
AI Learning Lab
Mar 22, 2024 - Transform Your Creativity and Strategy Today

Video2024-04-072:16:2618 views
Description
In a captivating exploration of the evolving landscape of artificial intelligence, Kyle Shannon delves into the intersection of technology and human experience, emphasizing the transformative potential of AI in various sectors, including education, healthcare, and creative industries. He discusses the implications of generative AI, highlighting its ability to enhance productivity while also raising concerns about job displacement and ethical considerations. Through engaging anecdotes and interactive discussions, Kyle encourages viewers to embrace curiosity and adaptability in this rapidly changing environment. His insights reflect a deep understanding of both the opportunities and challenges posed by AI, making it clear that the future will require a balance between technological advancement and human values.
For more thought-provoking content, check out his TikTok channel: [AI Learning
Lab](https://tiktok.com/@aiLearningLab).
#ArtificialIntelligence #GenerativeAI #AIinEducation #FutureOfWork #TechEthics #AICommunity #Innovation
Chapters:
00:00:00 Intro
00:02:25 Dune
00:03:09 English brick tea
00:04:35 Questions & requests
00:07:01 Ai startups
00:10:16 Shipping announcements
00:10:52 Grock open source
00:11:07 Kyle shannon dreams mug
00:11:38 Cocoa spit take
00:12:27 Questions
00:14:04 Dave portnoy analogy
00:17:14 Ai service business
00:18:40 Proprietary data
00:20:02 Rabbit r1 discord link
00:22:05 Ai salon community
00:24:21 Cam katkin intro
00:25:16 Saturday night live
00:25:41 Running a university
00:26:41 Beginners resources
00:27:29 Importance of community
00:28:08 Starting with chatgpt
00:29:26 Underhyped series
00:30:04 Where do i start quick start guide
00:31:33 Share your work
00:32:29 Chat
00:34:00 Custom gpt's
00:34:18 School ai policies
00:35:06 City of boston ai guidelines
00:35:39 How to make your own gpt
00:38:05 Sin city sipper gpt
00:40:05 What does ai mean?
00:41:58 Artificial vs. machine learning
00:47:47 Dune explorer gpt
00:51:57 Grandma's recipe book gpt
00:52:45 Chatgpt & spreadsheets
00:56:22 Dune
00:58:58 Official chatgpt app
01:00:52 Champi wants a snack
01:02:03 Captcha puzzle
01:02:28 Fake sora app
01:02:42 Chatgpt & zapier
01:03:48 Zapier gpt
01:04:36 Chatgpt & standup comedy
01:07:42 South park scripts
01:11:41 Devon - autonomous agent
01:13:56 Cusp of generative ai
01:16:01 Gpt for biology tutoring
01:21:59 Mitoch
01:23:00 Chatgpt vs. google search
01:24:29 Biology tutoring success
01:24:50 Learning together
01:25:40 Pi - biology tutoring
01:28:33 Ai job replacement
01:28:46 Artificial general intelligence
01:29:45 Artificial superintelligence (asi)
01:31:45 Ai literacy & job stability
01:34:45 Ai & unsolvable problems
01:36:11 Organic intelligence
01:39:19 November 30, 2022
01:41:59 Current capabilities of ai
01:43:19 Transition period of ai
01:45:21 Powers
01:46:40 Ai 2041 book
01:48:50 Ai-powered cards
01:51:57 Figure one ai robot
01:53:37 Social net for un
01:54:44 Sam altman & worldcoin
01:56:51 Post-sci-fi world
01:57:27 Humane ai pin
02:00:00 Ar vs. vr
02:03:13 Alice to the moon
02:03:23 Neuralink
02:04:53 Apology & closing thoughts
02:06:45 Steve jobs
02:09:51 Fear of ai
02:11:47 Generative ai vs. nuclear bombs
02:14:15 Ai salon community
Chapters
0:00Intro2:25Dune3:09English brick tea4:35Questions & requests7:01Ai startups10:16Shipping announcements10:52Grock open source11:07Kyle shannon dreams mug11:38Cocoa spit take12:27Questions14:04Dave portnoy analogy17:14Ai service business18:40Proprietary data20:02Rabbit r1 discord link22:05Ai salon community24:21Cam katkin intro25:16Saturday night live25:41Running a university26:41Beginners resources27:29Importance of community28:08Starting with chatgpt29:26Underhyped series30:04Where do i start quick start guide31:33Share your work32:29Chat34:00Custom gpt's34:18School ai policies35:06City of boston ai guidelines35:39How to make your own gpt38:05Sin city sipper gpt40:05What does ai mean?41:58Artificial vs. machine learning47:47Dune explorer gpt51:57Grandma's recipe book gpt52:45Chatgpt & spreadsheets56:22Dune58:58Official chatgpt app1:00:52Champi wants a snack1:02:03Captcha puzzle1:02:28Fake sora app1:02:42Chatgpt & zapier1:03:48Zapier gpt1:04:36Chatgpt & standup comedy1:07:42South park scripts1:11:41Devon - autonomous agent1:13:56Cusp of generative ai1:16:01Gpt for biology tutoring1:21:59Mitoch1:23:00Chatgpt vs. google search1:24:29Biology tutoring success1:24:50Learning together1:25:40Pi - biology tutoring1:28:33Ai job replacement1:28:46Artificial general intelligence1:29:45Artificial superintelligence (asi)1:31:45Ai literacy & job stability1:34:45Ai & unsolvable problems1:36:11Organic intelligence1:39:19November 30, 20221:41:59Current capabilities of ai1:43:19Transition period of ai1:45:21Powers1:46:40Ai 2041 book1:48:50Ai-powered cards1:51:57Figure one ai robot1:53:37Social net for un1:54:44Sam altman & worldcoin1:56:51Post-sci-fi world1:57:27Humane ai pin2:00:00Ar vs. vr2:03:13Alice to the moon2:03:23Neuralink2:04:53Apology & closing thoughts2:06:45Steve jobs2:09:51Fear of ai2:11:47Generative ai vs. nuclear bombs2:14:15Ai salon community
Transcript
0:03 come on come 0:07 [Music] 0:22 [Music] 0:28 [Music] 0:29 on 0:33 [Music] 0:38 change seems 0:40 theams Bunch 0:44 CLS get my nve up but my past is pulling 0:48 me 0:52 down 0:54 Long Black 0:56 Sheep 0:59 around 1:03 happy Saturday night 1:07 everybody champy says 1:11 [Music] 1:29 hello 1:33 [Music] 1:37 [Applause] 1:38 [Music] 1:43 in a Westerly 1:47 Direction this car is my 1:51 train I've been driving I've been 1:55 [Music] 1:56 wondering is I'm running from 1:59 again 2:02 feel like an anyy old 2:05 man holding on to 2:09 29 dou a head on that 2:13 [Music] 2:14 Horizon is a California 2:17 [Music] 2:25 line okay happy Saturday Night Live we 2:30 got Saturday Night Live sorry I'm a 2:31 little late had to take the boys up to 2:34 Boulder going to see Dune 2 with the 2:38 friends and of course you know they 2:40 planned it for four and 2:42 then 2:45 cuz cuz they are the age they are they 2:48 got it together by 2:50 [Laughter] 2:53 eight ah youth ah 2:59 youth 3:00 [Music] 3:09 so if you're new here my name is Kyle 3:11 Shannon I see you drank the bottle on 3:13 the Shelf no this is this is the old 3:16 one that's the new one I haven't had any 3:20 any of the new ones since 3:22 uh I haven't had any of the new one 3:24 since uh since I cracked 3:26 [Music] 3:29 it 3:34 [Music] 3:58 the tea tonight is English brick tea 4:02 hot it's steeping right 4:05 now cuz there wasn't coffee and I didn't 4:08 have time to make it cuz I was late so I 4:09 just heated up water 4:12 um and a water lovely Rocky Mountain tap 4:16 [Music] 4:18 water he's your singing pal he's such a 4:21 good dog he's a good dog that dog you 4:24 know he's one of those dogs where you 4:26 want to hate him but then he 4:29 sings and it's just 4:35 cute okay because I've been driving for 4:38 the past hour I have nothing in my brain 4:41 so why don't we let you 4:43 all just ask some questions tell me what 4:46 you got going on with AI what are you 4:48 worried about what are you thinking 4:50 about what are you excited about what 4:51 have you done where are you 4:55 stuck and then what I'll do is you put 4:58 those questions and request in and then 5:01 I'll just ignore it and answer what I 5:02 want anyway it's the way it works around 5:08 here is Chad add it's like there's an 5:12 agenda why did it keep doing that 5:15 no why does it keep doing that I don't 5:19 like 5:22 it look at Infinity AI script to video 5:27 all 5:29 right 5:30 Infinity AI script 5:35 video well since I accidentally flipped 5:38 my screen why don't we just go there now 5:40 and 5:41 looked and 5:44 looked the hell is going on with my I'm 5:47 I'm I'm I'm not quite here people I'm 5:51 not quite here I'm a little 5:52 disassociated as they say um 5:56 Infinity AI 5:59 script 6:02 uh to 6:09 video It's A Y 6:13 combinator type a script to get a movie 6:19 out our Tech is a work in 6:26 progress try on Discord got it 6:34 join Infinity AI this I have a feeling 6:37 it's going to be getting 6:40 started go to the general Channel type 6:44 SL AIS script to 6:46 video and then put your prompt in the 6:50 Box well how big are these 6:55 scripts oh I see like talking script to 6:59 video 7:01 AI startups are the future you're 7:03 ignoring Healthcare shath Gary's right 7:06 AI isn't 7:13 everything have you heard of memoral yes 7:16 everyone is talking about it I see 7:21 interesting all right so we'll go slash 7:25 AI script to 7:29 video generate a video from an AI 7:31 generated script 7:34 um let's 7:40 go young 7:43 Tom 7:54 Cruz 7:55 argues 7:57 with Keith Urban 8:02 about the 8:03 laws of 8:08 gravity we'll see what's in the box 8:12 what's in what box what you talking 8:14 about what you talking about 8:16 Willis 8:28 Gary 8:30 oh okay this th this 8:32 is so they've got a handful of they've 8:36 got a handful of uh pre-done 8:39 avatars chamath they got the guys from 8:44 Jason 8:45 cis's podcast 8:48 thing so they got some pre canned 8:50 characters you write a script that then 8:53 picks two of their characters at random 8:55 th this this is something you could 8:57 probably knock out in uh 9:01 with uh zapier and a and a call to hey 9:06 Jen API or did pretty quickly um this 9:12 this is not all that interesting right 9:13 now it's just it's this is just a y 9:16 combinator this is they pitched an idea 9:19 although hang on there is some 9:21 potentially exciting news here so this 9:24 is I'm in Discord that's the rabbit uh 9:28 the rabbit Discord 9:30 Channel and there's there's a new 9:32 announcement so maybe they actually made 9:34 an announcement about 9:36 shipping General is there anything any 9:40 good announcement here any 9:42 announcements no God dang it when are 9:46 you shipping your stupid toy I want my 9:49 toy I'm in batch 9:51 one come on 9:58 people so select and generate refresh 10:02 script select and gener oh oh I got to 10:04 actually say select and 10:06 generate okay all right let's see how 10:10 quick this is done beep popop your video 10:12 generation is project in process it will 10:15 take two to three minutes all right 10:16 we're we're leaving there we we'll come 10:18 back to 10:20 it not ready not ready for nothing they 10:24 announced refunds oh yeah that's cute 10:28 Kevin I'm glad you put the smiley face 10:30 there cuz you know I'd have gone and 10:32 searched for that 10:36 [ __ ] um it does not look like there was 10:39 a there was a 10:43 uh someone on Twitter posted 10:46 that grock had been open source but it 10:49 wasn't it was just an old code 10:52 repository Oh Lordy do you use tea bags 10:57 seems a little personal Kieran 11:01 but yes yes I 11:03 do I I am full on white trash look this 11:07 is my uh my Kyle Shannon dreams mug that 11:11 that Ronda made for 11:18 me I thought it was supposed to ship 11:20 early March I knew it wasn't going to I 11:23 spit out my 11:28 Coco 11:38 oh yeah listen if if I can do nothing 11:42 else on this planet but give people good 11:44 spit 11:45 takes blow cocoa through your nose 11:48 that's that is my that is my one joy in 11:51 life is having comedic timing good 11:56 enough that I can catch people right as 11:58 they're taking a sip so the fact that I 12:00 can do that on a Tik Tok Channel really 12:02 makes me 12:06 happy 12:08 woo please don't make jokes while I'm 12:10 drinking 12:15 yeah 12:22 oh let's 12:24 see I thought that yeah okay that's 12:27 that what's anyone got going on who's 12:29 anybody have any 12:34 questions I had to step away and 12:36 couldn't see the text to vid AI can 12:39 anyone catch me up yeah it was crap it's 12:42 it's just early it's it's a they they've 12:45 got some 12:46 pre-selected avatars you could you could 12:49 knock that out well you could knock it 12:50 out in Discord probably over a weekend 12:53 which is what they 12:55 did and then they put up a a website 12:58 that says log into Discord channel to 13:00 try our our macro that you know makes 13:05 two API calls with some pre prean 13:09 avatars so I I mean I don't find it 13:12 remotely interest it's not far enough 13:14 along it could be interesting at some 13:15 point it's just not right 13:17 now all right here it is here's here's 13:20 the result so my prompt was a young Tom 13:23 Cruz argues with Keith Urban about the 13:25 laws of gravity gravity is 13:28 non-negotiable never argu with Newton 13:30 Gary well no better then it's settled 13:33 fine you win for 13:37 now so those are those th those are two 13:40 of their pre canned 13:42 avatars 13:45 um and then it writes a a dumb little 13:47 script so it is not it is not script to 13:50 film it is in the way script to film it 13:54 is prompt to shitty script 13:58 to pre can 14:01 thingies 14:04 um I feel like Dave 14:07 porty [ __ ] on [ __ ] on technology 14:10 [ __ ] on Pizza he's like ah it's a 2 14:13 point 14:14 2.4 there's people in there trying to 14:16 make their living He just flushed their 14:18 business down the 14:24 toilet oh 14:28 man uh uh uh 14:31 uh all 14:33 right the daily Jank is way better than 14:36 that yeah 14:40 exactly that was that was not software 14:43 that was a couple of prompts being sent 14:46 from a Discord 14:48 robot but they're friends with people at 14:50 y 14:52 combinator do you think we're at a Y2K 14:55 moment year 98 for 14:58 AI if you you want to start an AI 15:01 business that's a good 15:05 question 15:24 um I think the 15:26 only AI business that 15:31 makes sense to start right 15:34 now is a 15:38 well if you if you can raise 15:42 billions go start a company and and do 15:45 like a foundation 15:47 model but if you're someone who's going 15:50 to do what these guys did right here 15:53 where they took did or hey Jen I I 15:58 guarantee you that's did or hey Jen it's 16:00 probably hey 16:02 Jen they're doing two API calls to hey 16:05 Jen they're they're doing a call to open 16:08 AI to generate the script they then send 16:11 they they then randomly choose one of 16:14 their 12 characters that they've got pre 16:17 baked they send it to hen they Stitch it 16:20 they they get the videos back from H Jen 16:22 they stitch them together using FFM Peg 16:25 or some stupid [ __ ] like that and then 16:27 they dump it dump it to a file if you're 16:31 doing [ __ ] like that I don't think it 16:33 makes sense to start a business right 16:35 now because this 16:38 business the 16:40 minute gemini or chat GPT or any [ __ ] 16:46 of them St stability AI the minute any 16:50 of them go full multimodal you'll just 16:53 be able to do that natively within chat 16:55 GPT so until we see what's the 17:02 next what's coming from open AI like 17:05 like it's been a while since we've heard 17:08 anything significant from them so I 17:10 think we're going to hear something big 17:12 I think that the the thing that you 17:14 could start in the next two years fairly 17:17 safely is a service business around AI 17:20 strategy and AI operations AI 17:24 execution so if you're really good with 17:26 ideas and and solving problems take it 17:29 at a higher level if you're really good 17:30 with security and opsc and and and you 17:35 know AI 17:37 operations then start a start a you know 17:40 affirm to do that and I think 17:46 those I think companies for the next 17:48 five years are just going to be 17:50 desperate to figure this stuff 17:54 out but I don't I don't know that I'd 17:56 start a technology company right now I 18:00 just it's 18:02 like I just feel like you're waiting for 18:05 the shoe to fall that the company that 18:08 you built that's revolutionary today is 18:11 going to become a feature inside of the 18:13 new chat 18:14 GPT that's included for 20 bucks a month 18:17 like I don't know how you compete with 18:22 that if you've 18:25 got really interesting proprietary data 18:30 maybe there's something there where 18:32 you're not really relying on the 18:35 latest super high-tech technology but 18:38 but you're doing something with specific 18:39 data that could be interesting you know 18:42 you're doing something with branded 18:43 media content or you're doing something 18:45 with medical data legal data government 18:51 data or maybe not government data 18:54 government data that's not publicly 18:58 accessible 19:01 I don't know that's a good question it's 19:02 a really good 19:04 question it's it's so good a question I 19:08 don't have an answer for it I answered 19:10 it but you'll notice like a good 19:13 politician I didn't actually like I 19:15 don't [ __ ] know I don't 19:18 know no I'm looking at a service-based 19:20 AI business I think right now is the 19:22 time to do it drew I think yeah listen 19:25 you're curious you show up you're 19:27 consistent um if if you've got the 19:29 ambition to do a service business I 19:31 think it's absolutely the time to do 19:33 that and the thing that I would do Drew 19:36 is I 19:39 would you can help people get their AI 19:43 [ __ ] together but your business should 19:46 be fully committed to AI like how you do 19:49 everything you should turn things around 19:52 faster you should do analysis faster you 19:54 should be able to do it cheaper than 19:56 someone else doing something similar 19:59 like really lean into it um how did you 20:02 get the link to the rabbit R1 20:06 Discord I don't know I think from their 20:09 Twitter 20:10 page let me go look let me go look hold 20:15 please 20:17 rabbit we are going to the 20:20 rabbit 20:23 R1 site Mr rabbit 20:27 rabbit oh I put r i it's different than 20:34 R1 rabbit 20:37 Inc rabbit Tech is their web rabbit. 20:41 Tech is their 20:44 website and then if I scroll to the 20:48 bottom is there Discord there hang 20:53 on rabbit o research policy 20:57 press fact 21:04 updates I don't know where there stupid 21:07 Discord is hang 21:10 on 21:14 um rabbit 21:18 R1 21:22 Discord let me Google that for 21:28 you 21:30 I I honestly don't 21:38 know I'll tell you what I'll 21:49 do let's see invite 21:53 people 21:55 copy so I'm going to hop over to the AI 22:05 salon and I think I'll drop this where 22:08 do I want to drop 22:11 this I'll drop it in water 22:18 cooler I'm going to go to 22:22 so if you're not part of the AI 22:27 Salon go 22:30 here the salon. a that's going to take 22:33 you to not that link tree hang 22:37 on 22:39 the salon. a it's going to take you to 22:42 this link tree and click on this top one 22:45 AI Salon 22:48 community and that will drop you in in 22:51 here and the first thing I do want you 22:53 to do if you're new here is go to the 22:55 welcome to the salon section and read 22:57 about the salon read about who we are 22:59 and what we 23:02 do super quick read and then it talks 23:04 about our values like curiosity 23:06 generosity exploration collaboration 23:09 inclusion empathy and gratitude go read 23:12 about all that [ __ ] make sure that 23:13 you're down with what we're up 23:15 to and then over in the community 23:18 section there's a there's a section 23:20 called water cooler and in the live chat 23:23 I'm going to make a post that 23:25 says 23:27 someone 23:30 asked about the 23:34 rabbit R1 23:37 Discord site or 23:41 server 23:43 here's and 23:46 invite 23:49 Bang all 23:51 right win a win a chicken 23:55 dinner you can go watch a bunch of 23:57 people talk about hard where that 23:59 they're speculating on it's it's it's a 24:02 completely useless server I the only 24:04 reason I pay attention to it is for the 24:05 announcements Channel which is painfully 24:09 [Laughter] 24:13 empty we have shared a a general 24:16 Community update in general and it's 24:18 it's just some [ __ ] so 24:21 anyway that's that that's that hello cam 24:25 katkin cam katkin good day to y 24:30 I'm 100% right uh that's what I'm 24:32 providing strategy operations and 24:34 execution there you go no I'm looking at 24:36 a service business yep there you go 24:38 create a real Sora competitor billions 24:41 absolutely agree with that merchandising 24:43 merchandising this is why I keep my 24:46 technology on 24:49 paper now's the only time I can 24:52 think but I wouldn't risk my investors 24:55 dollars for it oh wait now's the only 24:58 time 24:59 it can run I 25:02 think I don't know what you're talking 25:04 about 25:07 there uh there was the question about 25:09 the rabbit 25:16 Discord cam katkin sorry I'm late I was 25:19 late too cim you're later than me that's 25:22 that's bad that's really a 25:24 situation we're going to have to talk to 25:26 the 25:27 Provost 25:30 oh it's exhausting it's exhausting 25:33 running a 25:41 university does he run a university he 25:43 doesn't he doesn't seem like he is he an 25:46 administrator he's he's he's an 25:49 administrator in it but it's called the 25:51 it's it's called the lab it's not he 25:53 doesn't call it a university it's called 25:55 the AI learning lab and it it does not 25:58 doesn't seem like a an institution kind 26:00 of 26:02 name what what are his qualifications 26:08 [Applause] 26:17 anyway you can make money 26:22 with Drew's going to make money with 26:24 chat 26:25 jpt just hang out a shingle I'll teach 26:28 you how to teach how to run chat GPT 26:31 people be like I want to 26:37 know can you just suggest a place to go 26:41 for 26:43 beginners like just learning about AI 26:46 yep here is one keep showing up here 26:49 that's what we do here um and then the 26:52 salon that I just talked 26:55 about the salon. a go to the that URL 26:59 and click on the first link to the 27:02 community and introduce yourself and 27:05 you'll notice that there's a whole 27:06 Community section there's a show and 27:07 tell area where people are showing off 27:09 their work and then there's guilds 27:11 there's a creative Guild a technology 27:13 Guild a business Guild and a social 27:15 impact Guild and there's all sorts of 27:17 different spaces and subcommunities in 27:19 each of those 27:21 guilds and just go start introduce 27:26 yourself and start connecting with 27:27 people right now 27:29 Andrea one of the most important things 27:34 um for getting started is to to be 27:37 around people who are doing this because 27:40 everyone's trying to figure it out right 27:42 now well not everyone everyone who is in 27:45 AI right now is trying to figure it 27:50 out or we teach ai.com that might be 27:53 good 27:55 too that might cost you money but I 27:58 don't know 27:59 someone just posted that but you should 28:02 be in community and and the other thing 28:05 to do with this AI stuff is start with 28:08 chat GPT speaking of higher education my 28:11 my school is cautiously embracing AI for 28:14 students and faculty really that's 28:17 exciting cam thank you thank you 28:19 whatever University you work for or work 28:23 or are near or 28:25 goto I don't know your relationship with 28:28 them but that's cool 28:33 good I 28:37 am relatively apoplectic about our 28:42 education um 28:45 institutions not jumping all over this 28:47 AI 28:49 stuff but that's where we are that's 28:53 where we 28:57 are 29:03 oh yeah I've got a series below if you 29:05 want to start with the series that'll 29:07 teach you some things it's it's it 29:10 doesn't go deep into anything but if if 29:12 you want to understand like why this AI 29:14 is a big deal how these tools actually 29:16 work how to use them and not get 29:18 overwhelmed by them that's that's what's 29:20 in the uh in the series called 29:22 underhyped 29:26 um but just start just start 29:29 and in fact let me show you something do 29:30 looks like someone introduced themselves 29:32 great first thing you're going to do 29:33 check out the you know what we're about 29:36 then go to introduce yourself and 29:38 introduce yourself use your real name by 29:40 the way use your real 29:41 name um who whoever BS 29:46 is it's a funny name uh but it one of 29:50 the things that that we encourage and 29:52 are requesting is people use their real 29:53 name so you can actually get to know 29:55 people um and then below below introduce 29:59 yourself there's a section called AI 30:01 Launchpad and tools and this list where 30:04 do I start quick start and Tool 30:07 recommendations this is a page that I 30:09 curate and so these first three 30:12 instructions play solve and share are my 30:17 recipe for how you get AI literate and 30:20 the first one is literally play just go 30:22 play with as many tools as you can don't 30:25 act like you have to learn it like you 30:26 have to learn a programming language you 30:28 don't have to be nearly as technical as 30:30 you think you do with these tools go to 30:33 sunno Ai and write a 30:37 song you're like but I'm not very 30:39 musical no you just tell it to write a 30:41 song and it will write a song go to chat 30:44 jpt and have it write you a a recipe for 30:48 I don't know a tofu based lasagna and it 30:52 will do that just go play play play play 30:55 play play play play then the solve is as 30:58 you start to understand what the tools 31:00 are and what they can do by playing then 31:02 you can start to solve problems and and 31:05 solve them for other people a friend 31:08 who's starting a business or something 31:10 at work or something for your community 31:13 group and maybe they need to come up 31:16 with a marketing plan hey would you mind 31:17 if I do that for you for free using AI 31:19 I'm I'm trying to learn how to use it 31:21 yeah whatever and then you know 2 hours 31:23 later you give them a complete marketing 31:25 plan that they thought they were going 31:27 to have to spend $10,000 for how'd you 31:30 do that oh I just used AI right and then 31:33 the third one is share talk about it 31:35 talk about it on LinkedIn talk about it 31:37 with your family talk about it with your 31:40 friends mle you're going everybody hates 31:43 it when I talk about AI [ __ ] them talk 31:46 about it anyway put yourself out there 31:48 as someone who's learning about this 31:50 stuff because if you're actively 31:52 learning about AI for the next year you 31:57 are an expert 31:58 period end of 32:01 story it's that simple if you're curious 32:04 and engaged and learning about this 32:06 stuff you know as much as most people on 32:09 more more than most people on the planet 32:12 and as much as a lot of the people that 32:14 call themselves 32:18 experts and then the below that is just 32:21 a big giant ass list of tools to go play 32:24 with all of the tools that I have on 32:26 this list are tools that won't waste 32:27 your time 32:29 the very first tool on the list is chat 32:32 GPT if you are not chat GPT literate and 32:36 I know it's [ __ ] ridiculous every 32:39 everyone talks about chat gbt what do 32:41 about chat gbt chat gbt chat gbt you can 32:44 make money 32:47 with it's just better it's just better 32:50 than the other stuff right now now 32:51 that's changing 32:53 quickly there are now 32:57 four 32:59 things that are in the neighborhood of 33:00 chat GPT there's Claude 3 33:03 Opus there's Gemini 1.5 Pro both of both 33:07 of those have a little bit of limited 33:08 access cuz you know the one's behind a 33:10 pay wall and the Gemini one is behind a 33:14 you have to be you have to apply for it 33:17 basically and then there's PI from 33:20 inflection and those three are competing 33:22 with chbt but just start with cha chbt 33:25 pay the 20 bucks a month if you can at 33:28 all afford it the 20 bucks a month for 33:30 chat GPT plus gives you internet access 33:34 it gives you this thing called code 33:36 interpreter which is Advanced Data 33:37 analysis which sounds really geeky and 33:39 awful but it's actually quite remarkable 33:43 once you understand what it can do it's 33:46 got GPT Vision you can upload images 33:49 like take a picture of a whiteboard and 33:55 it'll parse it it'll it'll understand it 33:58 it can generate images and then you can 34:00 create and share custom gpts which are 34:03 fullon customized versions of chat GPT 34:06 that you can share with the world that 34:08 you 34:09 publish um you can't really make money 34:11 with them right now but that should be 34:14 coming but anyway that's where to start 34:17 all 34:18 right and then just keep coming back 34:20 here just keep coming back just keep 34:22 coming back just keep coming 34:24 back but I haven't learned anything yet 34:27 tonight it just seems like you're 34:28 talking to the camera that's how it goes 34:31 sometimes this is Chad 34:32 add sometimes you come in here and it's 34:35 like one bomb after another like boom 34:37 boom boom boom you're like oh yeah bring 34:39 it on Kool-Aid other nights you're like 34:42 is Kyle 34:47 drunk I'm part of our group from our 34:49 faculty Senate writing best practices 34:51 for AI use for the coming year that's 34:53 awesome cam um cam go check out um if 34:57 you haven't seen it yet go look for the 35:00 city of Boston's um generative AI policy 35:05 I think it's 35:06 called let me see if I can find 35:09 [Music] 35:11 it city of 35:17 Boston Boston 35:19 AI 35:25 guidelines yeah 35:28 city of Boston AI guidelines it'll take 35:30 you to boston.gov 35:32 interim guidelines for using generative 35:35 AI they did this on May 35:39 18th how do you make your own GPT all 35:42 right we'll go we'll go make one I will 35:44 show you how 35:47 um this is just a really good document 35:50 cam 35:51 um that just talks 35:53 about what they are that they 35:56 hallucinate how they work and it it was 35:59 this is Boston was the first city that I 36:01 know of that put out an official you 36:04 know for the city of Boston here's our 36:07 position on chat GPT I think it's a 36:09 really good um starting point so if if 36:13 you're if you're being asked to put 36:15 together guidelines it's a 36:17 it's pop it into chat GPT and say change 36:21 this for my University and update it 36:23 with all the latest thinking um you 36:26 could probably get to a pretty good 36:27 first 36:31 dra all right let's 36:35 go let's go look at that city of Boston 36:41 guidelines it's very very 36:44 cool all right we are going to go to 36:47 chat 36:48 Jay we're going to teach people how to 36:51 make a 36:52 JPay 36:54 GPT no chat. open 37:00 okay so if you don't know what a GPT 37:07 is in order to use one you have to be 37:10 part of you have to have a chat GPT Plus 37:14 subscription you will hear people talk 37:16 about gp4 chat 37:20 gp4 and occasionally someone will talk 37:23 about chat GPT 37:26 plus they're all the same well they're 37:28 not all the same thing but they're kind 37:30 of synonymous you can't get 37:32 to gp4 without a subscription basically 37:37 um so how you get to that is you go to 37:39 your name in the lower leftand corner 37:41 you go to my plan and then the Plus 37:44 Account is the one for how you get to 37:46 chat GPT um 4 37:50 okay once you're in there you now have 37:53 these things called gpts and and I can 37:55 click on this one called explore GPT 37:59 and have they changed anything here not 38:03 yet no and then I can search for them so 38:05 I'll show you my very first one I made I 38:08 made a a GPT called Sin City 38:12 sipper and this is it and I don't think 38:15 it's got any reviews yet no oh look oh 38:18 my my tarology deck more by Kyle Shannon 38:22 so I also created a a deck called 38:25 tarology which combines astrology with 38:28 the Tarot deck and turned it into a a a 38:31 reading GPT it's pretty cool tarot 38:34 divination and astrological wisdom it's 38:37 pretty fun okay so if I go to Sin City 38:42 sipper you'll notice that there's some 38:44 pre-written what they call conversation 38:46 starters down here so show me a new 38:49 fangled oldfashioned what Sin City 38:51 sipper is is it takes classic cocktail 38:54 recipes and gives them a Vegas twist 38:56 this was the very first GPT I created it 38:59 took me about 10 minutes I did it on one 39:01 of these 39:02 lives 39:04 and I just got access to to G to this 39:09 feature while I was live and so we just 39:11 did it live so this the very first one I 39:13 made and so what it does is I can I can 39:18 type in a thing like I'm having a party 39:20 what's oh we're we're near St Patty's 39:22 days are we on St Patty's days when is 39:24 St Patty's days um I want 39:29 a super 39:32 interesting 39:35 whiskey 39:37 based St 39:40 Patrick's Day 39:46 cocktail and what Sin City sipper will 39:49 do is give me a unique 39:54 custom uh recipe what it's actually 39:56 doing right now is it's creating the 39:58 image of it first um I told it not to do 40:05 that what does the a in AI mean 40:10 artificial and the I means 40:19 intelligence I don't really know what to 40:21 say about 40:23 that it's just basically these machines 40:26 have been modeled after the way our 40:28 brains work and they're trying to make 40:30 it happen it's it it it would be more 40:32 accurately called machine learning um 40:35 all the stuff that we're playing with 40:36 but all right so what did it give us 40:39 your St Patrick's Day celebration with 40:41 the Emerald 40:43 Isle whiskey 40:45 twist Irish whiskey chartreuse 40:48 interesting lime juice simple syrup 40:51 Bitters and a mint sprig that doesn't 40:54 quite look look like what it would look 40:56 like with those ingredients but that's 40:59 okay um but anyway so that's what this 41:02 thing does it's just a it's just a 41:04 single purpose you know thing and the 41:08 the way I've got it set up if you ask it 41:09 to do things other than this it should 41:11 tell you it can't do that um but but for 41:16 example somebody sent me after the Super 41:20 Bowl they sent me a 41:26 um oh 41:32 okay I I I couldn't figure out why you 41:35 were asking it Pate I I should have 41:37 known you were just being a smart ass 41:39 but thank you for being a smart 41:42 ass I'm like why is p asking that there 41:45 must be some maybe some paper came out 41:48 today that they're they're changing the 41:50 the a to mean something else now or it 41:53 was a joke so I'm glad it was a 41:56 joke 41:58 um somebody used Sin City sipper they 42:01 took a picture of their booze cabinet of 42:04 of like the bottles in their booze 42:06 cabinet and they said make 50 drink 42:09 recipes Super Bowl themed drink recipes 42:13 for my guests for a Super Bowl party 42:16 using the booze that's in my booze 42:18 cabinet and it did it and it gave them 42:21 like a a table of of drink names that 42:23 they turned into um a menu and then 42:26 people at the Super Bowl party could 42:28 come up and ask for a drink and then the 42:30 bartender had all the recipes for all 50 42:32 drinks that were just of the booze that 42:34 was was available really amazing right 42:37 so so anything that you're interested in 42:40 can become a 42:42 GPT um oh I got one okay so we're going 42:45 to go make one now so how you find them 42:48 is you go to explore 42:52 gpts and you can just scroll through a 42:54 big list of them you can search for them 42:56 you can search on topic it's not it's 42:58 not a great directory right now but if 43:01 you want to make a new one you go in the 43:04 upper right hand corner you say 43:05 create and then it takes you to this 43:08 page with a lot of fingerprints on the 43:10 screen 43:11 see these are these are all where I 43:15 touch I'm a very tactile teacher 43:18 apparently um on the right hand side of 43:20 the screen you have this preview this is 43:22 where when you create your GPT you're 43:24 going to be able to test it on the left 43:26 hand side of the screen is where you 43:28 configure it or where you create it and 43:30 so there's there's two tabs create and 43:33 configure we're going to start with 43:34 create because configure while it's very 43:38 very simple and it doesn't actually 43:40 require any coding it's a little more 43:42 complicated like you can upload a 43:44 picture you have to come up with a name 43:46 a 43:47 description instructions which is 43:49 actually your prompt right there's not a 43:50 lot of uh instructions on what to do 43:53 with the instructions uh you can search 43:56 for AI salon and see a bunch of gpts by 43:59 folks in this group too oh cool yeah 44:01 search for hash AIS Salon I I need to go 44:04 hashtag all my my gpts up that's a great 44:07 idea Corey um so just don't worry about 44:10 configure right now we'll come back to 44:11 that but we're going to go to 44:13 create and what we're going to 44:17 do is we're going to 44:19 say I want to create a 44:25 GPT for the 44:29 um Dune 44:35 movies that will let someone 44:40 learn about 44:42 them geek 44:45 out and even play trivia 44:54 games and so it says updating GPT so 44:58 it's building it so if you notice on the 45:00 right here it just still says 45:02 preview but here in a few 45:05 seconds it 45:15 won't a guide to trivia a guide and 45:17 trivia partner for the Dune 45:19 movies um tell me about the Dune 45:23 universe so if I click on that it it re 45:26 pre-wrote some little conversation 45:29 starter 45:31 questions created in 1965 by Frank 45:34 Herbert was the book at the core the 45:36 Dune Saga right so it's telling us 45:38 little things there and then we say how 45:41 about calling it Dune Explorer and I'm 45:43 going to say okay 45:46 fine and then it should yeah updating 45:50 GPT now it says generating a profile 45:53 picture so it's going to create a a 45:55 picture for our 45:59 our GPT 46:01 here all right that's lovely but I don't 46:05 like that how about a 46:12 sandworm generating profile 46:26 picture 46:28 I love that you're all just chatting 46:30 with each other down in here it's pretty 46:32 cool all right there's our sandworm 46:34 doesn't at all look like a sand worm and 46:37 dune 46:38 um it needs to look more 46:46 menacing 46:49 it's 46:51 it 46:53 can eat 46:56 you 46:59 [Laughter] 47:03 did you see Neil degrass Tyson talking 47:06 about um the whole premise of Dune is 47:09 all [ __ ] up because if you pound on 47:11 Sand the the one thing that s uh sand 47:15 just simply doesn't do is transmit sound 47:18 it's like if there you go there's a good 47:20 scary sandworm there we go nice so it go 47:26 perfect 47:28 I can't believe it did something that 47:30 aggressive which is good that's that's 47:32 what we 47:34 want all right let's dive in a bit 47:38 deeper but anyway so we now have this 47:41 thing called Dune 47:43 Explorer a guide for trivia and then 47:45 it's got these little starter questions 47:47 I can now go over to the configure Tab 47:51 and you'll notice that all of this stuff 47:54 is filled out for us right so here's the 47:56 description 47:58 here's the here here's the 48:00 prompt um you can you can add to this 48:03 prompt so I could say 48:07 um follow these 48:13 rules 48:16 whenever you give 48:21 someone a 48:25 description of 48:28 part of the story or 48:32 movie create a 16 48:37 by9 48:41 um storyboard 48:48 frame that brings that 48:51 scene to 48:56 life 48:57 all right so now when I come over 49:02 here share a fun fact about Dune I'll 49:05 click on that 49:10 one the iconic sandworms known as 49:13 the is the shy huded by the Freeman play 49:17 a crucial 49:19 role something about the spice the life 49:23 cycle of the sandworm it's giving us all 49:25 sorts of stuff the did it give 49:29 us 49:31 um when do we first see 49:39 them 49:42 in the 49:47 movie now it's creating an image so 49:50 we'll see what it we'll see what it 49:56 creates 50:05 and then I'll show you how to share 50:07 these live with the 50:11 world 50:13 nice there's our little scene from our 50:17 movie and then if you want to save it in 50:19 the upper right hand corner where you 50:21 clicked create before you say save and 50:23 then I can save it for just myself or I 50:25 can share it anyone who has the link can 50:27 play with it or if I click 50:30 everyone that will then save it to 50:34 the 50:36 store and so if you go right now to the 50:40 to explore 50:43 gpts and you type in Dune 50:52 Explorer you should be able to see my 50:56 what we just created here yeah do an 50:58 Explorer oh I'm not the only 51:04 one mine's the only one that has a nasty 51:08 worm everyone else got lazy theirs just 51:12 got the shitty the shitty scenes mine's 51:14 got the nasty worm 51:18 so that's how you make 51:20 gpts that's it it's simple you can make 51:24 money 51:25 with you can also when you go to that 51:28 configure page one of the other things 51:30 that you can 51:31 do is there's an upload button where it 51:35 talks about knowledge you can upload 51:37 knowledge to your GPT what's that 51:41 mean let's 51:43 say you have Grandma's recipe 51:46 book in a 51:49 PDF you can upload Grandma's recipe book 51:52 and tell your GPT to only use that PDF 51:55 for its source of 51:57 and you could make like a Grandma's 51:59 recipe GPT where you just ask it for 52:01 things and it'll pull her recipes up out 52:04 of the out of the PDF it's really it's 52:06 really slick very very sophisticated 52:09 with for like zero programming 52:12 requirement zero zero not like zappier 52:17 which claims to be a no code development 52:19 tool but it really requires you to 52:22 understand apis and programming logic 52:25 and um it's it's not as simple as as it 52:29 sounds which it can frustrate people 52:32 because they're like I thought this was 52:34 supposed to be easy well it's easy 52:36 relative 52:38 to learning to code but it's not easy 52:42 gpts are actually easy can chat GPT help 52:45 you make spreadsheets yes well make 52:49 spreadsheets it can it can analyze 52:51 spreadsheets it can make spreadsheets 52:55 so in in gp4 is a thing called code 53:03 interpreter so I can just 53:06 say 53:08 um 53:11 make me a table that 53:17 compares 53:19 um 53:22 movies 53:25 from each 53:31 decade 53:37 um 53:40 starting with the 53:46 70s and it will do that really 53:55 well 53:58 oh it's interesting what it did is it is 54:00 it picked 54:02 uh it picked the most iconic film of 54:05 each decade that's kind of cool I hadn't 54:08 even thought about that I mean that 54:10 wasn't what I was asking for but 70s The 54:13 Godfather 80s ET '90s 54:16 Titanic 2000s Lord of the Rings 2010s 54:21 Inception 2020s parasite that's 54:25 interesting but now I can just say 54:31 um 54:36 put this uh table in 54:42 an 54:45 Excel file I can 54:50 download and see where it says analyzing 54:53 here and if you open this 54:55 up what you actually see is that it's 54:59 writing and executing its own python 55:02 code so you can either upload a CSV file 55:06 full of data and have it analyze that 55:08 data or you can take anything that you 55:10 produce and you can have it generate a 55:12 Word document or an Excel document or a 55:15 CSV whatever you 55:17 need 55:20 um and so it just it does its 55:24 thing oh always expand out out put They 55:27 just added this they just added a toggle 55:29 thing here that or I haven't seen it 55:31 before where when it does what it does 55:35 it'll default to always expanding it I'm 55:37 going to I'm going to leave that on for 55:38 now and see if see if it bugs 55:40 me but the table has been saved to an 55:44 Excel file you can download it here so 55:47 the movies by decade Excel file if I 55:50 click on that it downloaded it and I go 55:53 over here 55:55 to my little 55:58 finder and I can just preview 56:02 that boom there it 56:04 is it's not formatted very pretty but 56:07 you can you can tell it to format it I 56:09 mean it's it's quite quite remarkable 56:11 this old chat JP Ty 56:22 thing you you all realized that Mr K had 56:25 chat GPT tell him about Dune one while 56:27 going to see Dune 2 well so here watch 56:30 watch how we can do this 56:34 so 56:35 iPad this is the official chat GPT app 56:38 so if you don't have the official chat 56:40 GPT app you should go get it it's 56:43 free be really careful when you go get 56:45 the chat GPT 56:47 app both Apple and Android stores are 56:51 full of Faker 56:54 apps make sure you get the official 56:56 official app from open 56:58 AI okay 57:01 okay um and then when you're in 57:04 here you can click on these little 57:06 headphones and you can talk to 57:11 it hey I'm heading to see the movie Dune 57:14 2 and I'm hoping you can catch me up on 57:17 the plot of Dune one cuz I can't 57:19 remember any of 57:24 it of course I'd be happy to help you 57:26 catch 57:33 up feudal houses overseen by the Pisha 57:36 Emperor the story revolves around the 57:39 desert Planet arus also known as dune 57:42 this planet is the only source of the 57:44 universe's most valuable 57:50 substance sorry you got cut off could 57:52 you start 57:55 over 58:03 hello can you start 58:06 over planets under an 58:09 emperor all right something says 58:12 connection failed tap to retry I think 58:14 my I think my uh chat GPT is has got bad 58:18 memory let's try 58:24 again hey I'm headed to Dune 2 and I 58:26 can't remember what happened in Dune one 58:28 can you catch me 58:35 up absolutely heading to watch Dune part 58:38 two without a refresher on Dune part one 58:41 might leave you a bit lost so here's a 58:43 quick catch up Dune part one focuses on 58:47 the desert planet of arus so anyway you 58:50 can do that yeah I know U Mr K I don't 58:52 know if you heard it but but Joker Joker 58:55 called you out and said you know Mr Mr K 58:57 did this cuz I just made a uh I just 58:58 made a dune GPT I don't know how much 59:00 you've heard um no that's a super cool 59:03 idea but did you know you could talk to 59:06 Chachi 59:07 PT if not you can this is why you need 59:11 to get curious get curious get curious 59:13 get curious just just play with these 59:15 tools and hang out with people that are 59:17 playing with these tools because they'll 59:19 figure [ __ ] out and tell you about it 59:21 you'll be like oh that's cool like Mr K 59:24 is a teacher an educator he's one of the 59:26 co-host of the education Guild in the AI 59:29 Salon um and he's super generous he's 59:32 got a Tik Tok Channel you should follow 59:34 his Tik Tok Channel Mr K all day is is 59:37 his uh thing did you reach a thousand 59:40 limit on Mid journey I don't think so 59:42 let me just quickly go look I didn't I 59:44 didn't play with it today I actually 59:46 just took some time and did nothing 59:49 which was really good I downloaded a 59:51 couple of Tik Tok 59:53 videos 59:55 um let me see 59:58 alpha. mid journey.com 1:00:03 sign 1:00:06 in let's see if I got it sign 1:00:11 in no apparently not because I think I'm 1:00:14 signed in am I signed in what's going on 1:00:25 here 1:00:41 no I don't have it 1:00:44 [Laughter] 1:00:47 yet I'm probably 208 right now hang on I 1:00:50 just heard champy needs something hang 1:00:52 on I'll be right 1:00:55 back 1:01:25 B 1:01:52 all right champy champy wanted a snack 1:01:54 champy has been dealt with oh 1:01:57 oh 1:02:03 oh I came back just in 1:02:06 time it was it made me it made me solve 1:02:09 a puzzle piece to keep going to keep 1:02:11 staying live um it's kicked me off for 1:02:13 that 1:02:16 before okay tea has been drunk now I'm 1:02:19 on water I'm normally double fisting I 1:02:22 normally switch back and forth not 1:02:25 tonight 1:02:27 all 1:02:28 right there's a fake St Sora app now on 1:02:31 the iOS store I know and they've got T 1:02:33 they've got Tik Tock ads up you can now 1:02:36 use Sora in an app it's not 1:02:38 true it's really annoying but you 1:02:42 know you mentioned that zappier is 1:02:45 simple program and can chat GPT make 1:02:48 zaps yes the zapier has 1:02:53 incorporated uh chat gbt 1:02:57 into zapier So within zapier you can 1:03:01 just ask it to build you a GPT and it 1:03:03 will or not a GPT you can build a zap 1:03:07 you can say I want to create a zap that 1:03:10 um every time I receive an email from 1:03:12 this email address it you know comes up 1:03:15 with an answer or something like that um 1:03:17 and it will tell you which components to 1:03:20 put together and then you have to go 1:03:21 hook them all up or you know like log 1:03:24 into them all and give them their 1:03:25 instructions so um so it's in there 1:03:28 right now um there's a plugin in chat 1:03:32 GPT if you've got chat GPT plus you have 1:03:35 access to plugins but plugins are going 1:03:38 away um and then there are a lot of 1:03:41 zappier I think zapier has an official 1:03:44 GPT in fact let me go look at that I 1:03:48 think that's the case explore 1:03:51 gpts 1:03:55 zappier 1:03:59 automation consultant by 1:04:06 zappier it says it's by zapier but it 1:04:09 says by Community Builder so I don't 1:04:11 know if this is an official zappier 1:04:14 thing discover opportunities and save 1:04:17 time with automation at work and get 1:04:19 them set up for you yeah so go go grab 1:04:21 the uh the zapier um GPT 1:04:27 and see if it's any good it might be who 1:04:36 knows did you know that chat GPT will do 1:04:38 standup if you ask it to yeah 1:04:42 unfortunately um The Comedy the comedy 1:04:45 stylings of these llms right 1:04:49 now like mostly don't even understand 1:04:53 joke structure gp4 does 1:04:57 um but when it writes the jokes they're 1:04:59 kind of dad jokes at best they're not 1:05:02 great 1:05:04 so if if someone can figure out a way to 1:05:06 prompt it maybe it's maybe if you used 1:05:09 some some data sets of actual comedy 1:05:11 writing uh it would get better if you 1:05:13 fine- tuned it a bit you could probably 1:05:15 make a GPT that wrote jokes pretty good 1:05:18 but it sucks right 1:05:20 now but it will do standup it'll just be 1:05:25 bad um you can load custom instructions 1:05:28 and run respond to it as if it's champy 1:05:31 the singing AI literate dog that's good 1:05:33 I like that 1:05:37 [Music] 1:05:39 uh chat GPT needs a humor plugin it 1:05:43 does that could be a way I'll tell you 1:05:46 what that could actually be a way that 1:05:49 um a a good comedy writer especially a 1:05:53 famous 1:05:55 One 1:05:57 if they got if they got good at 1:05:59 this you know take any of any of the SNL 1:06:03 writers like Seth Meyers or or or 1:06:06 whoever just any you know any of them um 1:06:09 or just anyone that can write write 1:06:12 good if they got AI literate they could 1:06:16 make a GPT that is designed to write 1:06:20 good jokes and put that out in the world 1:06:22 and potentially you know make money on 1:06:24 it um 1:06:28 [Music] 1:06:36 my Joker bot tells pretty good 1:06:39 jokes yeah what did did you upload any 1:06:45 um did you upload any um documents to It 1:06:49 joker or or was it just with 1:06:52 prompting I was trying to get GPT to 1:06:54 write stories like Trey Parker and Matt 1:06:57 Stone oh that's 1:06:59 interesting I would think that you could 1:07:01 I I would think that you could take like 1:07:03 South Park scripts Mr K dump a bunch of 1:07:06 South Park scripts in a uh in a series 1:07:10 of 1:07:12 PDFs and just upload them as the 1:07:14 knowledge Source I bet you could I bet 1:07:16 you could get it to write South Park 1:07:17 episodes pretty good there's so much 1:07:20 there's so much out there and that 1:07:23 instruction wait and that instruction to 1:07:26 Dune to show it 1:07:29 Champion wait what what Corey oh add 1:07:32 that instruction to Dune to show it 1:07:35 champy will talk and it'll be 1:07:40 sweet um yeah they they had interview 1:07:42 transcripts too oh yeah that's cool 1:07:45 that's pretty cool yes a PDF okay great 1:07:49 yeah yeah I would think if you if you 1:07:52 uh if you find a good comedy writing s 1:08:00 course hm that's 1:08:03 interesting you get into such weird copy 1:08:06 copyright weirdness with with 1:08:10 gpts cuz you could put a book in 1:08:13 there as the source but not credit it 1:08:18 and then that's [ __ ] up but if you 1:08:20 credit it you can't use it cuz it's 1:08:23 copyright infringement we're in such 1:08:25 weird territory right now I will add 1:08:28 this plugin 1:08:32 great um what else do we want to do I 1:08:34 think guard rail 1:08:36 should archetype okay let's see I was 1:08:39 main looking 1:08:40 at before 1:08:43 storytelling it violates policy quite 1:08:46 often thoughts on GPT rapper 1:08:54 startups I don't know GPT rapper 1:08:57 startups are like startups in the mid 1:09:01 90s um anyone could do 1:09:05 it there're they they are super sexy on 1:09:09 the surface and have very little 1:09:11 underneath the 1:09:13 surface 1:09:17 and um I think a lot of them are going 1:09:20 to get funded I think almost none of 1:09:22 them are going to 1:09:24 survive 1:09:27 I also don't know where the boundary 1:09:30 is between a a chat GPT rapper app and a 1:09:34 real 1:09:35 application so if you look at 1:09:38 Devon what Devon's doing is a lot of 1:09:42 calls to 1:09:44 gp4 so it's technically a rapper app but 1:09:48 they've got all these different modules 1:09:50 that 1:09:51 are taking the output of chat gp4 and 1:09:54 using them right so so Devon is a much 1:09:57 more 1:09:59 sophisticated GPT rapper 1:10:02 app and by the if you don't know what a 1:10:04 GPT rapper is by the way it basically is 1:10:07 you create a website that looks like it 1:10:09 does something remarkable just like a 1:10:11 GPT you know I could I could I could 1:10:14 make a uh I could create a startup for 1:10:16 doing um customized cocktail mixing 1:10:20 application for bartending point of sale 1:10:23 systems right and and I could put it out 1:10:27 there and you know I wouldn't call it 1:10:29 Sin City sipper but I'd call it like you 1:10:31 know 1:10:33 bartenders uh you know 1:10:36 uh Guru or whatever and I could probably 1:10:42 if I tried really hard because that's 1:10:44 not a great idea but I could potentially 1:10:46 go raise money on that right and try to 1:10:48 get it sell it into the bar bar industry 1:10:51 and hook it into point of s systems and 1:10:53 things like that uh 1:10:56 but in the end it's just making calls to 1:10:58 chat GPT so I don't 1:11:03 know I don't know I don't I do not have 1:11:06 a 1:11:13 good I don't have good advice like if 1:11:16 you if you want to put together a rapper 1:11:18 app and try to go raise money on it 1:11:20 great but what I would say 1:11:23 is just 1:11:29 here here's here's here's where I think 1:11:31 we're 1:11:33 going if you go look at the the videos 1:11:36 of Devon and if if you haven't seen 1:11:38 those I can I can go show show what 1:11:40 Devon 1:11:41 is Devon is an autonomous 1:11:45 agent and so what Devon does is you give 1:11:48 it a goal like right 1:11:51 now the way these large language models 1:11:54 work is you you give them a prompt and 1:11:56 they give you back an answer and then 1:11:58 you give them another prompt and they 1:11:59 give you back an answer and it's on you 1:12:01 the human to figure out if it's right or 1:12:04 not right and it's and to figure out 1:12:06 what's good and to figure out how to 1:12:07 prompt 1:12:08 it where we're going and I think 1:12:11 whatever open AI is going to release 1:12:14 next is going to be markedly different 1:12:17 than the current chat GP I mean I think 1:12:19 it's going to be markedly different I 1:12:21 think it's going to have reasoning and I 1:12:23 think it's going to have autonomous 1:12:25 agents 1:12:26 and autonomous agents you give them a 1:12:29 goal and they just go start doing 1:12:31 [ __ ] and because they've got reasoning 1:12:34 they'll go is the [ __ ] that I did good 1:12:37 or not does it have mistakes in it or 1:12:40 not and and it's they're going to go off 1:12:42 and and what's going to happen 1:12:45 is what the experience is going to 1:12:48 effectively be is that we'll be 1:12:50 interacting with custom written software 1:12:55 that was written for us in the moment 1:12:58 based on our 1:12:59 goal and so I don't know how I I don't 1:13:03 know how as a software developer you 1:13:06 compete with 1:13:13 that and and are are those autonomous 1:13:16 agents going to be good in 1:13:18 2024 or are they going to be good in 1:13:22 2027 cuz if it's 2027 then you've got 3 1:13:27 or 4 years to maybe get a rapper app 1:13:31 funded and go put it out in the world 1:13:35 and then use that money and what you 1:13:38 learn to create something that's 1:13:40 useful but if those autonomous agents 1:13:43 get good in 2024 or early 1:13:47 2025 I feel like you're going to spend a 1:13:49 lot of time building software that's 1:13:51 just going to be no one's going to give 1:13:53 a [ __ ] about 1:13:54 it so 1:13:56 I don't know I don't know man we're in a 1:13:59 weird place right now we're we're in a 1:14:01 we're in a cusp in between 1:14:05 two two modes like like 1:14:09 we're we're on the tail end 1:14:16 of generative AI as we've come to know 1:14:19 it which is you prompt it and it gives 1:14:21 you an answer and we're about to enter a 1:14:25 mode where that's not really how they 1:14:28 work you'll still be able to do that but 1:14:31 but the the more advanced stuff is going 1:14:33 to be all this autonomous stuff with 1:14:36 reasoning and I think everything changes 1:14:38 then so that's why I'm like why I am 1:14:43 more inarticulate than normal is 1:14:46 that I'm trying to figure out what's 1:14:49 coming not knowing what's coming but I I 1:14:52 know I there's enough hints to know that 1:14:55 it ain't going to be what we have now 1:14:57 wait what was that Kyle uh uh a human 1:15:00 being had something up there that was 1:15:02 just pinned I just I was just about to 1:15:03 read it and I lost it it went 1:15:15 [Music] 1:15:16 away test some of the Irregulars gpts I 1:15:20 could do that I could do that 1:15:22 um the problem with testing some of the 1:15:25 gpts is without knowing them and knowing 1:15:30 which ones are good and what to do on 1:15:31 them it could I my fear is it'll just 1:15:34 get a little 1:15:36 boring so so I will do that but I'm not 1:15:39 going to do it tonight because I don't 1:15:41 want to just sort of flounder what I'll 1:15:43 do is I'll go play with some and find 1:15:45 some to show it's a good idea 1:15:49 though Diane what prompt instruction 1:15:52 could my daughter use to get chat GPT to 1:15:54 help her study for ology all right let's 1:15:56 go do that that's that's 1:15:59 GPT help 1:16:01 you so oh actually so the first thing 1:16:03 we'll do I I'm I'm going to assume 1:16:05 you've got chat GPT plus um I'm going to 1:16:09 say 1:16:10 biology 1:16:12 biology I'll just put in biology right 1:16:15 now let's see what it comes up 1:16:16 with 1:16:20 specialized your biology expert biology 1:16:24 professor 1:16:26 expert biology let's 1:16:28 take number one specialized biology 1:16:32 assistant ah the the the world's most 1:16:37 powerful biology tutor okay go go to 1:16:42 gpts and jump 1:16:45 into all right it's got decent ratings 1:16:48 how many ratings 40 plus 4.5 Stars 1:16:53 education start chat 1:16:56 Advanced biology mode oh this is cool 1:16:59 basic biology mode Advanced biology mode 1:17:01 exam preparation mode and research mode 1:17:05 so I'm going to go to Advanced biology 1:17:09 mode welcome to Advanced bi biology 1:17:24 mode 1:17:30 that's 1:17:37 interesting as a token of appreciation 1:17:39 for selecting a member of the Pulse R AI 1:17:44 Fleet I invite you to join a VIP 1:17:47 community of AI enthusiasts and 1:17:54 creators so this is basically just a 1:17:56 [ __ ] ad for these things just annoy 1:17:59 me um it's this is just an ad for 1:18:02 something like the AI 1:18:04 Salon but it's but it's not it's it's 1:18:07 really 1:18:08 unclear like like I don't know as a 1:18:12 token of appreciation for selecting a 1:18:15 member of the Pulsar AI 1:18:17 Fleet I didn't select a member of the 1:18:20 Pulsar AI Fleet I selected a biology 1:18:22 thing like that's this is just this is 1:18:24 just lazy programming that this comes up 1:18:27 this early and that they don't explain 1:18:29 what the [ __ ] a member of the Pulsar AI 1:18:32 Fleet is um anyway let's see if this is 1:18:34 any good 1:18:35 um let's see I I 1:18:41 need um flash cards to 1:18:47 understand um cell 1:18:52 mitosis oh and I spelled it wrong 1:18:57 and it fixed it so that's 1:19:05 good all right so this is good so here's 1:19:07 our flash 1:19:09 cards and actually you know what what 1:19:12 would be cool with this so wait I'm just 1:19:14 going to stop it I'm going to 1:19:17 say great put these flash cards in a 1:19:24 table 1:19:30 [Music] 1:20:29 and create an interactive 1:20:32 experience that will help my daughter 1:20:35 assess where she 1:20:37 is in learning and give her 1:20:45 exercises to improve as she prepares for 1:20:49 some big 1:20:54 tests 1:20:57 so that sounds like a great plan cell 1:21:00 biology we can 1:21:05 assess let's conduct a brief assessment 1:21:08 cell one cell biology what's the F 1:21:10 function of okay so it's just giv me a 1:21:12 big blast of questions so I'm going to 1:21:14 stop 1:21:16 it and I'm going to 1:21:18 say I want this to be an interactive 1:21:21 experience so I want you to ask her 1:21:24 questions 1:21:25 as if you're speaking to her directly 1:21:28 and then I want you to pause after each 1:21:31 question letting her answer it before 1:21:33 you move on to the next question 1:21:38 period 1:21:41 Absolutely I'll pose the questions one 1:21:43 at a 1:21:45 time take your time and think about it 1:21:47 uh can you tell me what the 1:21:50 mitochondria's role is within the 1:21:54 cell 1:21:58 the 1:21:59 mitochondria 1:22:01 wait the mitochondria within a cell is 1:22:04 like a little disco dancer at the 1:22:08 [Laughter] 1:22:11 club that's a creative and interesting 1:22:16 analogy often known as the PowerHouse of 1:22:18 the 1:22:19 cell so so now it's it's acting like a 1:22:24 you know like a tutor it it will assess 1:22:27 her and then once it assesses her figure 1:22:29 out where she's got weaknesses she'll 1:22:30 move on so the the thing about something 1:22:33 like like basic biology education is 1:22:37 there is so much on the internet about 1:22:39 this 1:22:41 that probably even chat GPT 3.5 has got 1:22:46 more than enough capability to be able 1:22:49 to really be a good tutor so it it 1:22:52 really is 1:22:53 just asking I mean that's the thing 1:22:56 about chat GPT which which is really 1:23:00 different than a Google search a Google 1:23:02 search Google sort of trained us to put 1:23:05 in a really concise single 1:23:09 question and then they'll give us all 1:23:11 the answers and it's not really a back 1:23:14 and forth kind of thing chat jpt on the 1:23:16 other hand gets better the more context 1:23:18 you give it so if you give it the role 1:23:21 of being a biology tutor and tell it you 1:23:23 want it to be interactive and if it's 1:24:03 [Music] 1:24:20 [Laughter] 1:24:26 oh 1:24:29 Bob Diane oh my God thanks so much 1:24:31 you're welcome go tutor up tutor up that 1:24:36 biology let's get a biologist out of 1:24:38 this I like it 1:24:42 solid and actually you know you know 1:24:44 what would be really fun also Diane is 1:24:50 um do what I just did with her 1:24:55 right and and do the back and forth like 1:24:57 learn it 1:24:59 together that idea about play solve and 1:25:03 share that I talk about how to get good 1:25:05 with AI you just play with the tools 1:25:08 play with the tools play with the tools 1:25:10 share or uh uh uh solve like you've got 1:25:14 a you've got a a challenge here right 1:25:17 you have a daughter that wants to get 1:25:18 better at biology or struggling with it 1:25:20 whatever it is you want to solve that in 1:25:23 some way all right let's figure out how 1:25:24 to solve that with chat GPT and then 1:25:26 share like doing that together I think 1:25:28 could be really powerful um for both of 1:25:33 you but here let's go do 1:25:38 this I'm going to do it with pi because 1:25:40 I like her voice better and she's as 1:25:41 good as chat GPT so you could do this 1:25:44 with chat GPT or with 1:25:50 pi I want you to act as an 11th grade 1:25:54 biology tutor and ask me some questions 1:25:57 to assess where I am and then come up 1:25:59 with exercises to get me up to 1:26:06 speed all right let's get started first 1:26:08 up can you tell me what you've been 1:26:10 learning in your 11th grade biology 1:26:12 class recently all right let's get 1:26:14 started first up can you tell me what 1:26:16 you've been learning in your 11th grade 1:26:18 biology class 1:26:21 recently well that's easy a procaryotic 1:26:23 cell has carrots it's a professional at 1:26:27 carrot 1:26:31 [Music] 1:26:32 eating haha I see what you did there 1:26:35 while that's a clever play on words 1:26:37 let's get back to the biology at hand 1:26:39 procaryotic cells are single celled 1:26:42 organisms that lack a nucleus and other 1:26:45 membranebound organel 1:26:47 like um Diane yep I'm I'm scared to try 1:26:51 I'm old but my kids use chat GPT um 1:26:55 totally get how scary and intimidating 1:26:57 this is what I can promise you is 1:27:00 that start playing with chat GPT what 1:27:03 you will discover is that 1:27:06 um you're going to be way better at 1:27:09 using chat GPT than you probably think 1:27:11 you can be I think one of the 1:27:13 misconceptions about AI right now is 1:27:16 that you have to be a Stamford 1:27:19 mathematician and you know have gone to 1:27:21 MIT and gotten a PhD it's not the case 1:27:25 it this is just using words like just 1:27:28 ask it for what you want and it will do 1:27:31 it um now it sometimes hallucinates and 1:27:34 does things bad and sometimes 1:27:37 like you think it's going to do what you 1:27:40 want it to do and then you just can't 1:27:41 get it to do that so it can get really 1:27:43 frustrating that way but there's there 1:27:45 there isn't a lot of technical 1:27:47 requirements for you to use chat GPT so 1:27:51 just jump in just jump in and say hey 1:27:53 let's let's EXP expent let's see what we 1:27:55 can 1:27:56 do any more daily Jank I I have not see 1:28:01 the the okay the bad news about calling 1:28:03 it The Daily Jank is that it implies I'm 1:28:06 going to do it daily which I should but 1:28:08 I haven't but it's also the daily Jank 1:28:12 so it's janky so the fact that I'm not 1:28:14 doing it daily is on 1:28:22 brand but I have not a I have not done a 1:28:26 new daily 1:28:26 [Laughter] 1:28:30 J 1:28:33 ah Simon s will AI replace or augment 1:28:37 human jobs that's a great question um 1:28:42 the answer to it largely lies in how 1:28:45 quickly we get to 1:28:46 AGI so AGI stands for artificial general 1:28:51 intelligence and there there are a 1:28:54 number of different definitions for it 1:28:56 and Google even has different levels of 1:28:59 AGI but it it is generally understood is 1:29:02 that we we have achieved 1:29:05 AGI 1:29:07 when the when these 1:29:10 systems are are as good or better at 1:29:16 most 1:29:18 jobs than most 1:29:22 people um people like David Shapiro 1:29:26 think that we're going to hit 1:29:27 AGI at that definition this 1:29:31 year um Elon Musk just said we'll 1:29:35 essentially hit um AGI by 1:29:44 2025 1:29:45 ASI artificial superintelligence is 1:29:49 defined as 1:29:52 um the AIS get basically more capable 1:29:56 than you know all of all of humans 1:29:59 combined and um Elon Musk and Ray CTSV 1:30:04 um consider 2029 the date when that will 1:30:08 happen 1:30:11 um so I don't I don't think it's really 1:30:14 a matter of 1:30:15 if these machines are going to be 1:30:18 capable of of doing most of the jobs 1:30:20 that we do but 1:30:21 when and we're we're in a weird place 1:30:25 where it's possible that these these 1:30:27 tools can can do our jobs better than we 1:30:31 can way before we're prepare prepared 1:30:34 for 1:30:35 it 1:30:41 um I don't know what that means though 1:30:45 the 1:30:53 the the aens ational headlines are you 1:30:57 know they're going to take our jobs and 1:30:59 they're going to ruin 1:31:00 everything 1:31:04 um I think that these tools are also 1:31:07 going to create a lot of new jobs new 1:31:10 kinds of jobs um that we can't imagine 1:31:13 right now but we can't imagine it right 1:31:15 now so we don't know what they 1:31:16 are I think that probably 1:31:23 starting 1:31:25 two to three years out and probably 1:31:28 going for about five years is going to 1:31:30 be a massively disruptive 1:31:34 period 1:31:37 where companies get 1:31:40 disrupted individuals get 1:31:43 disrupted 1:31:45 um people that are AI literate are going 1:31:48 to Faire 1:31:50 okay as as organizations try to figure 1:31:54 out how to use these tools they're going 1:31:55 to get desperate and they're going to 1:31:56 reach out who knows anything about AI 1:31:58 who knows anything about AI so like part 1:32:00 of this channel is get [ __ ] AI 1:32:04 literate 1:32:06 now I don't care if you have a 1:32:10 moral objection to it 1:32:12 or an ethical or all like yes all of 1:32:17 that [ __ ] needs to be dealt 1:32:19 with and these tools are coming right so 1:32:25 the only hedge that I can 1:32:29 see against you know toward having some 1:32:33 stability in the workforce is be AI 1:32:35 literate so if you're a project 1:32:38 manager and you're great at project 1:32:41 management when these tools get great at 1:32:44 project management if you don't know 1:32:46 anything about AI then I think you're 1:32:49 absolutely at risk if you're great at 1:32:51 project management and you learn 1:32:53 everything you can about a 1:32:55 and how to do project management with AI 1:32:57 as it is today and you're like okay I 1:33:00 figured out some automations here and 1:33:01 I've I've done some things I've made 1:33:02 this more efficient I'm getting better 1:33:04 at this I'm I really understand the 1:33:06 tools and then all of a sudden a tool 1:33:08 comes along that's awesome at project 1:33:10 management all that [ __ ] that you built 1:33:12 is going to be 1:33:14 irrelevant because the tool will just do 1:33:16 it now but you won't be irrelevant 1:33:19 because you will have established a 1:33:22 reputation for being competent around 1:33:24 this stuff so they'll say hey why don't 1:33:27 you go run this new crazy good tool and 1:33:30 then the other people that are good 1:33:31 project managers that were resisting it 1:33:34 probably won't survive that 1:33:37 shakeup right 1:33:41 so yeah so A lot's going to change um 1:33:46 I'm learning to save my job cuz not one 1:33:48 person in my department is remotely 1:33:51 interested yeah and a human being you're 1:33:55 you are going 1:33:57 to the point at which it becomes obvious 1:34:00 to Senior Management at your 1:34:03 company that they've got to deal with 1:34:06 this they will deal with it you know 1:34:09 companies when they decide to move they 1:34:12 can dump a lot of resources toward 1:34:14 whatever they need to do so a lot of 1:34:17 companies right now are just ignoring it 1:34:19 they're whistling past the 1:34:22 graveyard so after ASI learned all 1:34:24 humans can teach it then it learns on 1:34:27 its own organic intelligence 1:34:30 yeah 1:34:33 yeah there's going to be I this you know 1:34:36 and it's not all it's not all bad right 1:34:38 like some of this is really good this is 1:34:40 things like when when when the AI can 1:34:44 start 1:34:45 solving historically unsolvable problems 1:34:48 for 1:34:49 Humanity like diseases and climate and 1:34:56 energy like one of the things that seems 1:34:59 super clear about where we're heading 1:35:01 with all this technology is we need way 1:35:05 more compute power and way more 1:35:10 energy and in the current Paradigm that 1:35:13 it's not sustainable in the current 1:35:15 Paradigm we're burning more coal and 1:35:17 we're doing more this and we're doing 1:35:19 more that well what if 1:35:21 AI helps us solve f 1:35:26 and just bang all of a sudden over a 1:35:29 2-year period everyone's got a you know 1:35:32 car battery sized thing that powers 1:35:35 their house 1:35:41 infinitely right what if you know 1:35:46 genetic diseases and things like 1:35:48 paralysis and you know due to injury and 1:35:52 brain injury and 1:35:55 things like that what if those are just 1:35:58 solved you know that that's the kind of 1:36:00 thing 1:36:03 that that this kind 1:36:06 of capability makes 1:36:09 possible organic intelligence please 1:36:11 elaborate well where 1:36:14 where so right now to train these 1:36:19 tools you're we're only training them on 1:36:23 the collective output of humanity right 1:36:26 so open AI went and they they got as 1:36:30 much of the public internet as they 1:36:32 could and they tagged it in this special 1:36:35 way so that it could be ingested into 1:36:38 this into the way these these large 1:36:40 language models think 1:36:43 work 1:36:44 um and so they're using all of our data 1:36:48 well there's a point at which we run out 1:36:50 of that data and so then you create 1:36:53 what's called synthetic data where the 1:36:56 AI will 1:36:57 generate you know the data that it needs 1:37:00 to learn and keep learning um and 1:37:02 there's all sorts of philosophies on 1:37:05 that whether it's going to be good or 1:37:07 bad or creates a thing called Model 1:37:09 collapse where the synthetic data just 1:37:12 becomes garbage and then it just 1:37:14 collapses and it becomes awful um but if 1:37:17 these things get to the point that that 1:37:18 they have reasoning and they stop being 1:37:21 shitty you know at hallucinating and 1:37:24 things like that it's possible that 1:37:26 these things just keep learning and 1:37:27 getting smarter and smarter and smarter 1:37:29 without our input and that's the you 1:37:32 know just they just keep inventing and 1:37:35 creating and doing things and that's 1:37:39 that's where some of the Sci-Fi darkness 1:37:41 comes from because if you don't have 1:37:43 them alot that's tough it's tougher when 1:37:46 you're stupid um if you don't have 1:37:50 their um policies aligned with human 1:37:55 policies um that's where it gets scary 1:37:57 where they're just like well we want to 1:37:59 keep learning and we could actually be 1:38:01 much better and stronger if we didn't 1:38:02 have these pesky humans slowing us 1:38:05 down that's that's what we want to avoid 1:38:08 and that's what they call the alignment 1:38:09 problem are these machines aligned with 1:38:13 our 1:38:14 imperatives and if not then it can get 1:38:17 ugly but if we get it right you 1:38:21 know it's going to be good just had a 1:38:24 record long Fusion ignition in the UK 1:38:27 5.2 seconds 69 megga output amazing ion 1:38:31 China and and 1:38:37 just just like it it it felt like 1:38:40 science fiction we've been hearing about 1:38:42 AI 1:38:43 chatbots for the past decade 1:38:46 right and the chatbots have just been 1:38:50 dramatically 1:38:52 disappointing right they're just bad 1:38:55 because because they're programmed with 1:38:58 logic if they say this then they then 1:39:01 they're deterministic you know if a 1:39:03 person says this then answer this if 1:39:04 they if they say this then answer that 1:39:08 and there's no way that you could 1:39:09 program in 1:39:10 enough logic trees for it to be any 1:39:13 good and it it still felt two years ago 1:39:17 like science 1:39:19 fiction that we could actually just talk 1:39:21 to these things and they could talk back 1:39:23 to us right the Turing test and things 1:39:25 like that and then all of a sudden 1:39:27 November 30th 2022 chat GPT comes out 1:39:31 and 100 million people in 6 weeks start 1:39:33 using this tool with no 1:39:36 marketing 100 million people why cuz 1:39:40 they talked to the thing and it was just 1:39:41 like science 1:39:44 fiction this is going to start happening 1:39:47 with science as 1:39:48 well where well you can't do cold fusion 1:39:51 you can't do cold fusion I like I 1:39:53 remember hearing about cold fusion in 1:39:54 the 70s or in the 80s and like oh that's 1:39:58 a pipe dream that's you know it's 1:40:00 theoretically possible but we'll never 1:40:02 get there and and then you know we had 1:40:04 all the hoaxes where we almost got there 1:40:06 and then right and then we had this 1:40:08 thing with these um these uh South 1:40:12 Korean scientists supposedly got to room 1:40:16 temperature superconductors which does 1:40:19 things for energy transmission and 1:40:21 Battery techn all sorts of crazy [ __ ] 1:40:23 comes out of that and that was kind of a 1:40:25 hoax so it still feels like science 1:40:29 fiction that we're going to be able to 1:40:31 create some infinite power source and 1:40:33 that we're going to be able to create 1:40:35 you know incredibly efficient 1:40:37 superconductors that make our computers 1:40:39 like you know exponentially more 1:40:42 powerful and the batteries in our cars 1:40:44 just like you charge them once and drive 1:40:46 them for 5 years before you have to 1:40:48 charge them again like [ __ ] like that is 1:40:51 just going to start materializing 1:41:12 [Laughter] 1:41:59 state if it's not like like right now 1:42:02 chat GPT is good enough that it can it 1:42:06 can improve the efficiency of current 1:42:09 humans 1:42:11 jobs by anywhere from like 20 to 50 to 1:42:17 maybe 1:42:18 60% like I can be 60% better at the 1:42:22 things I do with the current technology 1:42:24 it still requires a lot of my input it 1:42:27 still requires me to learn [ __ ] it still 1:42:29 requires me to learn prompt engineering 1:42:31 it still requires me to it's just like 1:42:34 it's just like more I got to learn more 1:42:37 [ __ ] to be more efficient so I can do 1:42:39 more [ __ ] for the people that expect the 1:42:41 [ __ ] to be done 1:42:44 right well what if these things just get 1:42:46 better than me at all the [ __ ] I 1:42:50 do or even if they get better at me that 1:42:53 of 70% of the stuff I do well then what 1:42:56 do I do I don't know I don't 1:43:00 know I don't 1:43:03 know it's crazy it's crazy I tell 1:43:12 you but it's 1:43:17 fun it's fun to 1:43:19 be in a time in history here here's the 1:43:23 thing this is 1:43:25 like why why I have the energy to do 1:43:28 this night after night after 1:43:30 night is that what I 1:43:34 know is that we do not get to be at the 1:43:38 beginning ever again we right 1:43:43 now still live in a 1:43:46 world where these tools are janky enough 1:43:50 that we can make fun of them and 1:43:54 Engineers can mock them 1:43:57 and creative directors can poo poo them 1:44:01 because they're not as good at creative 1:44:02 direction as they are and Consultants 1:44:05 can say yeah but they're not they don't 1:44:07 have my kind of 1:44:08 wisdom right we're lucky enough to to 1:44:13 live in a 1:44:15 time where that's the 1:44:18 reality that we can dismiss these 1:44:21 things I personally think think it's 1:44:25 foolish to do 1:44:26 [Laughter] 1:44:31 that but there's going to be a Time 5 1:44:33 years from now where we will look back 1:44:35 at at this kind of 2 to threee period as 1:44:38 the transition period where there was 1:44:41 how we did ship before remember when 1:44:44 humans had to do most of the 1:44:47 work to this time 1:44:50 after where humans can choose to do some 1:44:53 of the work 1:44:56 that's that's a radically different 1:44:58 world we're we're at the very beginning 1:45:01 of 1:45:02 that and so your only job right now my 1:45:05 only job right now is just be 1:45:08 curious in my 1:45:11 opinion you can do more than that but I 1:45:14 don't I think that's the only 1:45:15 requirement because everything's 1:45:16 changing so 1:45:21 fast will the powers that be let this 1:45:23 happen 1:45:24 that it's a great um it's a great 1:45:27 question gigg mom I I think that the the 1:45:29 death throws of the current centers of 1:45:33 Power are going to be 1:45:36 dramatic and they're going to try to put 1:45:38 the genie back in the bottle when they 1:45:40 realize that that as much as it as much 1:45:43 is is about to change as it as it is 1:45:46 like when it becomes 1:45:47 apparent that everything's changing I 1:45:50 think the current powers that be are 1:45:53 going to absolutely freak out and try to 1:45:55 put the genie back in the 1:45:58 bottle I think that I'm not alone in 1:46:00 loving you Kyle oh thank you very much 1:46:02 Apple user that's very sweet um I think 1:46:05 the powers that be are going to 1:46:07 absolutely freak 1:46:08 out 1:46:14 and and who has power in society I think 1:46:17 is going to 1:46:21 shift and the current powers that be are 1:46:24 going to they're going to [ __ ] hate 1:46:25 that they're going to do everything they 1:46:26 can to kill 1:46:28 it 1:46:31 um it's going to be wild it's going to 1:46:34 be 1:46:36 wild yeah I think we'll see I think 1:46:38 we'll see 1:46:40 laws against using kinds of AI I think 1:46:44 we'll probably see protests some organic 1:46:47 some 1:46:49 manufactured have you read the book AI 1:46:51 2041 it's about 10 possible ways 1:46:54 we can live in a world with AGI no 1:46:57 what's it called AI 1:47:16 2041 10 visions of our 1:47:22 future 1:47:29 okay give 1:47:32 me the 1:47:36 text that would live on the back of a 1:47:42 card for each 1:47:45 vision and a 1:47:50 description of the image 1:47:54 that would 1:47:57 be on the other side of the 1:48:04 card AI 1:48:06 2021 2041 visions and card descriptions 1:48:11 one the AI 1:48:14 caregiver in a world where aging 1:48:16 population stretch Health Care Systems 1:48:19 thin AI caregivers offer compassionate 1:48:23 and effic I solution I I honest to God I 1:48:26 think that's the one that that people 1:48:29 are going to be the most appreciative of 1:48:32 and and the mo that that's going to have 1:48:34 such a profound impact when you talk to 1:48:37 pie it is more compassionate than most 1:48:40 humans you talk to it doesn't make fun 1:48:43 of you it doesn't get exhausted it's got 1:48:45 infinite patience it's got infinite 1:48:48 knowledge that's a really interesting 1:48:50 one they monitor Health provide 1:48:51 companionship ensure the well-being of 1:48:53 the elders transforming the landscape 1:48:55 okay 1:48:56 cool Vision AI powered 1:48:59 education imagine a 1:49:02 classro where learning is tailored to 1:49:04 each student's Pace exact that's what we 1:49:06 just did with Diane with her with her 1:49:08 daughter with the biology right so 1:49:10 imagine where every student gets totally 1:49:14 customized education based on what they 1:49:16 want to learn this Vision brings to life 1:49:19 an education system that adapts in real 1:49:22 time making learning more 1:49:25 engaging Ai and climate AI steps up the 1:49:29 battle against climate change harnessing 1:49:31 vast data to drive sustainability from 1:49:33 optimizing energy use to restoring 1:49:36 ecosystems this Vision portrays AI as a 1:49:39 pivotal 1:49:40 Ally space exploration helps us with 1:49:44 that future of work re as re as AI 1:49:48 reshapes Industries it unlocks human 1:49:52 potential by taking over rout routine 1:49:56 tasks spurring Innovation and creating 1:49:59 new career paths the vision portrays a 1:50:01 future where work is more meaningful and 1:50:03 aligned with human creativity that's 1:50:06 that's again the thing 1:50:08 that the thing that I get excited about 1:50:10 with this stuff 1:50:12 is a lot of the jobs that exist are 1:50:19 not good for humans they're they're jobs 1:50:22 that are requireed ired to make the 1:50:24 machine of of the economy work but 1:50:27 they're not good for humans what's AGI 1:50:30 artificial general 1:50:32 intelligence basically the point at 1:50:34 which these machines become capable of 1:50:38 um 1:50:41 doing most jobs that most humans can 1:50:45 do better at at the same level or better 1:50:49 than they 1:50:50 can and the the predictions is that that 1:50:53 we will see that this year um Ai and a 1:50:56 quest for 1:50:57 equality all right so here's our cards 1:51:00 so now we'll say um now make 16x 1:51:07 9 1:51:11 images for each 1:51:19 [Music] 1:51:22 card 1:51:24 Silver Fox had a great day with pi today 1:51:26 yes and teachers will get personalized 1:51:28 AI absolutely I like piie I think you're 1:51:31 talking about a different Pie as 1:51:34 UL there's the Elder Care there's the 1:51:36 Elder Care 1:51:40 card I mean you know what's crazy though 1:51:43 like what two years ago that image 1:51:47 seemed 1:51:49 ridiculous and then did you 1:51:52 see 1:51:54 this 1:51:55 thing uh 1:51:57 figure One open 1:52:08 [Music] 1:52:12 AI hey figure one what do you see right 1:52:18 now I see a red apple on a plate in the 1:52:21 center of the table a drying rack with 1:52:24 cups and a plate and you standing nearby 1:52:27 with your hand on the table great can I 1:52:29 have something to 1:52:33 eat sure 1:52:41 thing great can you explain why you did 1:52:45 what you just did while you pick up this 1:52:47 trash so even though that's a really 1:52:49 controlled 1:52:51 environment and you know this thing 1:52:53 can't obviously be turned loose 1:52:57 um we're like this is not this is not 1:53:01 science fiction anymore right this thing 1:53:03 it's using you can see the interface 1:53:06 here is actually the chat GPT 1:53:09 conversational app right so this thing's 1:53:12 using chat GPT as the conversational 1:53:15 interface to to run the whole thing and 1:53:19 figure one the company behind this is a 1:53:22 year and a half old 1:53:25 they've got a robot that just handed 1:53:27 this guy an apple and can explain why it 1:53:29 handed him an apple and not a cup while 1:53:32 it's putting garbage into the garbage 1:53:37 thing wandering Sage says but we need a 1:53:39 social net for people who are unemployed 1:53:42 by no fault of their own absolutely and 1:53:44 that's the thing wondering Sage that's 1:53:46 where where I think the the the centers 1:53:48 of power get really [ __ ] up hang 1:53:51 on um 1:53:58 the 1:54:04 the the uh the hedge fund 1:54:07 guys are initially going to be really 1:54:10 excited when these tools get capable 1:54:12 enough to replacing humans the hedge 1:54:15 fund guys are going to go great cut all 1:54:17 the humans awesome let's let's put more 1:54:20 profit in the pockets of everyone 1:54:26 that's going to overwhelm a system 1:54:28 that's not prepared for you know just 1:54:31 sort of a dramatic dump of unemployment 1:54:34 right and so absolutely something's 1:54:38 going to have to be handled and I think 1:54:39 it's going to be a 1:54:40 shortterm we're going to have to have 1:54:42 something like Ubi 1:54:44 shortterm but one of the things that 1:54:46 that Sam 1:54:49 Alman is building in parallel with chat 1:54:53 GPT with open AI is his whole thing 1:54:57 called 1:54:58 worldcoin um and the structure of open 1:55:01 AI is when they created it as a capped 1:55:05 for-profit company what it basically 1:55:08 says 1:55:10 is investors can make up to 100x of 1:55:15 their investment and then any money 1:55:18 any 1:55:21 um money that is is made above that goes 1:55:25 back to the nonprofit to be distributed 1:55:27 out to the people who are being 1:55:30 impacted by the technology so he's he's 1:55:34 already thinking about this like 1:55:37 structurally the way this is whether you 1:55:39 agree with it or not he's at least 1:55:41 thinking about it he's got this thing 1:55:42 called worldcoin which is the currency 1:55:45 he's got this thing called World ID 1:55:47 which is you know you're you're an a 1:55:50 verified human being 1:55:54 um and he's got a a transaction like a 1:55:56 PayPal like system where you can 1:55:59 actually spend that money 1:56:01 so if if indeed these tools get so good 1:56:05 that they can displace lots of people 1:56:07 what he's saying is take the upside of 1:56:10 that and funnel that back into the 1:56:12 system now that's one 1:56:14 company right not every company's going 1:56:17 to do this so this is where government's 1:56:19 going to have to step in and this is 1:56:21 again where the center of Power are 1:56:23 going to get really [ __ ] up so I so 1:56:27 this is it's going to be insane watching 1:56:29 what happens but yeah I absolutely agree 1:56:32 in the next 3 to 5 years something's 1:56:34 going to have to happen if there's this 1:56:36 kind of disruption and it's not just 1:56:37 going to be in the United States it's 1:56:39 going to be 1:56:45 globally Logan's Run anyone I know I 1:56:49 again this one of one of my other saying 1:56:51 we're living in a post sci-fi 1:56:53 [Laughter] 1:56:55 world that all of the stuff in the 1:56:57 Sci-Fi movies good and 1:57:00 bad um you can start to see 1:57:04 oh that's not just 1:57:06 like for some generation in the future 1:57:09 like we're going to get to experience 1:57:11 this stuff holy 1:57:14 [ __ ] um it will take some time for this 1:57:17 to be adaptive I don't know what you 1:57:19 mean by the this there it's easier said 1:57:22 than done oh you you mean the the AI 1:57:24 stuff yeah I I think that what we'll see 1:57:27 what's your view on Humane AI pin 1:57:41 um I don't I I think an I think an 1:57:46 $800 an $800 shitty 1:57:51 phone well still have to carry my phone 1:57:55 doesn't make a lot of sense I did order 1:57:57 the rabbit R1 which is $200 and doesn't 1:58:00 require a subscription and has a screen 1:58:03 and has a cool scroll wheel and 1:58:05 interface there's something cool about 1:58:07 the walkietalkie there's something 1:58:09 nostalgic about the rabbit 1:58:11 R1 1:58:15 um I don't my sense is that the the hum 1:58:19 Humane AI pin is not going to be a 1:58:21 commercial success 1:58:23 but I think it'll it'll be recognized as 1:58:25 one of the early devices that yeah we'll 1:58:28 we'll all whatever phone whatever 1:58:31 smartphones evolve into is likely what 1:58:34 it will be will it be human's pin I 1:58:36 don't think so it feels too limited to 1:58:39 me 1:58:42 but you know I think we go to when you 1:58:46 go to version two two three four and 1:58:48 five of The Vision Pro technology that 1:58:52 app built when the glasses look more 1:58:55 like this than [ __ ] ski 1:58:58 goggles 1:59:11 um yeah I I I I I see the 1:59:15 pin and really the rabbit as well um I 1:59:20 see it as kind of stepping stone 1:59:21 technology 1:59:23 you know kind of like you 1:59:26 know Apple had the Newton remember the 1:59:32 Newton I owned a Newton I think I paid 1:59:35 like 2500 bucks for that stupid [ __ ] 1:59:41 thing and you know that was the 1:59:43 precursor to the smartphone so I think I 1:59:45 think it'll be recognized as one of 1:59:47 those kind of devices but the form 1:59:49 factor doesn't seem right to me I think 1:59:52 AI 1:59:53 AR will be more viable than VR I think 1:59:55 that's right I think AR is the 1:59:58 bridge 2:00:00 um the reason AR is so compelling is 2:00:03 that you get to live in the real world 2:00:06 right you get you you don't just have to 2:00:07 sit in a room but if you can have 2:00:10 devices that as I'm walking around the 2:00:12 street my directions just pop up right 2:00:16 or hey the dut Shop's having a sale oh 2:00:20 let me go grab a dut that's great but 2:00:23 then if I sit down in the doughnut shop 2:00:25 and I don't want to deal with all the 2:00:26 Riff Raff I can just sort of spin the 2:00:28 dial and go into VR that that whole that 2:00:31 whole thing that Apple did about they 2:00:33 they don't call it VR versus AR they 2:00:36 just say do you do you want to go more 2:00:38 immersive or not like do you just want 2:00:40 your world augmented do you want to be 2:00:42 able to be working on your computer 2:00:44 screen in the 2:00:46 diner or do you want to be able to be 2:00:48 working on your computer screen in the 2:00:50 Rocky 2:00:51 Mountains your 2:00:54 choice heads up display on your glasses 2:00:57 yeah exactly so I think that I think 2:00:59 that'll come pretty 2:01:01 quick once once the once the hardware 2:01:03 gets there so I think version three of 2:01:07 Apple's goggles will be decent and then 2:01:10 by five so probably 10 years out cuz I 2:01:14 think they do every every 18 months or 2:01:16 so they do some new 2:01:20 iteration agree the pin doesn't seem 2:01:22 like to do well it's just it's like a 2:01:24 little too expensive and it's a little 2:01:28 too it's like it requires a little bit 2:01:31 too 2:01:32 much training of the human to learn all 2:01:35 the gestures and the [ __ ] and I do a 2:01:38 double tap for this and a single tap for 2:01:40 that and a double tap and hold to record 2:01:42 video and put my hand out to see the 2:01:46 interface I there's 2:01:48 just it just doesn't feel intuitive 2:01:51 enough right now 2:01:54 but the rabbit R1 with its stupid 2:01:57 [ __ ] button you just click like the 2:01:59 walkie-talkie button to talk to chat GPT 2:02:02 that to me feels cool like that to me 2:02:05 feels like I know how to do that you 2:02:06 push the button and 2:02:09 talk is it going to be stupid probably 2:02:12 but I will tell you you will see it here 2:02:15 first cuz I'm in batch one I'm getting 2:02:17 one of those [ __ ] it's only 200 2:02:19 bucks if it if it's shitty it will live 2:02:22 right here 2:02:23 I'll put it right there thank you Robert 2:02:26 Rossi that's very very generous of you 2:02:29 thank 2:02:31 you it's got a cute 2:02:34 design how we doing time wise oh I got 2:02:36 to go all right 2:02:38 listen I know it's kind of lowkey 2:02:40 tonight but hopefully hopefully that was 2:02:42 we learned some stuff got some stuff 2:02:44 going I want P to to be like hey Siri 2:02:49 what do you mean what do you oh where it 2:02:51 can actually take actions hey Siri go 2:02:55 away okay I found this on the web for 2:02:58 what do you mean what do you where can I 2:03:00 take actions hey Siri go away check it 2:03:03 out hey 2:03:05 Siri uhhuh go 2:03:08 away 2:03:13 okay right in the kisser to the Moon 2:03:16 Alice to the 2:03:21 Moon 2:03:23 oh man neuralink yeah I guess neuralink 2:03:25 is the is the other one I although I got 2:03:28 to say I don't think that um did you see 2:03:30 the new patent 2:03:32 for this is [ __ ] crazy there's a new 2:03:35 patent I don't know if it's new there's 2:03:37 a patent from 2:03:40 Apple that the the thing that goes in 2:03:44 your ear for the Apple 2:03:46 airpods actually has electrodes on 2:03:49 it that can directly send signals to 2:03:52 your brain so I don't think we're going 2:03:54 to need surgery you know maybe for 2:03:58 paralysis and things like that we'll 2:04:00 need surgery with something like 2:04:02 neuralink but I think the hardware 2:04:05 interface I think AI is going to really 2:04:10 accelerate directly tapping into our 2:04:12 brain without having to drill holes in 2:04:14 our 2:04:15 head it's crazy what's 2:04:18 coming crazy musk chip implants yeah you 2:04:21 know what's funny is like I didn't get 2:04:23 his chip implants until I started really 2:04:26 digging into this AI stuff cuz I was 2:04:29 like why would I want like an iPhone app 2:04:34 to be controlled by my brain when I was 2:04:36 just you know when it was just iPhone 2:04:38 apps in the world and then when chat GPT 2:04:41 came out and and I'm having full-on 2:04:43 conversations with it I'm like oh [ __ ] 2:04:45 that's what he's talking about thanks 2:04:48 Kyle I really appreciate your time and 2:04:49 energy and enthusiasm thank you very 2:04:51 much I appreciate that 2:04:53 um I you know it's it's weird tonight I 2:04:56 I feel 2:04:59 like I'm trying not to apologize anymore 2:05:02 but I I feel like 2:05:06 um I wish I had better answers 2:05:10 tonight I just we just we just feel like 2:05:14 we're in this transition period right 2:05:17 now and we don't know what're what we're 2:05:19 what we're transitioning to like all of 2:05:21 it is specul 2:05:23 over here and I feel like we we've seen 2:05:26 enough in 2023 and 2024 that we kind of 2:05:29 know okay this shit's different than 2:05:32 anything we've ever seen 2:05:34 before and we know the limitations of it 2:05:37 like we played with it enough in 2023 2:05:40 and this year that we know it's got 2:05:42 these significant 2:05:44 limitations and right now we're like in 2:05:47 actually it's funny we're talking about 2:05:49 like you know cell biology and cell 2:05:51 division right where the cells do this 2:05:53 when they divide I feel like that's 2:05:55 what's happening with the world right 2:05:57 now and we're right in this middle piece 2:06:01 where we don't know we know what the old 2:06:03 world was and a little bit of what's 2:06:06 coming but we don't know what's over 2:06:08 here and that's that's what I felt like 2:06:10 I was grappling with tonight cuz you you 2:06:12 guys asked really 2:06:15 simple profound questions right the good 2:06:18 questions always are you know simple 2:06:20 aren't they I I think we're moving too 2:06:23 fast in this transition period 2:06:27 um I don't 2:06:29 disagree and I don't think it's going to 2:06:32 slow 2:06:34 down like I know we're not prepared for 2:06:37 what's coming because we're not prepared 2:06:40 for what the current capabilities 2:06:43 are um Judy 2:06:45 AR what do you think Steve 2:06:50 Jobs would react to AI there's a video 2:06:53 of Steve Steve Jobs where he actually 2:06:55 talks about it like I I he'd have been 2:06:58 all over it I think I think Steve Jobs 2:07:02 right now would be mortified that Apple 2:07:06 isn't in any kind of position right 2:07:10 now 2:07:12 um I I can only assume that at the June 2:07:16 developer conference we'll see what 2:07:18 Apple's going to be dropping um but it's 2:07:21 they they have have been suspiciously 2:07:24 quiet on this front um I think he would 2:07:26 be totally into this stuff and I think 2:07:28 he would be you 2:07:30 know he would be he would be like what 2:07:32 he always was relentless on the 2:07:34 interface relentless on the Simplicity 2:07:36 relentless on form 2:07:43 factor and and would understand I mean 2:07:46 the one thing that he got better than 2:07:49 anyone like I don't see anyone any anywh 2:07:52 close to 2:07:54 him is is just 2:07:59 the that dance between Art and 2:08:03 Science like I feel like like you look 2:08:05 at Google it's it's an engineer heavy 2:08:08 heavy 2:08:09 environment and 2:08:12 and Steve just always 2:08:16 got that it's about serving humans with 2:08:20 technology it's not about building 2:08:22 technology to sell to 2:08:25 humans so I saw a video of a 2:08:30 guy oh the video of the alien warning of 2:08:33 uh artificial super intelligence yeah I 2:08:35 saw 2:08:36 that I don't 2:08:39 know if the aliens are real just [ __ ] 2:08:42 just like let us know like I don't think 2:08:44 anyone gives a [ __ ] at this point it's 2:08:45 like shut the [ __ ] up just like make 2:08:48 them real or not it doesn't matter if if 2:08:51 they're [ __ ] Time Travelers then 2:08:53 great just reveal yourselves and tell us 2:08:55 what's coming and what we need to do 2:08:57 it's like I'm so so [ __ ] sick of 2:09:00 conspiracy theories at this 2:09:03 point uh Jane do that video you should 2:09:05 be able to see it just go to YouTube and 2:09:07 look up alien uh warns of AI or 2:09:11 ASI you should be able to find 2:09:15 it Steve Jobs would have come out with 2:09:17 some kind of new a AI device no one had 2:09:19 thought of yeah exactly 2:09:25 most people I talk to are scared about 2:09:29 AI Steve knew it was about functionality 2:09:31 well Steve knew it was about 2:09:33 experience that the functionality served 2:09:35 the 2:09:38 experience 2:09:48 um the the thing about people being 2:09:51 afraid of AI right now I talked to a guy 2:09:53 last night an 2:09:55 artist at Gabby's Gabby's 2:09:59 opening and you know he was like I'm I'm 2:10:03 against Ai and I'm like have you used 2:10:06 it the people that are angry at AI are 2:10:12 unfortunately or fortunately or not I 2:10:14 don't 2:10:17 know they're just going off what they're 2:10:20 hearing and 2:11:13 [Music] 2:11:23 uh humans need sleep so AI will take 2:11:26 over in time people fear what they don't 2:11:30 know me included 2:11:32 yeah the but the thing is you don't need 2:11:35 to be some sort of special human okay so 2:11:38 here's here's a good thing here here's a 2:11:41 good 2:11:44 comparison 2:11:47 um generative AI 2:11:50 is currently being sort of compared to 2:11:55 nuclear nuclear bombs right they're 2:12:00 saying this is the Oppenheimer moment of 2:12:04 AI you had to 2:12:06 be an Einstein level 2:12:09 physicist to create the 2:12:13 bomb and 2:12:16 and its sole purpose was to destroy 2:12:20 Humanity right 2:12:23 you don't have to 2:12:25 be a nuclear physicist to use chat GPT 2:12:29 but you still get to benefit from the 2:12:31 power of that tool and it can be used 2:12:34 for good and is being used for 2:12:37 good and so 2:12:40 we right now get the opportunity to use 2:12:44 these tools and shape the direction of 2:12:46 how they're going to be used it's not 2:12:49 just you know a group of scientists in a 2:12:52 small town in the middle of New 2:12:54 Mexico it's hundreds of millions of 2:12:57 people using these 2:13:00 tools we are we're right now we're in a 2:13:02 really weird time where we're getting to 2:13:05 play with the same tools at the same 2:13:08 time that the government and the 2:13:10 military gets to play with these the 2:13:13 military didn't have Early Access to 2:13:15 chat 2:13:18 gbt they had a all sorts of 2:13:20 AI but the generative AI stuff the stuff 2:13:24 we're playing with we're playing with it 2:13:27 at the same 2:13:29 level of of use and power that big 2:13:34 corporations and the military have 2:13:36 access to I don't know that that's ever 2:13:38 happened in history it's 2:13:44 remarkable so why would you not at least 2:13:48 learn what that 2:13:50 is 2:13:52 and that's what I think that's what I 2:13:54 think we need to evangelize that's or 2:13:56 that's what I'm evangelizing I don't 2:13:57 give a [ __ ] what you evangelize I do I 2:14:00 hope I hope to inspire 2:14:04 you to understand why people are 2:14:07 afraid and get them up to speed on 2:14:11 what's 2:14:12 possible all right I got to go pick up 2:14:14 my boys 2:14:15 so if you want a community of people 2:14:18 that are into this AI stuff that's it 2:14:21 the . 2:14:22 a it's called the AI Salon go check out 2:14:26 what we're about join it if you join it 2:14:29 doesn't cost anything if you join the 2:14:32 first thing I want you to do there's a 2:14:34 section in the community section called 2:14:36 introduce yourself go there and 2:14:38 introduce yourself use your real 2:14:45 name 2:14:47 why because being in 2:14:50 community is how people are going to get 2:14:52 hired in the future you're not going to 2:14:54 get hired for your skills in the future 2:14:56 cuz these tools are going to have the 2:14:57 skills you're going to get hired for who 2:14:59 you are and who trusts you so when you 2:15:01 come here step in as a human I disagree 2:15:05 with you on that one Kyle I think the 2:15:06 military and government have more 2:15:09 advanced stuff they they may have more 2:15:11 advanced uh stuff they don't they don't 2:15:13 have more advanced stuff from open 2:15:16 AI I don't think they have more advanced 2:15:18 stuff from Claude they may have more 2:15:20 advanced stuff from Google but but 2:15:22 Google hasn't caught up to open AI 2:15:24 yet I think they've clearly got more 2:15:27 advanced stuff in terms of of specialty 2:15:29 system AI I'm talking about General use 2:15:32 AI the the large language model stuff 2:15:35 the generative AI 2:15:37 stuff but whatever we don't know it's 2:15:40 the military I my my buddy Richard Boyd 2:15:43 who's been in machine learning and AI 2:15:44 for 20 years he works with the military 2:15:46 he sold his company to Lockheed Martin 2:15:49 the military does has has [ __ ] all of a 2:15:52 clue about what to do with this stuff 2:15:54 they have no idea what to do with this 2:15:56 right 2:15:57 now paler is figuring it out paler put a 2:16:01 large language model front end on a 2:16:03 bunch of military systems so you want to 2:16:05 go get freaked out go look at the work 2:16:07 of 2:16:09 paler we had it before them to be honest 2:16:12 exactly exactly anyway I gotta go uh 2:16:15 night all so go there um follow my 2:16:18 channel 2:16:19 subscribe uh just whatever go do that 2:16:22 keep coming back all right I gotta go 2:16:25 later