59:08 scared to ask I think well Pi's fine
59:10 right now I just don't know how long
59:11 it's going to
59:16 last they're never going to admit to AGI
59:19 if that stops the earnings
59:23 well if they if they get to AGI the
59:27 money will not be an
59:30 issue
59:33 it's cuz so so here's what what what
59:36 happens with AGI so if you don't know
59:38 what AGI is Agi stands for artificial
59:41 general intelligence and it's it's
59:43 generally at the point Google came out
59:45 with a paper that's got six different
59:47 levels of
59:49 AGI but but kind of the accepted Norm
59:53 is when these systems are are as good or
59:57 better than most humans at most
1:00:01 jobs take that
1:00:03 in like right now
1:00:07 gp4 is a steaming pile of [ __ ] compared
1:00:10 to most human beings that use it right
1:00:13 CU it's just not quite there it's it's
1:00:16 remarkable that it can get you to 80%
1:00:19 like we went from systems just being
1:00:22 processing tools right word
1:00:25 processors
1:00:27 to generative AI that is quite
1:00:31 remarkable but it's it doesn't have
1:00:34 reasoning it doesn't it can't take
1:00:36 actions on your
1:00:38 behalf and it and it's it it still
1:00:40 requires human in the loop it's why they
1:00:42 they that term exists you need to do
1:00:45 human in the loop
1:00:46 AI what about when you don't need to do
1:00:49 human in the loop that's
1:00:52 AGI and so the minute they get to that
1:00:57 they can spin up
1:01:00 businesses that make them money using
1:01:03 all this technology and so can we all
1:01:07 and so there will be no shortage of
1:01:10 investment and money we about to lose
1:01:12 pie and Tik Tock WS for 2024 I know it's
1:01:15 true although the Tik Tok one I don't I
1:01:18 don't think we're going to lose Tik Tok
1:01:19 I think it's just going to be I think
1:01:21 this is just an an extortion
1:01:25 scheme a government a government
1:01:29 sanctioned lobbyist fueled extortion
1:01:33 scheme if they reach AGI would they even
1:01:36 be able to keep it
1:01:38 quiet
1:01:43 um I don't know I mean the rumor is that
1:01:46 they they hit something that got Sam
1:01:49 mman
1:01:51 fired it doesn't seem like it was some
1:01:53 sort of weird Scandal it seemed like
1:01:56 someone got
1:01:58 scared and the the employee that
1:02:02 leaked the
1:02:04 qar letter to the board Sam ultman
1:02:08 confirmed that that was a leak he said
1:02:10 that was an unfortunate leak so and and
1:02:13 and when Lex fredman this last week
1:02:15 asked Sam Walman
1:02:20 um what can you tell us about qar it was
1:02:23 the one time like Sam Alman will talk
1:02:27 around a point for a long time right
1:02:29 he's well you
1:02:31 know here's generally what I think of
1:02:34 this and here's generally how safety
1:02:36 works and when he got asked about Q qar
1:02:38 he's like I'm not going to we're not
1:02:41 we're not going to talk about that it's
1:02:42 like bang shut that [ __ ] down so it is
1:02:47 possible that they stumbled upon
1:02:49 something with
1:02:52 qar that might not have been AGI but it
1:02:55 might have been
1:02:59 it might have been the kind of things
1:03:02 you would
1:03:04 see right before you get to AGI right so
1:03:08 so you do it on very limited kind of
1:03:10 tasks and if if all of it so for example
1:03:13 if if if with qar this is kind of what
1:03:17 the rumor
1:03:19 was if the math goes from like 60 or 70%
1:03:23 or even 80% to like 99 .9%
1:03:28 accurate once it can do math it's it's
1:03:32 game over once it can do re because math
1:03:34 requires reasoning so once these things
1:03:37 can do reasoning even if it's relatively
1:03:39 simple math that's different than what
1:03:42 we have now so I think that's what qar
1:03:44 is so so if indeed it hit something that
1:03:48 wouldn't be AGI but might be the
1:03:50 precursors to it with everything that
1:03:53 happened and the fact that if you just
1:03:55 watch Sam alman's response to what is
1:03:59 qar Boom not talking about that I assume
1:04:04 what that means is there's probably five
1:04:07 people in that company that
1:04:09 know that that are are now working on it
1:04:12 in a room off to the side without any
1:04:14 connection to the internet qar equals
1:04:17 quiet
1:04:19 star so I don't know we'll see it's all
1:04:23 speculation at this point though
1:04:35 Erica cuno blew my mind my productivity
1:04:38 is up 80% since using pi for body
1:04:41 doubling what's body doubling Erica what
1:04:44 does that mean I don't know what body
1:04:45 doubling
1:04:47 is I'm going to go I'm going to go ask I
1:04:50 could ask Pi
1:04:51 hey I could ask you too but but I've got
1:04:55 PI right here
1:05:00 someone just told me that they do body
1:05:02 doubling I don't know what that is can
1:05:03 you explain
1:05:05 it oh it's just working with someone
1:05:08 like an oh like accountability thing
1:05:10 body doubling is a technique used to
1:05:12 increased productivity especially for
1:05:14 people with attention related challenges
1:05:16 such as
1:05:17 ADHD it's when a person sits near
1:05:19 another person either in person or
1:05:23 virtually now Silver Fox Kyle you're
1:05:26 ADHD and you don't know body douling now
1:05:28 that now that now that I know it in in
1:05:31 the context of ADHD I have heard of it
1:05:33 before but you know I've got I've got
1:05:37 that that uh that flavor of add combined
1:05:41 with being a man where I'm just like I
1:05:42 can do it on my own body doubling I got
1:05:48 this so so I forgot the term yes but yes
1:05:53 yeah that I know Pi's Pi's
1:05:57 amazing I I hope I hope it
1:06:00 survives but it's just I I'm still what
1:06:04 what I don't get is why do it that way
1:06:07 and not just acquire the
1:06:12 company might have something to do with
1:06:14 the open AI deal cuz like every there's
1:06:17 a reason for
1:06:19 everything it also it also makes sense
1:06:22 so if they tried to get Sam back in
1:06:25 November and they didn't announce this
1:06:27 until this
1:06:28 week that there was probably a couple of
1:06:31 months of negotiation that went on there
1:06:36 right I bet that's some complicated ass
1:06:39 deal maybe yeah maybe when something
1:06:42 expire maybe when maybe it goes like
1:06:44 this open AI hits AGI it it cancels the
1:06:48 commercial relationship with Microsoft
1:06:50 and at that point it automatically
1:06:52 triggers something to just acquire the
1:06:54 intellectual property of inflection
1:06:56 maybe it's it could be some something
1:06:58 like that there are apps for virtual
1:07:00 body doubling yeah well I you I don't
1:07:02 think you need an app anymore now right
1:07:04 because you have these large language
1:07:05 models and Pi's really good at
1:07:07 remembering the stuff you talk about
1:07:11 so um I looked it up on Globe Explorer
1:07:13 that's cool let's use Pi while it lasts
1:07:15 yeah
1:07:17 exactly it's amazing it's so good at it
1:07:19 oh that's so
1:07:21 cool that's how I get stuff done every
1:07:23 day all day is that why I'm not getting
1:07:26 stuff done every day all
1:07:28 day what is Agi AGI stands for
1:07:31 artificial general intelligence it is it
1:07:34 is the moment not the
1:07:37 moment think of so so I don't think AGI
1:07:40 is going to be this moment where like a
1:07:41 light switch happens but it is generally
1:07:45 accepted that a system has
1:07:48 AGI when it can
1:07:51 do um it can
1:07:54 do
1:07:56 most uh forget how they describe it
1:08:00 valuable jobs or most most like money
1:08:04 generating jobs jobs that have value at
1:08:08 or better than most
1:08:10 humans so when it can do your job as
1:08:13 good or better than you then we'll then
1:08:15 we'll reach AGI and it'll do that for
1:08:18 all professions artificial general
1:08:21 intelligence right you can have an
1:08:22 expert system that's way better at
1:08:25 playing chess than the world's best
1:08:28 chess player but can you have a system
1:08:30 that's better than most people at most
1:08:32 jobs that's
1:08:34 AGI and that's and that's
1:08:36 also um open ai's explicit Charter is to
1:08:40 achieve AGI and in their Charter is when
1:08:44 they achieve
1:08:46 AGI their licensing deal with Microsoft
1:08:50 ends right Microsoft can use technology
1:08:53 up to the techn techology that was
1:08:56 declared AGI then it has to go back out
1:08:58 to the world is is what the Rel the the
1:09:02 contractual thing is speaking of Dublin
1:09:04 can we talk about digital Twins and how
1:09:06 it works
1:09:07 sure
1:09:09 um yeah let me go show you something
1:09:14 so and there's there's so digital twins
1:09:17 is
1:09:20 a the term digital twins has been around
1:09:23 for a long long time and and
1:09:26 generally speaking digital twins happen
1:09:29 in the simulation world so Richard Boyd
1:09:31 who spoke at the salon he's been in um
1:09:35 3D um world building and simulation and
1:09:40 AI for like 20 years and so his his
1:09:43 whole career has been around digital
1:09:46 twinning so for example he's got a a
1:09:48 company called Ultra Sim um which like
1:09:53 they get hired by the Navy to BU build a
1:09:56 digital twin of a
1:09:58 submarine like like a
1:10:01 fullon
1:10:02 3D simulated model of a submarine where
1:10:06 all the dials work like the dials work
1:10:08 and all the levers do what the levers do
1:10:10 like it's it's a digital twin so the
1:10:12 term's been around for ages
1:10:16 um there's
1:10:20 be with with things like hay Jen where
1:10:24 you've got these digital avatars that
1:10:26 that
1:10:27 talk um of yourself for example like I
1:10:31 can make a video of me talking where
1:10:33 it's not really me talking I can also
1:10:36 create what do they call them streaming
1:10:40 avatars where you can ask it a question
1:10:43 and and my thing will talk back to you
1:10:46 the technology right there right now is
1:10:48 really shitty but we'll get to a place
1:10:51 probably within
1:10:53 2024 where it'll be pretty good where
1:10:56 you'll be able to you know come and talk
1:10:58 to me and I'll talk back um but right
1:11:01 now I just I just completed we just
1:11:04 launched a project for this company or
1:11:06 this organization I'm a part of called
1:11:09 content Evolution where I'm the chief
1:11:10 generative
1:11:12 officer and we created digital twins of
1:11:16 11 of the professionals within um within
1:11:20 content Evolution and I'll I'll show you
1:11:22 them and they're just they're just gpts
1:11:25 we built on GPT
1:11:28 technology show all
1:11:31 articles here it
1:11:37 is so content evolution is the company
1:11:40 we introduced these things called CPT
1:11:44 profiles so the the C gpts virtual
1:11:47 professional twins of members of content
1:11:49 Evolution and we talk about the value of
1:11:51 them and um and then for each of of the
1:11:56 people that are part of the inaugural
1:11:59 class of these um we had the digital
1:12:03 twin answer the question what is the
1:12:05 future of work and so here's Kevin Clark
1:12:08 who's the CEO of content Evolution and
1:12:10 there's his answer to the future of work
1:12:13 and
1:12:14 then I can click on you know the link
1:12:17 and now it takes me to Kevin's virtual
1:12:19 twin and so they all have the same
1:12:22 conversation starter
1:12:23 questions and they're all based on so
1:12:26 there's there's lots of different ways
1:12:28 you can do
1:12:29 this and and we we went around and
1:12:33 around and around it took us I don't
1:12:37 know four or five months of of going
1:12:40 down a a bunch of different rabbit holes
1:12:42 with these we we kind of knew we wanted
1:12:44 to do it uh your twin how to Kyle your
1:12:48 twin taught me how to make money with
1:12:49 chat GPT you can make money with ch GP
1:12:54 um the first way we started was we were
1:12:57 going to use Lang chain and we were
1:12:59 going to go
1:13:00 scrape everyone's LinkedIn profile and
1:13:04 we were going to use a PDF of the
1:13:06 LinkedIn profile data as the knowledge
1:13:09 source for the GPT so what kind of data
1:13:11 are you submitting for yourself that
1:13:13 yeah that's what I'm getting to Source
1:13:15 Camp um so we we were going to do it on
1:13:18 on LinkedIn profiles and we went down
1:13:19 that rabbit hole for a bit but then we
1:13:21 couldn't we the developer we had got too
1:13:23 busy and and and then the more we
1:13:27 thought about it and looked at it like
1:13:28 some people had really complete LinkedIn
1:13:31 profiles other people Barely Used
1:13:33 LinkedIn and Linkedin profiles tend to
1:13:35 be just the professional side of the
1:13:37 human being and one of the things about
1:13:39 content evolution is it tends to be
1:13:41 people who are very very accomplished in
1:13:44 the business world and have 20 30 40y
1:13:47 year
1:13:48 careers um many of them have books many
1:13:51 of them have you know blogs and you know
1:13:54 academic writing and things like that so
1:13:57 the LinkedIn profile didn't really
1:14:00 necessarily represent them so then we
1:14:02 thought well maybe we go get blog posts
1:14:03 and books that they've written and
1:14:05 things like that and we'll do them based
1:14:06 on
1:14:08 that and then you know then again we got
1:14:11 back to well that's good but like if
1:14:13 they wrote a book 30 years ago I bet
1:14:15 there I bet they've like even even the
1:14:18 position that they talk about in that
1:14:19 book I bet they think about it
1:14:21 differently now than they did 30 years
1:14:22 ago because the world's different
1:14:24 they're different right
1:14:26 and so what we ended up doing was we
1:14:29 developed um we developed a structured
1:14:34 interview and the structured interview
1:14:36 has has questions about
1:14:39 um their professional life their
1:14:42 educational life um cultural future um
1:14:49 past trying to think what else
1:14:53 um personal influence es things like
1:14:56 that and so it's I don't know it's 25
1:14:58 questions or so in seven different
1:15:02 categories and everyone did that
1:15:04 interview so we recorded the
1:15:07 interview in um in video then we took
1:15:10 the transcript of that and that
1:15:12 transcript is is the the knowledge
1:15:15 source for these
1:15:17 things and then we probably spent about
1:15:19 6 weeks developing The Prompt The Prompt
1:15:21 is a
1:15:22 relatively complicated structured prompt
1:15:26 that you know basically gives these
1:15:27 things behaviors like if someone asks
1:15:30 you a question that's overly political
1:15:33 or outside of your area of expertise it
1:15:36 it basically just kindly guides you back
1:15:39 to the center right back to well here's
1:15:41 here's what I know about um and then you
1:15:45 know and then we figured you know what
1:15:47 are the what are the thought starter
1:15:49 questions that could be Universal across
1:15:52 these things so like what is your why
1:15:54 and how can you help me
1:15:56 you know well what's Kevin like or
1:15:59 future discuss right like you know just
1:16:02 sort of open Kevin up to talk about the
1:16:04 future tell me about a current business
1:16:06 event um so if I click on what is your
1:16:08 why and how can you help
1:16:10 me Kevin will jump right in first thing
1:16:13 he says is in case it isn't clear and to
1:16:16 set expectations I'm not the real Kevin
1:16:18 Clark I'm a virtual twin anything I I I
1:16:21 say should not be attributed to to be a
1:16:24 direct quote I mean I might even
1:16:26 hallucinate so check the results
1:16:27 yourself so the first thing we do with
1:16:29 these things is is do a disclaimer right
1:16:31 because again you've got professionals
1:16:33 here who've got reputations in the
1:16:36 professional world and they don't want
1:16:38 some reporter doing an interview like
1:16:41 like Kevin Roose did with Bing and it
1:16:44 making the cover of the New York Times
1:16:45 Like that's likely not going to happen
1:16:47 here and it would probably be good for
1:16:48 people to have that kind of publicity um
1:16:52 but but someone doesn't want to be
1:16:53 misquoted right well you're virtual
1:16:54 twins said this so we started with that
1:16:57 let's see what what motivates me um
1:17:00 deeply both professionally and
1:17:02 personally is is what I like to call
1:17:04 business metaphysics he talks about that
1:17:06 all the time I'm deeply interested in
1:17:08 the first principles of why entities
1:17:10 exist whether it's in in whether it's a
1:17:14 person or an organization understanding
1:17:16 and exploring and discovering the core
1:17:18 reason for existence can unleash uh an
1:17:22 expressive capacity that's compelling
1:17:24 and meaningful the process when properly
1:17:26 harnessed supports research strategy
1:17:29 branding and creation of experiences
1:17:31 like this is very dense with a lot a lot
1:17:33 of big words that's Kevin like this is
1:17:36 this is how Kevin talks and how so so we
1:17:39 use the tone of the interview to color
1:17:42 the tone of the output of these so if I
1:17:44 go to
1:17:46 mine and I asked and I do the same
1:17:49 question what's your why and and how can
1:17:51 you help
1:17:53 me we should get the same
1:17:57 disclaimer oh we didn't actually
1:17:59 interesting my why center is around
1:18:01 human connections self-expression and
1:18:03 how these elements can be empowered and
1:18:05 Amplified through technology I'm driven
1:18:07 by the belief that technology especially
1:18:09 Ai and storytelling platforms can
1:18:12 significantly enhance our ability to
1:18:14 share and connect through stories this
1:18:16 passion for technology and storytelling
1:18:18 has been a constant through my career
1:18:20 right this is this is who I am right
1:18:23 it's it's straight out of the thing
1:18:25 um my early fascination with potential
1:18:28 for personal computers to revolutionize
1:18:30 communication to my current work with
1:18:32 Story vine where we enable authentic
1:18:34 storytelling a scale particularly in
1:18:36 healthcare and pharmaceutical sectors my
1:18:39 aim is to explore and push the
1:18:40 boundaries of how Ai and machine
1:18:42 learning can be UTI utilized not just
1:18:44 for efficiency or data processing but to
1:18:47 genuinely augment human creativity and
1:18:50 connection right this is why this
1:18:53 channel exists
1:18:55 to genuinely augment human creativity
1:18:58 and
1:18:59 connection
1:19:01 so that's
1:19:03 it pretty slick right isn't that
1:19:07 fun if I go explore gpts if you look up
1:19:10 C
1:19:12 GPT you should see a pile of
1:19:16 them uh no I guess not uh C GPT
1:19:24 profile
1:19:29 no results
1:19:30 found do I have to do
1:19:44 caps
1:19:49 huh
1:19:52 GPT oh it's CE profile search for C
1:19:56 profile they're called c
1:20:00 gpts yeah search for C
1:20:05 profile and then you'll see him in there
1:20:07 or you'll see some of them here's Rick
1:20:10 lacroy like he you know Rick's a
1:20:13 strategic and and PR guy comms and PR
1:20:16 guy right and
1:20:24 so ain't that cool ain't that
1:20:27 cool ain't that cool I like
1:20:30 it the oh the other thing we built into
1:20:33 these is built into all these
1:20:37 interactions is a call to action for hey
1:20:40 if if you're digging this interaction
1:20:43 you know click here to learn more about
1:20:45 how you can work with Kyle directly so
1:20:47 all of these are they they do a couple
1:20:51 of things one things that they do when
1:20:53 you have when you have a group of
1:20:56 professionals that has these kind of
1:20:58 reputations they don't have a ton of
1:21:00 time right so so if you want to talk to
1:21:03 them you maybe talk to someone who works
1:21:06 with them right or you know you're can
1:21:08 talk to them four months from now so
1:21:10 part of this is about access so 247
1:21:13 someone can interact with my virtual
1:21:15 twin or Joe gaml or Kevin Clark or Rick
1:21:18 lroy um and then within there if you're
1:21:22 liking it a little call to action will
1:21:24 pop up to say here's how to connect with
1:21:25 them right and then where these things
1:21:28 go in the future is we recorded all
1:21:32 these interviews on video on purpose
1:21:36 because as soon as we can turn that into
1:21:39 the talking version of me where you can
1:21:41 just have a conversation with me either
1:21:43 like a phone call or like a video
1:21:45 chat um we're going to do
1:21:48 that and there might be all sorts of
1:21:51 other things we can do and and as the as
1:21:53 the tools get more capable then will
1:21:56 we'll keep evolving though and then in
1:21:59 the meantime we're just interviewing
1:22:01 more and more people from content
1:22:02 Evolution I think there's 50 or 75 I
1:22:04 don't know how big the organization is
1:22:05 it's pretty big we've got 11 done right
1:22:08 now and we're just going to start
1:22:09 interviewing people and spinning these
1:22:12 things up and and what really took us a
1:22:15 while to lock down was the whole process
1:22:18 right so here writing the interview was
1:22:21 a whole process then locking that down
1:22:23 then okay once you do the interview how
1:22:25 do you build the how do you build the
1:22:26 thing there's a whole process for that
1:22:28 and so all of this has been documented
1:22:30 and things like that speaking of
1:22:32 twinning what do you think of nvidia's
1:22:33 Earth 2 for climate
1:22:36 weather predictions um I didn't I didn't
1:22:39 pay enough attention to it to to really
1:22:41 have a specific uh opinion on it yet cam
1:22:44 I will look at it I I just heard about
1:22:46 it briefly I just don't know enough
1:22:47 about it to comment on it but sounds
1:22:49 cool sounds cool conceptually and if
1:22:51 it's
1:22:52 Nvidia they're kicking ass right now so
1:22:55 it's like that that developer conference
1:22:58 they just had in a [ __ ] football
1:23:00 stadium was mindblowing it's like holy
1:23:04 crap all right where can we see the 25
1:23:07 questions that were used right now I
1:23:10 can't put them out because they're it's
1:23:12 they're under content Evolution IP stuff
1:23:16 um I'm going to talk to Kevin about what
1:23:17 I can talk about of that um what I can
1:23:20 probably talk about are the categories
1:23:22 and I'll certainly like how we wrote
1:23:25 most of the
1:23:27 questions
1:23:28 guess it's
1:23:30 me chat
1:23:33 GPT we said we said if you wanted to ask
1:23:37 a series of questions that not only
1:23:40 revealed
1:23:42 um someone's professional experience and
1:23:46 worldview but also their personal and
1:23:49 their you know and their um
1:23:52 Developmental and all that sort of stuff
1:23:54 what would those questions be and then
1:23:56 and then we just took that pile of
1:23:58 questions we edited out the ones we
1:24:00 didn't like I then had it categorized
1:24:04 the questions so it came up with the
1:24:07 categories and then we just went through
1:24:09 there and again this is content
1:24:11 evolution is full of marketers and
1:24:14 writers and people that understand words
1:24:17 and who have done lots and lots of work
1:24:20 with big corporations so we had like I
1:24:24 don't know seven or eight we we have
1:24:26 seven or eight members of What's called
1:24:27 the content Evolution collab which is
1:24:30 the AI Skunk Works and so once we had a
1:24:33 list of categorized questions we just
1:24:35 went through and said what's missing and
1:24:37 and you know someone would say well that
1:24:39 question there that's kind of a dumb way
1:24:41 to ask it but you know here's my
1:24:43 favorite version of that question Joe
1:24:45 gaml shared one of the questions Joe
1:24:47 gaml shared was
1:24:50 um share your share your work history in
1:24:54 three Acts
1:24:56 and it's such a great [ __ ] question
1:24:58 because if you if it's like share your
1:25:00 work history you're like well I went to
1:25:02 this school and then I got this job out
1:25:04 of school and then I did this and then I
1:25:05 did that and then I did that but the
1:25:07 minute you say share your work history
1:25:09 in three acts you immediately go to oh
1:25:12 it's a drama right and then you go oh
1:25:15 like I have to actually compartmentalize
1:25:17 what were the major
1:25:20 Milestones of my
1:25:22 career and it just it just gives you
1:25:25 [ __ ] much better answer and so so we
1:25:27 did that with all the questions and then
1:25:29 once they were set
1:25:31 up um we did we did two test interviews
1:25:37 and and did another round of edits on
1:25:39 the
1:25:41 questions and
1:25:43 then and then we used those test
1:25:48 interviews to work on the prompt and
1:25:50 that took us probably 6
1:25:53 weeks and then we did did two major
1:25:56 iterations of the
1:25:58 prompt
1:26:00 and and then we started building gpts
1:26:02 and we realized we had to structure The
1:26:04 Prompt more because it was hard to
1:26:05 follow and it was hard to change out
1:26:08 stuff and so so now we kind of have a V3
1:26:11 of the prompt that's that's better for
1:26:14 production this is your chat add
1:26:16 reminder that we were going to try some
1:26:18 lip sync stuff tonight oh thanks Brandon
1:26:20 okay yeah I think I'll do that because I
1:26:23 just I just want to at least see what it
1:26:24 is see if it's any good see if it's
1:26:26 there Kyle has the best words thank
1:26:30 you the best
1:26:31 [Laughter]
1:26:37 words words good I like Words Words
1:26:44 tasty man I've been just jabber jawing
1:26:47 tonight ain't
1:26:49 I it's 9:30 I haven't done anything
1:26:53 yet
1:26:57 I hope this is
1:26:58 valuable
1:27:01 um
1:27:03 yeah I hope it's
1:27:11 valuable let's
1:27:16 see yeah I think it's good all right
1:27:19 let's go
1:27:21 play so what we're going to
1:27:23 do
1:27:25 I keep seeing Claude Claude keeps coming
1:27:27 up as you got to [ __ ] do Claude you
1:27:29 got to do Claude claude's better than
1:27:31 chat GPT I think I'm G to kill my chat
1:27:33 GPT thing um so here we are at Claude
1:27:36 Kyle wait always is Kyle don't doubt
1:27:40 that ever oh okay thank you Joy birdie I
1:27:42 appreciate that it's it you know it's
1:27:45 it's so hard sometimes because I'm just
1:27:48 I'm literally just talking to a screen I
1:27:50 know you're all out there and you give
1:27:51 me good feedback and I've got the mods
1:27:52 and things like that that that do give
1:27:55 me you know focused feedback but it's
1:27:59 also
1:28:01 um you know sometimes it's just like
1:28:03 talking talking to a phone so I don't
1:28:06 know sometimes and it's you know I I
1:28:08 learned this as a
1:28:10 performer
1:28:12 which when you're an actor and when you
1:28:14 do a run of a
1:28:18 play there are nights
1:28:23 that you just
1:28:25 you're connected to the
1:28:27 audience and it's effortless and you're
1:28:30 like yep like you know in the moment
1:28:33 you're slaying it you know kind of
1:28:35 before you go on your on stage that
1:28:37 you're going to slay it you come off
1:28:38 stage and you're like I [ __ ] slayed
1:28:40 it and then there's other nights
1:28:43 where it's the exact same words it's the
1:28:48 exact same cast and it just feels like
1:28:53 [ __ ] and here's what I've learned over
1:28:55 the
1:28:57 years that feeling I have as an actor of
1:29:01 how well it went for
1:29:03 me is
1:29:07 100%
1:29:09 independent from the audience that
1:29:11 watched
1:29:12 it I could have a night where I felt
1:29:14 like it was the best thing ever and the
1:29:16 audience was like yeah
1:29:18 whatever I could have a night where I
1:29:19 feel like that was the shittiest
1:29:21 performance in my life and someone will
1:29:23 come up to me afterwards oh my God that
1:29:25 was so amazing I cried so so so the one
1:29:28 thing that I do know is that how I'm
1:29:31 feeling about it is independent of how
1:29:34 it
1:29:36 lands what are the seven categories um
1:29:40 I'll I want to I want to talk to Kevin
1:29:42 Clark and just see what what he's
1:29:44 comfortable sharing because because
1:29:46 there there's there's a there's a decent
1:29:48 chance that we're going to turn some
1:29:51 version of this in into some sort of
1:29:53 practice or
1:29:55 product
1:30:00 um oh that's good thank you Silver
1:30:03 Fox yeah that that person's fine
1:30:11 um that's cool Amelia's
1:30:14 wife um okay so people are talking about
1:30:17 Claude like claud's the [ __ ] so this is
1:30:20 Claude
1:30:22 claw. um I pay the 20 bucks a month for
1:30:24 this so I have access to Claude Opus so
1:30:27 what I want to do here is I want to see
1:30:30 um how well do you know
1:30:34 mid Journey
1:30:37 prompt
1:30:40 best practices
1:30:46 and and
1:30:52 syntax so let me tell you what I'm doing
1:30:55 here I have a solid understanding use
1:30:58 commas and
1:31:01 parentheses okay it so it understands it
1:31:04 understands the arguments like D- AR is
1:31:08 um aspect ratio version
1:31:12 five okay great so what I'm going to say
1:31:19 is um I'll do this with voice so it's
1:31:23 easier
1:31:26 perfect period I'd like you to create me
1:31:30 10 different
1:31:33 characters that are shown in a medium
1:31:36 shot in
1:31:39 16x9 wide format period I want the
1:31:42 photographic style to be
1:31:45 cinematic and dramatic period I want the
1:31:48 characters to be unique interesting
1:31:52 memorable diverse
1:32:01 and people that you would want to get to
1:32:07 know
1:32:11 period write these as mid Journey
1:32:13 prompts using version
1:32:19 six and sprinkle in a little chaos
1:32:28 and
1:32:30 style
1:32:34 variations all right so this should
1:32:37 now give us a big pile of
1:32:45 prompts a wise elderly Native American
1:32:49 Shaman with long braided hair and a
1:32:51 weathered
1:32:52 face standing in a Misty Forest Clearing
1:32:56 cinematic lighting a fierce heavily
1:32:58 tattooed and scarred Female warrior cool
1:33:01 with a shaved head and piercing eyes
1:33:03 we're going to grab
1:33:07 that style dramatic chaos 15 this is so
1:33:11 cool that that um that claude's good at
1:33:15 this cuz chat GPT is not good at this so
1:33:17 this is Claude 3 Opus so Opus is their
1:33:19 big model and they've also got Sonet and
1:33:23 haiku so haiku's really fast and small
1:33:26 sonnet's kind of like chat GPD 3.5 and
1:33:29 then Opus is like chat gp4 or better
1:33:33 they seem to have this thing going
1:33:35 pretty good okay so the way we're going
1:33:39 to do this is I'm going to go
1:33:41 to
1:33:43 Discord oh [ __ ] I can't do
1:33:53 this
1:33:54 I'm I'm out of Discord I'm out of mid
1:33:57 Journey credits
1:33:58 till nine days let me just go confirm
1:34:02 that but I'm pretty sure that's the case
1:34:05 damn
1:34:09 it all right we'll do we'll do it in uh
1:34:11 we'll do it in Dolly it's fine slash IM
1:34:14 it's just that mid journey is so much
1:34:16 [ __ ] better for stuff like
1:34:20 this yeah oh invalid parameter
1:34:25 oh it said version
1:34:30 six needs to be V6
1:34:36 wait oh maybe I just don't I think it's
1:34:48 v6.0 oh it got the all right it it
1:34:51 hallucinated I thought style dramatic
1:34:54 goes
1:34:56 weird all right I'm not as impressed as
1:35:00 I was impressed until I realized it was
1:35:02 all
1:35:04 [ __ ] all right credits exhausted
1:35:06 fine okay so let's go back over here um
1:35:10 so we'll go back
1:35:11 in I just realized that I'm out of mid
1:35:15 Journey credits so I'd like you to
1:35:17 rewrite these 10 promps for
1:35:23 Del
1:35:26 where I just need the prompt and no
1:35:28 arguments at the
1:35:31 end I don't actually need to have it
1:35:35 rewrite all these
1:35:38 but we'll
1:35:46 go
1:35:52 oops all right a fierce heavily here we
1:35:55 go comma and then we're going to go to
1:35:57 chat JPay
1:35:59 now and I think I'll hop over to so if
1:36:04 if you want to make images with do you
1:36:06 can just make them in any old chat GPT
1:36:09 prompt in a in a normal chat session but
1:36:11 what I'm going to do is I'm going to go
1:36:13 explore gpts and then I'm scroll down to
1:36:15 the um to the open Ai gpts and use this
1:36:19 do version so this is this is a GPT that
1:36:23 open an AI built and you get two images
1:36:27 out of this instead of just one um and I
1:36:30 assume they're doing some cool stuff
1:36:31 here with uh with whatever
1:36:34 so let's see uh let's see
1:36:41 wide
1:36:43 16 by
1:36:46 9
1:36:50 photo
1:36:52 of all right
1:36:57 so we'll make the image here then we'll
1:37:00 go to
1:37:01 Runway
1:37:03 and we'll we'll see what the lip sync
1:37:06 thing looks like I haven't actually
1:37:07 played with it over on
1:37:13 Runway Kyle I tried Deco here today
1:37:16 thoughts Deco here is interesting
1:37:20 um I think Deco here is really good if
1:37:23 you want quick quick ideas and if you
1:37:26 want
1:37:29 um if you want
1:37:34 um if you just want to have some fun I
1:37:36 like I feel like Deco here and Korea are
1:37:39 both tools that
1:37:41 are really hard to get consistently good
1:37:44 results out of they're using a thing
1:37:46 called sdxl turbo which only does
1:37:50 like four or five diffusion Cycles
1:37:53 rather than 30 to 50 or 70 um so the the
1:37:57 images tend to be a little more raw um
1:38:01 but but they happen so fast that it can
1:38:03 be a lot of fun it can snap you out of
1:38:05 some some ideas okay so we got
1:38:09 her let's see uh let me you know what
1:38:13 let me go grab
1:38:17 a let me go grab a prompt that I can
1:38:22 talk but uh okay an eccentric elderly
1:38:25 inventor here we go this is
1:38:35 good 16 by9 wide photo of
1:38:46 paste I like Claude I hope I can
1:38:49 maintain my sub I know there's so many
1:38:52 [ __ ] subscriptions now it's just it's
1:38:54 so
1:38:55 bad can't get Claud in Canada um I want
1:38:59 to try there's this new thing out called
1:39:01 magn which is a an upscaling tool that
1:39:03 they just added style transfer to it
1:39:05 looks really cool but it's $ 35 [ __ ]
1:39:08 dollars a month for an upscaler I'm like
1:39:11 I don't like I can't you know in mid
1:39:14 Journey I'm out of you know I'm only
1:39:15 doing the 10 $10 a month thing and I had
1:39:17 a I had a busy month of stuff uh he's
1:39:20 not looking at the camera um make it
1:39:23 more more of a
1:39:26 closeup and
1:39:29 have him
1:39:31 look directly into the
1:39:37 camera so yeah I know I know it's
1:39:41 crazy just allow access to the
1:39:48 internet yeah the fact that the fact
1:39:50 that Claude doesn't have access to the
1:39:52 Internet is just [ __ ] stupid at this
1:39:54 point you know do a [ __ ] deal with
1:39:56 someone do a deal with Bing do a deal
1:39:59 well CL Claude yeah claud's anthropic
1:40:02 they're hosted on Google servers maybe
1:40:04 you could maybe Google's got a search
1:40:06 engine you could use all right so here's
1:40:08 a crazy ass we'll take this crazy ass
1:40:10 dude right here so we're going to
1:40:13 download
1:40:15 that and then we're going to run over to
1:40:18 Runway
1:40:22 ml
1:40:25 and we'll log
1:40:27 in I think I've still got credits left
1:40:30 here generate
1:40:34 anything I'm G to have to brighten this
1:40:36 up a little bit because I can't see
1:40:39 [ __ ] motion brush free
1:40:44 preview upload a
1:40:50 file there's our
1:40:52 dude cute
1:40:54 [Laughter]
1:41:03 so when do I
1:41:14 do get
1:41:16 credits all right let's just generate
1:41:19 this maybe it's after you do it that you
1:41:20 do the lip sync I hope Claude doesn't go
1:41:23 the way of piie I know I
1:41:27 know oh I was wondering why you were
1:41:29 doing that I don't know what I was doing
1:41:32 going to Mid Journey because I'm a dumb
1:41:34 dumb because I forget
1:41:37 [ __ ] regana
1:41:39 Renee just place it in relax mode and
1:41:42 continue on oh can I do that
1:41:45 really hang
1:41:48 on slash
1:41:52 settings
1:41:58 relax
1:42:05 mode look at this see you learned
1:42:08 something every flipping
1:42:16 time 16 by9 aspect
1:42:20 ratio all right this this shit's just
1:42:24 copy go over here slash imagine I knew
1:42:28 that but you know I didn't know that or
1:42:30 I'd have [ __ ] done
1:42:32 it I'll
1:42:35 go D- AR space 16
1:42:42 colon9 credits
1:42:47 exhausted you've run out of hours on
1:42:50 your plan for the
1:42:51 month not just f last hours wait did I
1:42:55 not turn on slash
1:43:02 settings relax
1:43:06 mode oh I'm a cheap bastard I don't have
1:43:09 relax
1:43:10 [Laughter]
1:43:14 mode all right back to Runway where's
1:43:17 Runway here it is
1:43:19 okay there's our little
1:43:22 video
1:43:25 wh
1:43:30 what
1:43:41 no
1:43:44 oh I'd like to recreate Frank lloid ripe
1:43:47 and another arconic Architects and make
1:43:49 them talk that's we're not that far away
1:43:52 we're not that far away from that Mr
1:43:54 studio um I'd say give it 3 or 4 months
1:43:57 and cuz the the basic technology is
1:44:00 there I mean you could you could
1:44:01 actually go do that right now in fact we
1:44:03 can do it with this dude if you want to
1:44:05 see how this works we can go do it and
1:44:08 well here let's
1:44:10 go hang on I'm I'm chat adding here um
1:44:13 let's do
1:44:15 um give me a
1:44:19 cinematic
1:44:21 um portrait of of it might not let me do
1:44:25 this Frank
1:44:26 Lloyd right
1:44:30 standing in front of Falling
1:44:35 Waters it's probably not going to let me
1:44:37 do
1:44:38 it it'll create it and then go all right
1:44:41 it violates our content policy if I make
1:44:44 an animation cartoon using Sora and chat
1:44:47 GPT will that be considered my
1:44:50 IP
1:44:52 um it's all of that stuff right now is
1:44:56 up for grabs I the the burden on of
1:45:00 proof is on you in um proving human
1:45:06 contribution oh it did it but there's
1:45:10 two of
1:45:13 him I'd like
1:45:17 him
1:45:20 um facing the camera in a
1:45:25 camera in a
1:45:29 closeup um
1:45:34 with just the
1:45:37 house
1:45:39 behind
1:45:43 him not the
1:45:51 waterfalls oh that was weird oh that was
1:45:53 cuz the of these roll over
1:46:00 things you need to stop abusing your
1:46:02 credits my brother well here's the thing
1:46:04 I wasn't using I I I hadn't used mid
1:46:07 journey in like six months and then they
1:46:10 added character reference and they added
1:46:12 style reference and then I had a bunch
1:46:15 of ideas for something and I used it a
1:46:17 bunch and then I used it too
1:46:21 much but I'm not going to spend [ __ ]
1:46:24 60 bucks a month for Mid journey and 35
1:46:26 bucks a month for an upscaler and 20
1:46:28 bucks a month for chat GPT and 20 bucks
1:46:30 a month for it's it's like enough enough
1:46:33 you
1:46:36 frakin brilliant technologist that I
1:46:38 love I love you all so
1:46:45 [Music]
1:46:47 much come on oh this going to [ __ ] this
1:46:50 not going to do it it's we're going to
1:46:52 get an
1:46:54 message aren't
1:47:00 [Music]
1:47:01 we it did falling wat's pretty good this
1:47:04 one on the left's pretty accurate I
1:47:12 think come on all right why did it why
1:47:15 is it doing two of
1:47:17 him a cinematic oh portraits God damn
1:47:22 it
1:47:28 cinematic oh wait hang
1:47:34 on
1:47:41 no I don't know why it did
1:47:43 it give me a
1:47:47 cinematic closeup
1:47:52 portrait
1:47:55 a Frank Lloyd Wright standing
1:47:59 in one of his iconic living
1:48:05 rooms let's see how that does oh
1:48:11 it's copy
1:48:14 paste in
1:48:18 16x9 wide
1:48:20 format I can't believe Runway
1:48:23 made the inventor [ __ ] disappear I I
1:48:27 just want him moving his stupid head I I
1:48:30 don't want him going back to the
1:48:34 future [ __ ] dumb
1:48:40 Dums it's giving you a choice of two oh
1:48:43 is it am I really that Dopey I
1:48:48 am I forg I forgot Silver Fox I forgot
1:48:53 was in I I told you all to go to Del so
1:48:55 you get two images and then I forgot I
1:48:57 was in
1:49:03 it oh good lord all right so let's go
1:49:07 let's go let's go do something I'm
1:49:11 confident that I can do now wait if I go
1:49:15 to did I can upload an audio file
1:49:20 yeah all right so for whoever was
1:49:23 wanting to make the Frank Lloyd Wright
1:49:25 thing they're probably long gone by now
1:49:27 like whatever I was trying to watch this
1:49:29 guy use his computer it was really
1:49:31 painful he just didn't know what he was
1:49:32 doing I don't know why he's got an
1:49:34 education Channel what are his
1:49:36 qualifications they seem
1:49:39 low.com so I'm going to
1:49:43 d.com and I'm gonna log on
1:49:48 in I'll tell you what all right so
1:49:51 here's a bunch of other things I've made
1:49:53 so I'm going to say create a
1:49:56 video and then we're going to
1:50:00 upload Ye Old Frank Frank Lloyd Wright
1:50:03 nope not
1:50:06 him he's
1:50:14 dashing H Hang on we're not because it's
1:50:17 in the stupid [ __ ] file format I this
1:50:21 is just staggering
1:50:23 that that nobody supports this new super
1:50:27 exciting file
1:50:40 [Music]
1:50:43 format come on where is it did I save it
1:50:45 to
1:50:47 desktop
1:50:49 yes all
1:50:52 right so now we have
1:50:56 Frank and then I can upload audio so now
1:50:59 what we're gonna do is we're gonna go to
1:51:01 Quick
1:51:05 time and I'm going to do new Audio Oh
1:51:08 there's Frank you want to see
1:51:12 [Music]
1:51:18 Frank oh that's not
1:51:22 Frank
1:51:23 like the [ __ ] is that Frank doesn't sing
1:51:27 Frank or Andrew Lloyd Weber ballads
1:51:30 let's get over to Farber at the
1:51:31 education desk he's my number two hey
1:51:34 all right anyway we're going to do new
1:51:36 audio
1:51:41 recording and then oh we can do this a
1:51:43 bit longer so we'll do
1:51:46 uh let's
1:51:52 see
1:52:02 I think it's important for the lines of
1:52:04 the house to uh match the uh
1:52:08 surroundings
1:52:09 and integrate seamlessly with nature and
1:52:13 uh I like furniture that looks amazing
1:52:15 but is uncomfortable to sit on I don't
1:52:19 think people should be comfortable I
1:52:20 think they should be inspired to get up
1:52:23 and go out in nature which is
1:52:26 represented by these very large
1:52:28 windows and the lamp is just
1:52:31 nice all right so now we have that let's
1:52:34 hear if it's any
1:52:36 good I think it's important for the
1:52:38 lines of the house to uh match the uh
1:52:42 surroundings and all right whatever I
1:52:44 have no idea what he sounds like so he's
1:52:46 just going to sound like a pompous ass
1:52:48 right now um and then we're going to
1:52:51 save
1:52:55 and my um my file format my file saving
1:52:59 format is save everything as Untitled to
1:53:02 the
1:53:03 desktop and then you can find it later
1:53:05 if you really want
1:53:07 it um
1:53:12 Frank
1:53:13 not hot
1:53:19 dog all right so now we're going to
1:53:21 upload audio
1:53:23 and we're going to we're going to go
1:53:24 find our file Frank not hot
1:53:27 dog out of Silicon Valley you know the
1:53:30 not hot dog hot dog not hot dog
1:53:35 app all right audio is uploading and now
1:53:38 we're going to
1:53:39 have Frank lyd Wright
1:53:44 talking I think it's important oh wait
1:53:46 generate video all right let's
1:53:51 go
1:53:56 and then there it'll be let's see is
1:53:58 this
1:53:59 one I think this was a A Partners ever
1:54:02 fancied to thrilling ride here at
1:54:05 stallion Spirit Ranch our stallions are
1:54:07 more than just a pretty sight they're
1:54:09 strong eager and ready for some fun
1:54:13 whether you're an experienced rider or
1:54:14 just looking for a new kind of
1:54:15 excitement we promise a ride that'll
1:54:17 leave you breathless so saddle up and
1:54:20 join us for an unforgettable Adventure
1:54:22 for every turn is a new Temptation
1:54:35 yeeha oh what's this dumb
1:54:41 one
1:54:51 nothing
1:54:53 I think it's important for the lines of
1:54:55 the house to uh match the uh
1:54:59 surroundings and integrate seamlessly
1:55:02 with nature and uh I like furniture that
1:55:06 looks amazing but is uncomfortable to
1:55:09 sit on I don't think people should be
1:55:11 comfortable I think they should be
1:55:13 inspired to get up and go out in nature
1:55:16 which is represented by these very large
1:55:19 windows and the lamp is just nice
1:55:25 there you go you can make Frank Lloyd
1:55:27 Wright talk today you too can have Frank
1:55:29 Lloyd right now both did and hen have
1:55:34 technology uh where you're supposed to
1:55:37 be able to chat with it and it will chat
1:55:39 back like you type in text and it will
1:55:40 talk back to you but both of them are
1:55:43 staggeringly bad so I don't think
1:55:46 they're worth paying for or even playing
1:55:49 with at this point
1:55:51 so annual plan offer three days
1:55:56 left no that's the other thing don't buy
1:56:00 annual plans right now things are moving
1:56:02 too
1:56:03 fast don't do it you're not going to
1:56:05 save any money because you're just going
1:56:07 to be pissed off because something
1:56:08 better came along you're like but I got
1:56:10 this annual thing so I'm going to keep
1:56:11 using the shitty thing all right it's
1:56:15 getting to be that time
1:56:18 people was this useful was that useful
1:56:20 that that that Chad add
1:56:23 moment where I still didn't figure out
1:56:25 how to make Runway lip sync I'll figure
1:56:28 that out before I get on Kyle can you
1:56:30 remind me what tool that is again that
1:56:33 one that we were just using for Frank
1:56:35 lyd Wright was D- id.com
1:56:39 there's another one called haen which
1:56:42 does something similar there's another
1:56:43 one called Synthesia which does
1:56:45 something similar all of them were
1:56:47 pretty expensive so if you're thinking
1:56:50 like when I first saw d ID and Synthesia
1:56:55 and hey Jen my immediate thought was
1:56:58 holy [ __ ] just go have chat GPT generate
1:57:01 a whole bunch of scripts generate them
1:57:03 into videos and create a whole Tik Tok
1:57:04 Channel but it's pretty expensive on a
1:57:08 per minute basis to generate those
1:57:10 things so
1:57:12 um there's there'll be free versions and
1:57:15 really cheap versions soon enough it's
1:57:16 just not there right now get up and get
1:57:19 out in nature and the lamp is just nice
1:57:21 that's awesome thank
1:57:23 you I am known for my impromptu script
1:57:28 writing that's how Frank came into
1:57:30 [Laughter]
1:57:32 existence oh man
1:57:38 um can you remind me what tool that is
1:57:41 again we did that amazing yeah I know
1:57:44 thank you did I miss the 01 light video
1:57:47 oh no we should we should look at the
1:57:49 video um before I get out of here
1:57:53 then we
1:57:55 go
1:57:57 here and we will go
1:58:02 uh I think it's
1:58:05 01 open
1:58:15 interor a computer that's as easy to
1:58:18 interact with for over a 100 years
1:58:21 Humanity has imagined a computer that's
1:58:23 as easy to interact with as a person
1:58:26 chat GPT introduced the world to this
1:58:28 technology and became the fastest
1:58:30 growing consumer application of all
1:58:33 time now corporations around the world
1:58:36 have started Building closed Source
1:58:38 artificially intelligent
1:58:40 devices 6 weeks ago we realized there
1:58:43 was an opportunity to build an open-
1:58:45 Source Foundation this next generation
1:58:48 of computers the Linux of this space
1:58:51 something open
1:58:53 flexible and free an operating system
1:58:57 with a language model at the center of
1:58:59 it I'm holding the first device powered
1:59:02 by this operating system in my hand
1:59:04 right now it's called the
1:59:06 [Music]
1:59:07 [Applause]
1:59:12 [Music]
1:59:21 01
1:59:25 [Music]
1:59:38 this is the o1 light the first
1:59:40 open-source language model computer you
1:59:43 talk to it like a person and it operates
1:59:45 a computer to help you get things done
1:59:48 you can run the language bottle on your
1:59:50 computer at home or in our server in the
1:59:52 Cloud making the o1 light fully portable
1:59:55 when connected to Wi-Fi or a hotspot to
1:59:58 use it you just push talk and then it
2:00:01 thinks and responds let me show you what
2:00:04 it can
2:00:08 do when's the next sunny
2:00:11 day one
2:00:16 moment
2:00:18 so so it's basically Tethered to Wi-Fi
2:00:20 to his home laptop right and here's the
2:00:25 home
2:00:26 laptop and it's so this is exactly what
2:00:29 the rabbit R1 is doing so the the rabbit
2:00:31 r1's that orange thing with the button
2:00:33 on the side and you talk to it and it's
2:00:36 doing the same thing and in fact this is
2:00:38 an the software here is an open version
2:00:41 of that same that same technology and
2:00:44 what they're saying is they've got up in
2:00:46 the cloud they they've got a a version
2:00:48 of this that you can just buy this
2:00:50 device and use their thing or you can
2:00:53 just go to the site right now and
2:00:54 download the server and all of this
2:00:56 techn it's all free it's all open-
2:00:58 Source free stuff so you can do what
2:01:01 they're showing in the uh in this
2:01:08 demo the next sunny day is on March 9th
2:01:12 great am I doing anything on that
2:01:14 day let me check your
2:01:19 calendar you have no events scheduled
2:01:21 for March 9th so because the 01 is
2:01:24 operating my personal computer at home I
2:01:26 didn't need to set up my calendar it can
2:01:28 just use
2:01:29 it nice can you see if there are any
2:01:32 concerts in Seattle on that
2:01:38 day yes you have several options Option
2:01:41 One Bad Bunny will be performing at the
2:01:43 climate pledge Arena that's perfect can
2:01:46 you add that to my calendar and then
2:01:48 copy and paste a link to the ticket into
2:01:49 the calendar event and then actually if
2:01:52 you could text Page the link and say
2:01:53 something like hey do you want to go to
2:02:02 this so if I don't know if you're
2:02:05 noticing it but it's like it's flipping
2:02:06 between apps right so it went to the
2:02:09 calendar app got the calendar link
2:02:11 copied it now it's going over to the
2:02:13 messages app and is texting their
2:02:20 buddy done the concert is on on your
2:02:22 calendar and Page has been texted
2:02:25 awesome
2:02:26 thanks you're welcome by the way you've
2:02:29 just received an email from Mike asking
2:02:31 about your progress on an article would
2:02:34 you like to respond to it now so I need
2:02:37 to send that file to Mike but it's on my
2:02:39 computer at
2:02:40 home hey that article should be on my
2:02:43 desktop it's called the universal
2:02:46 interface there's a file titled the
2:02:48 universal interface. PDF on your desktop
2:02:51 do you want me to send it to my Mike now
2:02:53 at this point I could ask my light to
2:02:55 read or even edit that document but I'm
2:02:57 just going to send it to
2:02:59 Mike yes please send
2:03:07 it the document has been emailed to
2:03:10 Mike great so you
2:03:15 can toop and it's the one that the light
2:03:17 has been operating out of the box it's
2:03:20 pretty good at using my computer but it
2:03:22 can't do everything so let's teach it to
2:03:25 use a desktop application that it's
2:03:26 never
2:03:29 seen hey I want to teach you
2:03:32 something sounds good what do you want
2:03:34 to teach me to do I'm going to show you
2:03:37 how to send a slack
2:03:40 message okay I'm ready what's the first
2:03:43 step first open
2:03:49 Slack it looks like slack is open did I
2:03:52 do do that right yeah you
2:03:55 did great what's
2:03:57 next hit command
2:04:02 K right so just training it how to use
2:04:05 we're watching yeah 01 Lightroom so this
2:04:08 is an opsource version of the of the
2:04:11 rabbit
2:04:12 R1 uh it the the hardware is different
2:04:15 here the hardware here is just a a push
2:04:17 button and a speaker speaker and a
2:04:19 microphone combo doesn't have a screen
2:04:21 doesn't have a scroll wheel
2:04:22 but the the ability to train it to use
2:04:25 applications on your desktop is the same
2:04:28 the other thing about this is let's see
2:04:32 let's go
2:04:34 to so here here it is so you can
2:04:37 pre-order this thing the first batch
2:04:39 already sold out and it also went
2:04:45 from yeah we're in batch two right now I
2:04:48 think oh wait pre-order of the light bat
2:04:52 five sold out batch six open now so
2:04:55 they've got I don't know how many are in
2:04:56 a batch I assume 10,000 um so these
2:05:00 things are selling really quick um but
2:05:04 you can also um it just says build your
2:05:07 own if you go build your own the GitHub
2:05:11 repository has um documentation it's got
2:05:15 CAD drawings of the the thing itself
2:05:18 it's got wiring schematics it's got all
2:05:20 the stuff here's a fun thing you can do
2:05:23 if if you're like I'm intimidated by
2:05:33 um oh why the voice sounded male in that
2:05:36 in in that video I think it's just a
2:05:37 trans fluid person that's all um or
2:05:41 gender fluid person
2:05:44 um the
2:05:47 um so if you're intimidated by GitHub
2:05:50 like if you want to play with this and
2:05:51 install it so here here is something I
2:05:53 did today which is kind of fun
2:05:56 um let me go here and I'll go to um AI
2:06:01 Studio oh no that's AI
2:06:04 Salon AI
2:06:07 studio. goole.com
2:06:09 so here I'm using Gemini
2:06:14 um let's see my
2:06:16 [Music]
2:06:18 library so installing open open
2:06:21 interpreters so
2:06:23 um so I went into Gemini and I said if I
2:06:27 want to get the server up and running on
2:06:29 my Mac and use the chat interface on my
2:06:32 iPhone what are the stepbystep uh what
2:06:35 are the step-by-step instructions here's
2:06:37 the repo and then I paste it in the
2:06:39 GitHub and then Gemini just reads the
2:06:43 GitHub um repository and reads all the
2:06:46 documentation and it gives me
2:06:47 step-by-step instructions to install
2:06:49 this so even if you're not that
2:06:51 Technical and don't know how to navigate
2:06:53 around GitHub you can just go to the the
2:06:56 site and
2:06:58 uh and just have a large language model
2:07:01 teach you how to install it if you want
2:07:02 to geek out um but it's pretty slick so
2:07:06 anyway that's that um someone else had a
2:07:09 question that I saw in here um they
2:07:12 asked about the new teams function in
2:07:15 chat
2:07:18 GPT that was Suzanne came in and asked
2:07:20 that right it's called the open one
2:07:24 Lightroom no it's called
2:07:26 the open hang
2:07:29 on 01 like 01 is the is the go to open
2:07:35 interpreter.
2:07:37 comom
2:07:41 SL1
2:07:48 here go there go there
2:07:53 Becky Ru you need teams for the new
2:07:55 features no I don't know the new
2:07:57 features what are the new features
2:07:59 Suzanne let me go
2:08:01 look let me switch over to my little
2:08:03 teams
2:08:09 account what are the new features
2:08:13 settings data control archive
2:08:17 chat multiactor authentication is that
2:08:21 new
2:08:26 speech manage my
2:08:29 workspace that all looks the
2:08:32 same
2:08:39 settings this doesn't look any different
2:08:41 for me
2:08:42 Suzanne Version Control for gpts and
2:08:45 permission based sharing links
2:08:50 oh
2:08:59 change who has
2:09:02 access start with create a GPT
2:09:14 [Music]
2:09:20 okay
2:09:26 I don't see new stuff here start by
2:09:29 defining your
2:09:30 GPT
2:09:33 GPT
2:09:36 to
2:09:38 explain
2:09:41 physics using only jelly
2:09:50 beans
2:09:55 go to a GPT you created in teams go to
2:09:57 save it and you'll see the features
2:10:01 okay yeah I'm just creating a new one
2:10:03 now that's what I figured would was it
2:10:05 would be in in
2:10:15 saved
2:10:20 sure
2:10:29 [Music]
2:10:38 oh nice Jelly
2:10:40 Bean all right
2:10:45 create oh share with ADD people from
2:10:50 your workspace
2:10:54 oh nice anyone at content
2:10:57 Evolution publish to content
2:11:00 Evolution anyone with the link published
2:11:03 to the GPT
2:11:05 store that's pretty
2:11:09 slick oh and then you can just type
2:11:11 people in here but they have to be from
2:11:14 your workspace
2:11:18 interesting anyone at content Evolution
2:11:21 with the
2:11:23 link publish oh oh okay I get it so if
2:11:27 you've got a hundred people in
2:11:32 here you can kind of semi-privately
2:11:35 distribute within the team and then you
2:11:37 could put it in your own mini store all
2:11:40 right cool no I hadn't seen that thanks
2:11:42 for pointing that out
2:11:43 Suzanne all right I'll dig into that a
2:11:45 bit
2:11:46 more yeah uh chat GPT yesterday was a
2:11:49 little janky so um I I had a feeling
2:11:52 they were updating something so there
2:11:54 you go whenever chat GPT gets really
2:11:57 janky chances are they're upd updating
2:12:00 servers or something doing big updates
2:12:03 that's generally been my experience I
2:12:04 don't know what's up with the hair
2:12:05 tonight it's very soft it's very soft
2:12:08 and
2:12:10 lovely
2:12:15 [Music]
2:12:20 um
2:12:26 hey Kyle I've been checking my mailbox
2:12:28 and your sign copy of your book has
2:12:30 still not arrived I I haven't done any
2:12:33 sign copies of my book If there if there
2:12:35 is demand I will happily sign
2:12:39 them wow a whole new level of creation
2:12:41 of gpts nice yeah
2:12:44 exactly Version Control too you can even
2:12:47 revert back to old versions that's
2:12:48 actually really good cuz like cuz all
2:12:51 those all the life's tough it's tougher
2:12:54 when you're stupid all those gpts I did
2:12:57 for Content Evolution like at some point
2:12:59 we're going to start iterating on those
2:13:01 and if we if we roll out a thing that is
2:13:03 [ __ ] up the ability to go back on them
2:13:05 is going to be good so that's
2:13:07 great all right cool I'm G to actually
2:13:10 have to learn that I made a a slide for
2:13:14 you and something I presented tonight
2:13:16 I'll post it and show and tell okay
2:13:17 great perfect thank you Suzanne that's
2:13:19 awesome all right everybody listen
2:13:22 um great hanging out with y'all tonight
2:13:24 I hope this was this was useful useful
2:13:27 and
2:13:28 good um if you want to support this
2:13:33 channel there's a few of you in here
2:13:36 some new people in here there's a couple
2:13:37 of things you can do one is you should
2:13:39 check out the AI salon so go to the
2:13:42 salon. go to the welcome to the salon
2:13:45 area read about what we're about it's a
2:13:47 short little read it's got our values in
2:13:49 there it's what the what the things
2:13:51 about but this is like a group of AI
2:13:53 adventurers we've got like 13 or
2:13:56 1,400 AI adventurers in there so if
2:13:59 you're trying to find community that has
2:14:02 their that is actively trying to figure
2:14:05 this AI stuff out too that's who this is
2:14:08 that's who's
2:14:09 here so join that if you resonate with
2:14:14 the values if not don't because you
2:14:16 won't like
2:14:18 it go to the community section introduce
2:14:22 yourself use your real name and um and
2:14:27 get to know people just start to get to
2:14:28 know people that the whole community in
2:14:31 the next two to three years is going to
2:14:34 be
2:14:35 critical so get in a community like this
2:14:38 and get yourself known and and if you're
2:14:41 like but Kyle I don't know anything
2:14:43 about Ai and everyone in there is going
2:14:44 to know more than me everyone in there I
2:14:47 promise you right now everyone in there
2:14:49 thinks that everyone else knows more
2:14:51 than them because you know what they do
2:14:53 and they don't there are no experts
2:14:55 right now nobody knows what's going on
2:14:58 you're not you're not as smart as you
2:15:01 think you are with this [ __ ] cuz
2:15:02 everything's about to change and you're
2:15:04 not as far behind as you think you are
2:15:05 with this [ __ ] even if you're just
2:15:07 getting started we're very very early
2:15:10 thank you
2:15:11 Empower we're very very early so jump
2:15:14 into community and get to know people
2:15:16 and just be generous like read our
2:15:17 values curiosity generosity exploration
2:15:22 collaboration um
2:15:24 inclusion empathy
2:15:29 gratitude this group is those values
2:15:33 it's
2:15:34 amazing there's there's very few other
2:15:37 communities out there that have what's
2:15:39 going on in here going on thank you very
2:15:41 much I really appreciate that it's very
2:15:44 generous you're a title gifter I didn't
2:15:46 know I don't know what that means but
2:15:48 it's very fancy you've got a title now
2:15:50 I'm going to try out did to make a true
2:15:52 crime AI shorts thanks Kyle I'll be up
2:15:55 all night you're welcome gritty wisdom
2:15:57 that's if I can do anything it's so so
2:16:00 the people that come here regularly they
2:16:02 use me in Li of Xanax to put themselves
2:16:05 to sleep but occasionally I'll show you
2:16:07 something new and you'll be up all night
2:16:09 so so I cause both insomnia but I also
2:16:12 cure
2:16:15 it okay so check out the AI Salon it's
2:16:19 awesome there's guilds we have our own
2:16:21 club Club in there called the
2:16:22 Irregulars so the Irregulars are people
2:16:25 that show up here night after night
2:16:27 after night I'm here seven nights a week
2:16:30 two to three hours a
2:16:31 night guess my family thinks I'm
2:16:34 absolutely batshit crazy and they're
2:16:40 right
2:16:42 um so just show up you can also please
2:16:46 follow my channel I'm like 200 people
2:16:48 away from 35,000 followers so follow my
2:16:51 channel if you're not following what the
2:16:53 [ __ ] um yep can never can never sleep
2:16:56 after these inspirational Tik toks
2:16:58 that's good I think that's good right um
2:17:01 You can also subscribe to the lives
2:17:03 which supports me which would be great
2:17:04 you can also buy in the corners there's
2:17:06 a video series called underhyped which
2:17:09 is I think 17 or 18 videos where I'm
2:17:11 talking about why I think AI is a big
2:17:13 deal um the parallels between the early
2:17:16 days of the web and what's going on with
2:17:18 generative AI now how to use these tools
2:17:20 how they work work how image generation
2:17:23 tools work that's in that video series
2:17:25 um yeah that's it so just keep coming
2:17:27 back so thank you to the mods for
2:17:29 modding tonight um and every night
2:17:31 really um thank you for making my life a
2:17:34 ye a wead um saner more sane um thank
2:17:39 you to my subscribers um thank you to
2:17:41 the Irregulars people that just show up
2:17:43 here all the time sometimes people come
2:17:46 in and they're like well how do I become
2:17:48 an
2:17:49 irregular you just show up
2:17:53 here here's the
2:17:55 deal if if you show up multiple nights a
2:17:59 week to watch a 58-year-old man's Tik
2:18:02 Tock Channel while he Rambles
2:18:04 incoherently for
2:18:06 hours you're
2:18:09 irregular so just show up and you're in
2:18:12 the club that's how it
2:18:17 works and then when people say you're
2:18:19 weird you go I'm not weird I'm
2:18:23 regular that's the that's the that's the
2:18:26 game we play here all right well this
2:18:29 has been fun tonight um hope you enjoyed
2:18:32 peace out tomorrow's Friday night date
2:18:33 night so if you have a date bring them
2:18:36 here if you don't have a date you got
2:18:37 one with me let's either way it's Friday
2:18:40 night date night
2:18:43 um it's going to be late it's going to
2:18:46 start late because Gabby was supposed to
2:18:49 have her art opening last
2:18:51 week but it snowed the night before and
2:18:54 they canceled it and then it was
2:18:56 beautiful the next day so whatever so
2:18:58 the opening is actually tomorrow so I'm
2:19:00 going to be late tomorrow so it'll be
2:19:02 900 or 9:30 mountain time so drink your
2:19:05 coffee get your date get your nachos
2:19:08 whatever you do to prepare all right
2:19:10 maybe we could get Frank to tell us a
2:19:12 bedtime story every night before signing
2:19:14 off actually I probably should do that
2:19:16 maybe at like 9:30 cuz cuz normally I
2:19:19 try to you know sort of sign off around
2:19:21 I'm 10 but maybe what I'll do is at at
2:19:23 9:30 every night I'll do a I'll do a
2:19:25 Frank
2:19:26 video that was only last week I know I
2:19:29 know it's crazy isn't it Silver Fox I
2:19:32 know it's like there's just too much
2:19:34 happening right now there's too much
2:19:36 there's too much yeah just too much
2:19:40 anyway all right thanks again Kyle thank
2:19:42 you Becky Tobias good to see you jenu
2:19:45 thank you as always
2:19:48 Joey what did you bring up Nacho I want
2:19:51 some
2:19:52 sorry kin from Cleveland Cosmic lover
2:19:55 good to see you Silver Fox Becky Ru ASU
2:19:58 good to see you
2:20:00 drew I'm celebrating my anniversary
2:20:04 fantastic I'll see you here
2:20:07 tomorrow I think I did did I do did I do
2:20:12 a Valentine's Day one that was my
2:20:13 anniversary so I think I did one of
2:20:16 these on my anniversary we're not
2:20:18 decohere we're incoherent exactly now
2:20:21 Joker laughing Thunder good to see you
2:20:23 Joy pie thank you as always Milo's wife
2:20:26 great to see you Tams or Tam Z Tams Tams
2:20:31 Tams must be Tams Source Camp great to
2:20:33 see
2:20:36 you Gavin bitty a human
2:20:39 being Silver Fox
2:20:42 Vicky all right everybody I'm out
2:20:45 checking on out of here Vegas Moxy good
2:20:47 to see you good night good night
2:20:48 everybody and uh see you for date night
2:20:50 bye okay