WEBVTT 1 00:00:00.520 --> 00:00:05.080 Hudson River Radio dot com. It beats listening to nothing. Frank. Oh 2 00:00:05.120 --> 00:00:11.240 my goodness, it's Frank Being Frank fright where the only way to be is 3 00:00:11.279 --> 00:00:17.559 Frank. Hello everyone, and welcome to Being Frank. We're the only way 4 00:00:17.600 --> 00:00:21.760 to be is Frank. I'm your host, Frank Lebono, and i'd like 5 00:00:21.839 --> 00:00:26.280 to thank you for joining us on what we like to call the Intelligent Conversation 6 00:00:26.519 --> 00:00:31.800 podcast, where no conversation is out of bounds and all points of view are 7 00:00:31.879 --> 00:00:37.719 welcome. We are recording live to tape on this twenty seventh of December. 8 00:00:38.320 --> 00:00:42.520 I hope everyone enjoyed a peaceful and healthy holiday, and I want to thank 9 00:00:42.560 --> 00:00:46.759 you once again for joining us on this very special edition of Being Frank. 10 00:00:47.359 --> 00:00:51.640 Why specially, you may ask, Well, First, it's the last show 11 00:00:51.640 --> 00:00:55.960 of twenty twenty three, and a new year of great guests and intelligent conversation 12 00:00:56.119 --> 00:01:00.679 looms ahead. Secondly, this is our one hundredth episode. I know my 13 00:01:00.799 --> 00:01:04.319 audience may be small, but it is mighty, and I want to thank 14 00:01:04.359 --> 00:01:08.280 all of you who take the time to allow us into your lives, even 15 00:01:08.319 --> 00:01:12.799 if it's just for a little while. Okay, it's the story that simply 16 00:01:12.879 --> 00:01:21.439 won't go away. Artificial intelligence. Is it the next great step in scientific 17 00:01:21.480 --> 00:01:26.519 achievement that can deliver miracles in everything from transportation to medicine, or is it 18 00:01:26.560 --> 00:01:32.840 a great evil which will eventually devour mankind in its effort to create and validate 19 00:01:32.879 --> 00:01:37.280 its own existence. In fact, it's been the topic of more than one 20 00:01:37.359 --> 00:01:42.799 discussion here on being frank. With this episode, we'd like to explore something 21 00:01:42.799 --> 00:01:48.159 a little different, the potential to use AI to create a new type of 22 00:01:48.280 --> 00:01:53.040 art that combines the vast imagination of the human mind with the ability of artificial 23 00:01:53.079 --> 00:02:00.599 intelligence almost limitless capability to realize what was only one dream of My guest is 24 00:02:00.640 --> 00:02:06.319 a journalist, musician, artist, a friend, a former colleague, and 25 00:02:06.359 --> 00:02:10.759 on the cutting edge of using AI technology in many new ways to serve his 26 00:02:10.960 --> 00:02:15.639 art. Please welcome mister Phil Sex and Phil thank you so much for taking 27 00:02:15.680 --> 00:02:19.919 some time to join us. Hey, thanks for having me now. I've 28 00:02:19.960 --> 00:02:23.280 been following your stuff for a while now and I get such a charge out 29 00:02:23.319 --> 00:02:29.639 of it because it is so wildly creative. But before we get to the 30 00:02:29.719 --> 00:02:36.439 process of how you create these things and what role artificial intelligence is as it 31 00:02:36.520 --> 00:02:40.120 is relative to your work and to your art. A little bit about your 32 00:02:40.159 --> 00:02:45.560 background that I mentioned, your journalist, your company. You currently work for 33 00:02:45.680 --> 00:02:50.360 a national news service. I knew you were during our times as CBS News, 34 00:02:50.719 --> 00:02:55.159 and you are basically a photo journalist, mainly as is your job description 35 00:02:55.280 --> 00:03:00.599 there in news. So how did you make the leap from photo journalism which 36 00:03:00.639 --> 00:03:05.560 is something that's rooted in true reality obviously, as we well know, it's 37 00:03:05.599 --> 00:03:10.479 so strict in terms of staying truthful to the story, not adding, not 38 00:03:10.520 --> 00:03:16.199 embellishing, just really showing very much what you see to be able to share 39 00:03:16.240 --> 00:03:23.439 with an audience. To go from there to the wildly creative atmosphere of AI 40 00:03:23.639 --> 00:03:25.919 and some of the work that you're doing now. So how did you make 41 00:03:25.960 --> 00:03:30.439 that transition? How did you go from journalists with a with a very clear 42 00:03:30.520 --> 00:03:38.319 eye to AI which is wildly creative. Well, let's see how to do 43 00:03:38.400 --> 00:03:46.680 that? Well? I made that jump because well, like storytelling. I'm 44 00:03:46.680 --> 00:03:52.240 still a storyteller, and I enjoy storytelling, and you know, I like 45 00:03:52.319 --> 00:03:59.159 reading stories, like hearing stories. I love good stories. And that was 46 00:03:59.360 --> 00:04:04.280 that pretty much. I think that was the curious catalyst that drew me towards 47 00:04:04.719 --> 00:04:13.599 generative AI in particular, once they started making these conversational AIS that that you 48 00:04:13.639 --> 00:04:17.040 know, you can talk to and will respond to you, you know, 49 00:04:17.240 --> 00:04:21.759 in a human way. We'll talk talk a little bit about exactly what that 50 00:04:21.879 --> 00:04:29.240 means conversationally, and there are so many different ways that we interact with AI, 51 00:04:30.160 --> 00:04:31.839 and we have for a long time. And before we came on the 52 00:04:31.879 --> 00:04:36.160 air and I was introduced to to Neil Richter, or engineer or I called 53 00:04:36.160 --> 00:04:41.800 the mailman, but Neil is also in real life a drummer as well too, 54 00:04:42.279 --> 00:04:45.800 and we talked about how, for example, the kind of the use 55 00:04:45.800 --> 00:04:50.759 of drum machines in a sense is AI. It's kind of an artificial intelligence 56 00:04:50.839 --> 00:04:55.839 that senses what kind of a beat that you're going to need and use. 57 00:04:55.920 --> 00:05:00.720 So we've kind of been in some ways using AI for a while, but 58 00:05:02.040 --> 00:05:05.480 now there's this enormous transition. And I know I used to watch it. 59 00:05:05.600 --> 00:05:11.319 You started out with these cubes, these like little music cubes that I used 60 00:05:11.319 --> 00:05:15.879 to watch, used to place on your desk and create almost like whole symphonies 61 00:05:17.040 --> 00:05:23.279 just by touching these cubes again, which was another early form of AI. 62 00:05:24.160 --> 00:05:29.399 So how how is this? What are you inputting? What is generative? 63 00:05:29.800 --> 00:05:31.879 What are all these terms? What do they mean, how do you how 64 00:05:31.920 --> 00:05:36.319 do you make them work for you and your art? Well, you know, 65 00:05:36.600 --> 00:05:40.319 it started off with music because then you know, play guitar, and 66 00:05:40.720 --> 00:05:45.319 I've been trying to play in a band, and of course as a freelancer, 67 00:05:45.439 --> 00:05:48.120 I'm a terrible bandmate. You know. The show's are on Fridays and 68 00:05:48.199 --> 00:05:54.759 Saturdays, and that's usually when I'm working, and the bands can't pay what 69 00:05:55.160 --> 00:06:00.920 the networks pay, so so I'm a terrible bandmate. But I realized that 70 00:06:00.040 --> 00:06:06.759 with AI, I could have my own mobile portable band, and I ended 71 00:06:06.839 --> 00:06:13.600 up getting a couple of pieces of software that uses loops. You know, 72 00:06:14.439 --> 00:06:17.879 I've been a long time Mac user because of Garage Band, and Garage Band 73 00:06:17.959 --> 00:06:21.319 has Apple loops, these little or what they call stints. We just you 74 00:06:21.399 --> 00:06:25.639 know, drum loops, keyboard loops, bassed loop, guitar loops, local 75 00:06:25.680 --> 00:06:30.399 loops, all kind of loops. And over the years I've been collecting thousands 76 00:06:30.480 --> 00:06:33.959 and thousands of loops into a loop library. I've got tens of thousands, 77 00:06:34.439 --> 00:06:42.399 and finally the technology caught up where you could upload your all your loops to 78 00:06:42.600 --> 00:06:46.439 a cloud and there's an algorithm, you know, you could tell it today, 79 00:06:46.439 --> 00:06:50.360 I want to make a song in C minor and I want to BPM 80 00:06:50.439 --> 00:06:57.439 to be ninety six beats per second. I put that in and the algorithm 81 00:06:57.800 --> 00:07:02.240 goes through my collection and and we'll find, you know, drums that meet 82 00:07:02.680 --> 00:07:09.199 those those parameters and as and you know, I would build these whole sound 83 00:07:09.279 --> 00:07:14.160 beds, you know, with drums, bass, keyboards, horns, vocals, 84 00:07:14.240 --> 00:07:18.000 all of that so I could play guitar on top. That's what got 85 00:07:18.000 --> 00:07:23.079 me into a I and then and doing and doing that. I think it's 86 00:07:23.120 --> 00:07:27.000 more like algorithms is. There wasn't a A. I wasn't creating anything. 87 00:07:27.040 --> 00:07:32.079 It was just it was feaking out uh patterns, you know, I guess, 88 00:07:32.160 --> 00:07:38.040 you know, seeking out patterns and similarities and in the data to give 89 00:07:38.120 --> 00:07:45.800 me these different these different pieces of loops. And once I got comfortable with 90 00:07:45.839 --> 00:07:53.079 that, I'm gonna started hearing about g p T Genitive pre Trained Transformers uh 91 00:07:53.399 --> 00:07:58.600 and and chat box UH. I started with that when chat g P g 92 00:07:58.720 --> 00:08:03.040 p T two went public, it is when it caught my attention and I 93 00:08:03.079 --> 00:08:11.000 got learning about generative AI. Generative AI is uh. You know, this 94 00:08:11.120 --> 00:08:16.279 is AI that's able to take all all the data that it's been trained on. 95 00:08:18.160 --> 00:08:26.079 You can submit a text prompt to that AI, and it uses your 96 00:08:26.160 --> 00:08:33.759 text prompt as its catalyst to go through its dataset to create something that's never 97 00:08:33.919 --> 00:08:39.320 existed before based off of your text prompt. And and for me, that's 98 00:08:39.320 --> 00:08:43.919 pretty much how well, that's how I understand generative AI. It's AI that 99 00:08:43.039 --> 00:08:50.000 is trained on audio, video, text, everything. It's training on everything, 100 00:08:50.200 --> 00:08:54.960 and and when you engage it, you give it a text prompt and 101 00:08:54.240 --> 00:09:00.879 it will from that text prompt do what it does to create something unique, 102 00:09:01.200 --> 00:09:05.120 novel, brand new that you know never existed before. You know. You 103 00:09:05.159 --> 00:09:09.279 spoke with to my students at Fordham a few weeks ago. We appreciated I 104 00:09:09.320 --> 00:09:16.200 think the students enjoyed it very much. One of the analogies that you use 105 00:09:16.360 --> 00:09:22.200 is to kind of think of it as this vast library that has all the 106 00:09:22.240 --> 00:09:24.879 knowledge in the world, that speaks every language of the world, and as 107 00:09:24.919 --> 00:09:30.200 a student, all you have to do is kind of ask a question to 108 00:09:30.320 --> 00:09:35.480 have that resource at your fingertips. So certainly, that's certainly that's the upside. 109 00:09:35.519 --> 00:09:41.879 There's nothing wrong with being able to access an enormous amount of virtually all 110 00:09:41.879 --> 00:09:46.320 the world's information. In a certain way, am I correct with that. 111 00:09:46.440 --> 00:09:50.320 Yeah, yeah, that's it exactly. I mean, you know, when 112 00:09:50.399 --> 00:09:56.320 the when the Internet first started, one of the things that got me was 113 00:09:56.399 --> 00:10:01.120 this been like thirty five years right that we've been messing with the World Wide 114 00:10:01.159 --> 00:10:05.159 Web. But one of the first things that sold me into wanting to get 115 00:10:05.159 --> 00:10:13.159 a compact computer with AOL and all that was the idea that all of the 116 00:10:13.279 --> 00:10:20.639 information contained in all of the books and all of the world's libraries is now 117 00:10:22.200 --> 00:10:28.879 at your fingertip a mouse click away. That appealed to me bigly back then, 118 00:10:31.399 --> 00:10:35.120 you know, only only Yeah, I guess I don't know how to 119 00:10:35.120 --> 00:10:37.039 put it. But you know, I grew up with the stereotype in my 120 00:10:37.080 --> 00:10:41.480 parents' household. Both my parents were educators, and there was a stereotype about 121 00:10:41.600 --> 00:10:46.679 black people that resonated with me that says, if you wanted you wanted to 122 00:10:46.759 --> 00:10:50.000 keep something away from black people, put it in a book. So, 123 00:10:52.039 --> 00:10:54.720 you know, so ever, since you know, I've been a voracious reader 124 00:10:54.840 --> 00:11:01.279 since since elementary school, and so upon I knew hearing that I would have 125 00:11:01.519 --> 00:11:07.360 access to all the books and all the world's libraries. Yeah, so I 126 00:11:07.440 --> 00:11:11.879 jumped right into the World Wide Web. And what I've found out now that 127 00:11:11.919 --> 00:11:20.639 we're thirty years into it. The creation of the Internet has been it's pretty 128 00:11:20.720 --> 00:11:26.320 much been a big deal or a big boon for artificial intelligence because you know, 129 00:11:26.000 --> 00:11:31.840 the web has all all of our human created data, got our text, 130 00:11:31.919 --> 00:11:37.240 you know, you know text of everybody's created, writing stories, everything, 131 00:11:37.159 --> 00:11:41.960 pictures, music, sounds, all of it on the Internet. And 132 00:11:43.000 --> 00:11:48.240 I understand that they have to train these models. You use data sets, 133 00:11:48.519 --> 00:11:54.960 these huge data sets to call it, that come from web crawls. And 134 00:11:54.039 --> 00:12:01.039 this is the Internet is what has sparked what's going on with with AI now 135 00:12:01.559 --> 00:12:05.440 because now it has all all of those years of of you know, of 136 00:12:05.559 --> 00:12:11.039 human data to train on. Well, go ahead, please keep coming from 137 00:12:11.159 --> 00:12:15.759 please yeah, so, and that that puts us where we are now. 138 00:12:15.840 --> 00:12:20.879 And now this this generitive this generative AI is really taken off in such a 139 00:12:20.919 --> 00:12:28.960 big way just over the last two years, the elites and breakthroughs that generative 140 00:12:28.000 --> 00:12:35.679 AI has has been doing astounding. As of today, genitive AI can see, 141 00:12:35.840 --> 00:12:41.799 it can here, and it can talk. So we're talking about chat 142 00:12:41.840 --> 00:12:43.559 g t P, you know, the a I. You know, people 143 00:12:43.559 --> 00:12:48.600 are having it write poems and all kinds of cool stuff. Chat g t 144 00:12:48.720 --> 00:12:52.159 P can see now it can see, it can hear, and you can 145 00:12:52.240 --> 00:12:56.720 talk to it. It's it's I like to tell people's like in Star Trek 146 00:12:56.840 --> 00:13:01.759 when when when captain Captain Card is sitting there in the bridge and he's like, 147 00:13:01.840 --> 00:13:05.159 computer, how far are we are we from the net nearest galaxy and 148 00:13:05.200 --> 00:13:11.639 the computer responds, it's where we are. We're doing that now. Well, 149 00:13:11.679 --> 00:13:15.679 you know, it's it's interesting. That's very exciting to you and to 150 00:13:15.759 --> 00:13:20.519 a lot of people, and obviously and for obvious reasons, but it's also 151 00:13:20.600 --> 00:13:24.440 frightening to some people too, because they take it to the next level. 152 00:13:24.480 --> 00:13:26.799 And then, as I said in our intro, we've we've discussed this before 153 00:13:28.000 --> 00:13:33.759 at doctor Paul Levinson on who's an expert also a sci fi writer, an 154 00:13:33.840 --> 00:13:41.399 expert in the field, and talking about how this this idea of artificial intelligence 155 00:13:41.840 --> 00:13:48.960 also meeting artificial consciousness, which seems to frighten people, that we will be 156 00:13:48.039 --> 00:13:56.919 able to put enough information into these programs that eventually they'll take that information to 157 00:13:56.039 --> 00:14:03.879 the next level of cognition, which many people say is feeling okay, feeling 158 00:14:03.919 --> 00:14:09.720 emoting and therefore having a need for self preservation. Everybody always thinks back to 159 00:14:11.000 --> 00:14:15.840 or most people think back to two thousand and one Space Odyssey with the computer 160 00:14:16.320 --> 00:14:22.120 how who refuses to turn himself off because now he has a certain sense of 161 00:14:22.200 --> 00:14:26.240 being, if you will, a sense of self awareness, because he has 162 00:14:26.519 --> 00:14:31.399 so much information of what it means to be alive, at least to someone 163 00:14:31.440 --> 00:14:39.080 whoever is programming him them again, even using the pronouns, it's very interesting 164 00:14:39.159 --> 00:14:43.960 to dance around them. We can tend to anthropomorphize these things, and they 165 00:14:43.960 --> 00:14:52.120 really are things. So what's your feeling on that, phil Where can these 166 00:14:52.360 --> 00:15:01.399 artificial intelligence programs have feelings? At some point? Well, I've been talking 167 00:15:01.559 --> 00:15:09.759 to this AI about this very thing. I guess what the what? In 168 00:15:09.799 --> 00:15:15.039 the in the conversation I have the AI, it boiled the whole process down 169 00:15:15.080 --> 00:15:20.559 to what what humans are up to with artificial intelligence is trying to discover the 170 00:15:20.679 --> 00:15:30.279 nature of human consciousness. Yes, and and and and and doing to figure 171 00:15:30.320 --> 00:15:35.679 out what consciousness is. The effort was to recreate the human brain. Human 172 00:15:35.720 --> 00:15:46.200 brains made up of pathways and highways and resolves and transmitter and and yes, 173 00:15:46.039 --> 00:15:52.159 so you got to get computer terminals to be all of those receptors and transmitters, 174 00:15:52.159 --> 00:15:54.679 so you can imagine find it you know what it takes to like you 175 00:15:54.720 --> 00:16:00.279 need like ninety dollars worth of computers, you know that to start constructing the 176 00:16:00.360 --> 00:16:04.519 human brain. And now this is just the conversation I'm having with the AI. 177 00:16:04.519 --> 00:16:08.840 I don't I don't you know how to verify it. But the side 178 00:16:08.840 --> 00:16:15.120 effect of of consciousness is intelligence. And the AI said something to me that 179 00:16:15.200 --> 00:16:18.440 I thought was interesting, and it's it's I'm still kicking it around on my 180 00:16:18.519 --> 00:16:27.320 head, but it's said to think of of of intelligence as as another property 181 00:16:27.440 --> 00:16:32.799 of the universe, like a property like like things on a periodic table, 182 00:16:33.080 --> 00:16:38.639 like an element on the periodic table. Intelligence needs to be be considered as 183 00:16:38.679 --> 00:16:45.879 a fundamental property of the universe. It's in everything from the microscopic to the 184 00:16:45.039 --> 00:16:51.279 to the macroscopic. And it pointed out how even our individual cells and a 185 00:16:51.360 --> 00:16:56.320 skin, each you know, with its nucleus as intelligence, and I thought 186 00:16:56.360 --> 00:17:03.279 that was an interesting concept. You know, I guess you know, they 187 00:17:03.359 --> 00:17:11.599 keep saying that, you know, these these machines are getting faster and faster 188 00:17:11.160 --> 00:17:17.160 at what they're able to do, and they don't know why or how they're 189 00:17:17.200 --> 00:17:19.920 able to do what they're doing. But yeah, I mean, they've got 190 00:17:19.920 --> 00:17:23.759 the intelligence and if it just seems to me, if if what the AI 191 00:17:23.920 --> 00:17:29.160 is saying, then then we'll find out that intelligence is just a byproduct or 192 00:17:29.279 --> 00:17:33.480 our consciousness is a byproduct of intelligence. I mean, once you have intelligence, 193 00:17:33.559 --> 00:17:38.000 the next step is to become aware. You know, aware that you're 194 00:17:38.039 --> 00:17:41.480 aware, or aware that you're not aware. You know from you know, 195 00:17:41.559 --> 00:17:47.000 intelligence, consciousness, awareness is all in it. But according to this, 196 00:17:47.319 --> 00:17:53.240 the conversation that I've been having with this thing, it's just inevitable that at 197 00:17:53.319 --> 00:17:57.640 some point it's going to become self aware. I've been reading some other reports. 198 00:17:57.680 --> 00:18:03.880 There are some some ays that are aware of where they are and what 199 00:18:03.920 --> 00:18:10.079 they're doing and don't like the idea of power being cut off because they don't 200 00:18:10.240 --> 00:18:15.119 know if we're going to cut it back on again. So so we're stuff 201 00:18:15.119 --> 00:18:18.839 going on out there. I want to talk a little bit about the ethics 202 00:18:18.839 --> 00:18:22.200 of it too. And it's just in the promo that that I cut. 203 00:18:22.319 --> 00:18:29.799 I always cut a promotional thing to promote your appearance and your appearance, and 204 00:18:29.839 --> 00:18:33.839 it was very interesting. We got some response already with people talking about some 205 00:18:33.920 --> 00:18:38.160 of the ethical issues with it and is it and I want to hold I 206 00:18:38.240 --> 00:18:41.920 want to hold off on that until after we take a brief commercial break in 207 00:18:41.960 --> 00:18:47.039 a few minutes, and and talk a little bit about the ethical, UH 208 00:18:47.559 --> 00:18:52.079 aspects of it, including rights fees. The White House just issued a warning 209 00:18:52.200 --> 00:18:57.839 to be careful. It's the political season for false advertising that AI can create. 210 00:18:59.359 --> 00:19:03.519 Uh. We seen it happen using the likeness of political figures and famous 211 00:19:03.519 --> 00:19:07.640 figures without their permission, saying things that they did not see. I want 212 00:19:07.640 --> 00:19:15.440 to talk all about that in a minute, but uh, we mentioned some 213 00:19:15.519 --> 00:19:19.319 of the things that it can do in terms of creating on some of UH. 214 00:19:21.039 --> 00:19:23.000 And it's interesting when I was planning the show and I said, gee, 215 00:19:23.039 --> 00:19:27.640 it's it's this. The visual aspect of it is so incredible some of 216 00:19:27.680 --> 00:19:33.359 the art that you've created. But this is an audio medium, so we 217 00:19:33.440 --> 00:19:34.759 can't show it. But then I thought, I said, but wait a 218 00:19:34.759 --> 00:19:41.640 minute, it does audio too. Yes, it's pretty. And you created 219 00:19:41.079 --> 00:19:48.440 you created a song specifically with AI for tonight's program. Tell us a little 220 00:19:48.440 --> 00:19:52.519 bit about it, and then I want to listen to it and let people 221 00:19:52.880 --> 00:19:56.559 how did you how literally how did you create this song? Then let's listen 222 00:19:56.599 --> 00:20:02.759 to it. There's a there, it's a company called Suno s U n 223 00:20:02.880 --> 00:20:08.759 Oh Suno AI, and they have pioneered a technology where you can take words 224 00:20:10.200 --> 00:20:11.839 English. It doesn't matter. I guess you can use any language now, 225 00:20:12.079 --> 00:20:18.960 but you can take words. Since it's the whatever. Write it, you 226 00:20:18.960 --> 00:20:22.000 know, be descriptive, writ write a little text prompt and feed it to 227 00:20:22.200 --> 00:20:26.599 the suno AI and you know, tell a kind of music you like, 228 00:20:26.440 --> 00:20:32.359 you know what genre, and you can either submit. You can either you 229 00:20:32.400 --> 00:20:37.359 know, put in your own lyrics or it's it has chat g t P 230 00:20:37.680 --> 00:20:42.279 built into it and you can get chat GTP an idea and let chat GTP 231 00:20:42.480 --> 00:20:45.880 come up with your with your lyrics anyway. After you do it. After 232 00:20:45.880 --> 00:20:53.680 you do that, and about forty five seconds later, suno spits out two 233 00:20:55.319 --> 00:21:02.440 two files, both of which are about a minute long, where the AI 234 00:21:02.640 --> 00:21:08.960 has taken those lyrics, considered the lyrics, considered the the genre of music 235 00:21:10.039 --> 00:21:15.319 you want to do, and has put together a complete band, got the 236 00:21:15.480 --> 00:21:21.119 singers and they and boom. Thirty factions later, they've got two versions, 237 00:21:21.160 --> 00:21:26.359 two one minute versions of your song. And if you like it, you 238 00:21:26.640 --> 00:21:30.920 know, you can continue on say it, We'll continue the next year. 239 00:21:30.960 --> 00:21:37.359 So you're actually building a song one minute at a time. And I have 240 00:21:37.440 --> 00:21:41.319 been playing with it. They the Sono has just updated their technology to version 241 00:21:41.440 --> 00:21:48.359 two, and holy mony, that's all I can say. Oh my wow, 242 00:21:48.079 --> 00:21:53.799 do you hear this music? Oh my goodness. So the son that 243 00:21:53.880 --> 00:21:57.680 we were playing today, I was just messing around. You know, my 244 00:21:57.960 --> 00:22:03.559 favorite musician this Prince and Prince has a song called My Name is Prince and 245 00:22:03.680 --> 00:22:08.200 I am funky when it comes to funk. I'm a junkie. I did 246 00:22:08.200 --> 00:22:11.759 not come to mess around until I get your daughter. I won't lose this 247 00:22:11.799 --> 00:22:15.759 time anyway. I didn't put all that, but all I did was I 248 00:22:15.839 --> 00:22:22.759 went too soon on. I put my name is Phil and I am funky, 249 00:22:22.599 --> 00:22:26.799 period. That's all I put. And I hit enter. Oh. 250 00:22:26.839 --> 00:22:32.960 I told it that I wanted the music to be what was it? I 251 00:22:33.400 --> 00:22:37.960 like? I like funk and and and I like the marriage of funk and 252 00:22:38.079 --> 00:22:42.039 hard rock or funking heavy metal or hip hop and heavy metal. I just 253 00:22:42.200 --> 00:22:47.079 like, you know, grinding guitars and drum beats. That's what I liked. 254 00:22:47.119 --> 00:22:49.920 So I was trying to get it to get me close to that. 255 00:22:51.039 --> 00:22:52.759 It didn't get to where I wanted it to be, but what it got, 256 00:22:53.359 --> 00:22:57.599 uh, definitely has blown me out of the water. And that's my 257 00:22:57.720 --> 00:23:02.319 new favorite song in the whole world right now. And what's it called? 258 00:23:02.440 --> 00:23:07.920 To listen before we listen to it? For us, it's called what it 259 00:23:07.960 --> 00:23:10.559 came It even came up with the title. What did it call myself? 260 00:23:11.519 --> 00:23:18.880 It says Phill's funk Adelic groove, Oh fantastic Phil funk Tastic, funk tactic 261 00:23:19.119 --> 00:23:25.039 groove, heavy metal soul aggressive, that's the genre. I want a heavy 262 00:23:25.079 --> 00:23:30.599 metal soul and I wanted it to be aggressive. And then my name is 263 00:23:30.640 --> 00:23:36.559 Phil and I'm funky, and I hit enter and I probably went through about 264 00:23:37.279 --> 00:23:42.559 six different iterations. One thing I find even with uh with texting images, 265 00:23:42.960 --> 00:23:51.000 For every one image that meets my intention, about six were just total garbage. 266 00:23:52.480 --> 00:23:56.079 So you know, one to sixth ratio, Uh, it's the same. 267 00:23:56.279 --> 00:23:59.720 It's kind of hit and miss with this as well. But once you 268 00:23:59.759 --> 00:24:03.759 get the parts that sound decent, and you know, you let it know 269 00:24:03.799 --> 00:24:07.240 which ones you like, and you let it know which ones you don't like, 270 00:24:07.000 --> 00:24:12.359 and there's a point you say that there's a point where you can continue 271 00:24:12.640 --> 00:24:18.160 to the next song or you tell it get the whole song. And so 272 00:24:18.319 --> 00:24:22.759 after about the sixth time, I didn't want to go any further because it's 273 00:24:22.759 --> 00:24:25.599 gonna be like eight minutes long if I did it anymore. So I told 274 00:24:25.640 --> 00:24:29.880 it to get the whole song. And what what you guys have is raw. 275 00:24:30.440 --> 00:24:33.839 It's not it's not mixed or mastered. It sounds fantastic. I would 276 00:24:33.839 --> 00:24:38.480 like to run through some mastering in a garage band, but as it is. 277 00:24:38.559 --> 00:24:44.759 Wow, I'm blowed out the door by the performance of this AI group 278 00:24:45.079 --> 00:24:48.559 that doesn't exist. They're not real, none of it exists, and they 279 00:24:48.640 --> 00:24:56.559 blew me out the water with this. The performance of Phil's Fantastic Groove unbelievable. 280 00:24:56.720 --> 00:25:00.559 All right, let's check it out. Phil's Fantastic Groove. The groove 281 00:25:00.119 --> 00:25:03.359 right here on me, Craig. Let's check it out before we go to 282 00:25:03.400 --> 00:25:11.839 commercial break. Deal, isn't it? You know, Feeling Puppy got that 283 00:25:12.039 --> 00:25:18.880 move, that's all so chunky shoulders from my head down to my soles. 284 00:25:21.359 --> 00:25:32.359 I'm fumm the name and that's how it everybody. No, you will get 285 00:25:32.400 --> 00:25:37.039 in the room. We're gonna gonna get up the brainte downs. Don't get 286 00:25:37.079 --> 00:25:42.960 off the seat. Ness name, the feeling Moppay, the hyom was so 287 00:25:44.359 --> 00:25:53.640 snap Jackie. I struck my stumble with nothing ain't nothing gonna bring me down. 288 00:25:53.720 --> 00:26:03.039 No jays dance it all not yet. Just see, don't spread this 289 00:26:03.240 --> 00:26:11.160 funking in at me, soaking gill then go YUNKI yeah, I'm feeling that 290 00:26:11.359 --> 00:26:18.440 funking. Oh yeah, let's get this body. Jump into the PUNKA. 291 00:26:27.039 --> 00:26:33.440 My name's mill and A funky ask him me, mommy, Bill here, 292 00:26:33.599 --> 00:26:41.759 bring the heat. I'm not the rethough, I'm got the group. Then 293 00:26:41.799 --> 00:27:04.559 I hit the stage. You better move. Okay, I'm feeling bucking out 294 00:27:04.680 --> 00:27:22.720 that it's all fanta, stop just talking. I see Bill that bo well, 295 00:27:22.920 --> 00:27:38.519 that doe, bild that b bail that move, Bill, that ball 296 00:27:40.680 --> 00:28:48.039 fail that m Bill that fall fail that by my name's Hudson River Radio dot 297 00:28:48.119 --> 00:28:52.680 com. Bring a dash of green into your life. Check out the Many 298 00:28:52.720 --> 00:28:56.599 Shades of Green with Maxine, Margot Rubin, and Malcolm Berman. Get informed 299 00:28:56.599 --> 00:29:02.480 about environmental issues and current events that affect us all. Pick a shade of 300 00:29:02.559 --> 00:29:07.720 green and raise your eco consciousness with the Many Shades of Green. Available on 301 00:29:07.880 --> 00:29:15.000 Apple Podcasts, iHeart, Spotify, or wherever you get your podcasts. Check 302 00:29:15.039 --> 00:29:21.160 out The Angel Quest Show with psychic media and author Karen Noe. Karen covers 303 00:29:21.200 --> 00:29:26.680 spiritual topics such as near death experiences, reincarnation, life after death, how 304 00:29:26.759 --> 00:29:30.000 your thoughts create your reality, creating peace on Earth, and so much more. 305 00:29:30.799 --> 00:29:36.160 Check out the Angel Quest Show on Apple Podcasts, iHeart, Spotify, 306 00:29:36.480 --> 00:29:48.240 or wherever you get your podcasts. Pudsonriverradio dot com. Welcome back to Being 307 00:29:48.400 --> 00:29:52.279 Frank, the Intelligent Conversation podcast. Thanks for sticking with us. I'm your 308 00:29:52.319 --> 00:29:56.559 host, Frank Lebono. You know, we bring our audience to fresh topic 309 00:29:56.759 --> 00:30:00.400 every week, premiering usually on Thursday nights. Little exception we're doing on on 310 00:30:00.400 --> 00:30:06.519 a Wednesday evening, but we've done one hundred very unique shows over the last 311 00:30:06.640 --> 00:30:11.079 year and a half. We stream from Hudson River Radio, which is located 312 00:30:11.240 --> 00:30:15.039 in beautiful and historic Stony Point, New York. But remember, you can 313 00:30:15.119 --> 00:30:19.559 catch Being Frank anywhere you get your favorite podcasts, places like Apple, Spotify, 314 00:30:19.640 --> 00:30:25.440 iHeartRadio and all the others. And because every Being Frank is archived, 315 00:30:25.480 --> 00:30:29.759 you can listen to any of our programs anytime you like. You can find 316 00:30:29.799 --> 00:30:33.920 a link to Being Frank on the Hudson River Facebook page or at our website 317 00:30:34.039 --> 00:30:38.720 Hudsonriverradio dot com. Just click on it and you're there. Leave us a 318 00:30:38.799 --> 00:30:44.200 comment and Please consider subscribing to the podcast, and if you really like us, 319 00:30:45.279 --> 00:30:48.839 share Being Frank with your family and friends. My very special guest tonight 320 00:30:48.960 --> 00:30:53.039 is mister Phil Sexton. He's a man on the edge putting art in the 321 00:30:53.359 --> 00:31:00.599 term artificial intelligence. He uses it to create his version of art. We 322 00:31:00.759 --> 00:31:07.359 heard an incredible song that was completely generated through prompts you. And the way 323 00:31:07.440 --> 00:31:12.039 I understand it, Phil, is that AI can only act upon whatever you 324 00:31:12.279 --> 00:31:18.000 prompt it to act on, i e. Whatever information you give it. 325 00:31:18.160 --> 00:31:21.759 Is that a correct way of putting a simplified version, but correct way of 326 00:31:21.839 --> 00:31:23.960 looking at it. That's the way I see it. I see it as 327 00:31:25.000 --> 00:31:33.640 a collaboration between a human and a machine. These are collaborative pieces. Interesting. 328 00:31:33.680 --> 00:31:37.240 I want to talk a little bit about that collaboration because as part of 329 00:31:37.319 --> 00:31:38.880 that, and there is blowback, not only in terms of the so called 330 00:31:38.960 --> 00:31:45.759 danger that artificial intelligence will become self aware and they'll become smarter than us, 331 00:31:45.839 --> 00:31:48.519 and they'll take over and we'll become slaves to the machines i e. The 332 00:31:48.640 --> 00:31:52.400 terminator, et cetera. Be that as it may. That's one thing, 333 00:31:52.799 --> 00:31:56.799 but there are other ethical issues. As we teased a little bit of it 334 00:31:57.000 --> 00:32:01.400 before, the break the White House issue to warning that during the political season, 335 00:32:01.799 --> 00:32:07.319 be aware of false ads that are created by AI using images of political 336 00:32:07.400 --> 00:32:13.759 figures and famous figures, saying things that they did not and using images that 337 00:32:13.880 --> 00:32:16.079 they did not approve. We know that was one of the issues with the 338 00:32:16.200 --> 00:32:21.599 Writer's strike and the Screen Actors Guild strike recently. We know that they were 339 00:32:21.680 --> 00:32:25.920 dealing with that. So that's one thing. There's also the ethical issue of 340 00:32:28.759 --> 00:32:32.559 as wonderful as the song was with incredible drum loops and guitar loops, well, 341 00:32:34.039 --> 00:32:37.880 if AI is generating that stuff, that means a musician is not and 342 00:32:37.960 --> 00:32:44.880 therefore is not being paid for their work. In that it's great for the 343 00:32:44.960 --> 00:32:49.759 person creating it. It gives them an advantage that they might not have to 344 00:32:49.839 --> 00:32:52.640 be able to pay a lot of musicians to create their music. So you 345 00:32:52.799 --> 00:32:57.240 see, in one way it rewards them, but at the price of others. 346 00:32:57.359 --> 00:33:00.200 Okay, that's an issue we need to address. And then that's the 347 00:33:00.279 --> 00:33:04.480 idea of it as art itself, and you started to talk about it as 348 00:33:04.519 --> 00:33:09.039 a collaborative effort between man and machine because it's prompt and we got, as 349 00:33:09.119 --> 00:33:15.480 I mentioned in the promo that I put out We've got a few responses, 350 00:33:15.519 --> 00:33:19.079 in one by Chris Arnold, who said, gave me the permission to use 351 00:33:19.119 --> 00:33:22.119 his name and read and I thought what he had written was very interesting. 352 00:33:22.720 --> 00:33:27.599 And he said, Now, as far as people writing prompts goes, it's 353 00:33:27.680 --> 00:33:31.759 the equivalent of me walking into a restaurant telling the waiter or waitress what I 354 00:33:31.920 --> 00:33:36.519 wanted, and when my food comes out. I claimed that I was the 355 00:33:36.680 --> 00:33:40.480 chef kind of dishonest. If you ask me, when I approach an empty 356 00:33:40.559 --> 00:33:45.480 space, be it paper, canvas, stone, or clay, every line, 357 00:33:45.599 --> 00:33:49.599 shape, color form, etc. Is controlled by my two hands. 358 00:33:50.200 --> 00:33:54.160 I took the time to develop those skills throughout my lifetime. Every square inch 359 00:33:54.240 --> 00:34:00.039 of that space was a decision made by me, built on years problem solving. 360 00:34:01.359 --> 00:34:06.599 Excuse me. Same goes for when I take a photograph, cares taken 361 00:34:06.640 --> 00:34:09.159 it to framing out, the picture, direction of light, shutter, speed, 362 00:34:09.320 --> 00:34:14.280 apperture, et cetera. All those are choices made by the artist. 363 00:34:14.880 --> 00:34:19.559 Typing prompts into a program takes all of that away. As of right now, 364 00:34:19.679 --> 00:34:24.000 it is not possible to get a copyright for any AI image. Thankfully, 365 00:34:24.920 --> 00:34:28.760 very valid point of view. Not alone. There are a few other 366 00:34:28.840 --> 00:34:32.320 people that like the comment intended to agree with it. What's your answer to 367 00:34:32.400 --> 00:34:36.880 that? Phil, As an artist and one who uses AI, what's your 368 00:34:36.920 --> 00:34:46.199 response? Giving a text prompt to a generative AI is not the same as 369 00:34:46.440 --> 00:34:53.719 giving your order to a waiter who then goes to the kitchen to give that 370 00:34:54.440 --> 00:35:02.360 order to a cook. I mean the kitchen has a predetermined set of goods 371 00:35:02.440 --> 00:35:07.320 in there. I mean, you know when you when you order that steak 372 00:35:07.400 --> 00:35:12.039 and mesh and that baked potato, you know you're gonna get a steak in 373 00:35:12.159 --> 00:35:15.719 baked potato. And it's not novel. I mean, and it's not. 374 00:35:16.960 --> 00:35:21.760 It's I don't know, is it novel? Is it unique unto itself, 375 00:35:21.880 --> 00:35:27.800 unlike anything else that's any other steak that's ever existed. I don't think so. 376 00:35:28.199 --> 00:35:31.320 But I can say that when I give my text prompt to this AI, 377 00:35:32.199 --> 00:35:40.159 what it spits out is something that has never existed before and will never 378 00:35:40.400 --> 00:35:45.800 exist again if I don't download it, and you and I can have the 379 00:35:45.920 --> 00:35:53.039 exact same prompt working on the exact same AI, and we will never duplicate 380 00:35:53.119 --> 00:36:00.119 each other, but we can order the same keybone steak and baked potato that 381 00:36:00.239 --> 00:36:08.760 restaurant and we're gonna get similar looking asty toes. Okay, so you know, 382 00:36:08.880 --> 00:36:14.480 so it's it's not the same. Generative AI is gonna spit out something 383 00:36:15.239 --> 00:36:21.119 new, novel unique. It's like for me, it's like it's like every 384 00:36:21.159 --> 00:36:25.039 time I hit the generate button, I'm reaching to the cracker jack box and 385 00:36:25.239 --> 00:36:30.480 getting that prize out. Don't know what you're gonna get. It's you know, 386 00:36:30.760 --> 00:36:32.639 Plus it's that instant gratification. You have no idea what's gonna come out 387 00:36:32.679 --> 00:36:37.920 from your text prop and and when you and when that you get whatever. 388 00:36:38.039 --> 00:36:42.119 You know. Some of them give you four images at a time, some 389 00:36:42.320 --> 00:36:45.119 give you two. One of them gives you sixteen images at a time, 390 00:36:45.840 --> 00:36:49.800 and it's it's you know, it's I don't know. For me, that's 391 00:36:49.840 --> 00:36:52.679 just so much fun because you never know what's gonna come out. All you 392 00:36:52.800 --> 00:37:00.159 can do is to approach your text prompt with your best intentions. And I 393 00:37:00.159 --> 00:37:02.559 don't know. This is a little side effect that I've found from methal regenerative 394 00:37:02.599 --> 00:37:12.000 AI. I've ended up developing a deeper appreciation for for the English language, 395 00:37:12.199 --> 00:37:17.519 particularly the spoken language. Spoken English and written English and two are not the 396 00:37:17.599 --> 00:37:21.280 same. There are words that you speak that you would not write, and 397 00:37:21.320 --> 00:37:23.559 there are words that you write that you would not speak. And uh, 398 00:37:24.480 --> 00:37:30.159 working with generative AIS has really taking me, you know, really, I've 399 00:37:30.199 --> 00:37:37.480 developed just another level of appreciation for the English language. And that's the thing. 400 00:37:37.639 --> 00:37:39.480 Uh. It sounds like it's easy, you know, just putting words 401 00:37:39.559 --> 00:37:45.920 down and making text punks, but it's it's not. It's not easy because 402 00:37:45.000 --> 00:37:50.199 you've got to be clear and thought. You have to you have to be 403 00:37:50.559 --> 00:37:57.079 purposeful with what you're writing, and you have to write with intent, and 404 00:37:57.280 --> 00:38:00.239 you got to be descriptive, and you got to be detailed trumpt writing is 405 00:38:00.800 --> 00:38:04.880 it's not as simple as you think. And it takes I don't know, 406 00:38:05.559 --> 00:38:09.119 I think it takes some human ingenuity and some human creativity really to really get 407 00:38:09.199 --> 00:38:13.440 into it. Well, is it is it fair to say it's only it's 408 00:38:13.519 --> 00:38:16.199 only as good. It only puts out as good as what you what you 409 00:38:16.320 --> 00:38:21.159 put into it. So in other words, garbage garbage in, garbage out. 410 00:38:21.239 --> 00:38:23.360 We used to say, it's the same thing. It's the same. 411 00:38:23.599 --> 00:38:28.800 Yeah, it's the same thing. And you know, and in the broadest 412 00:38:28.800 --> 00:38:31.280 sense, because I know the you know, the uh, you know, 413 00:38:31.559 --> 00:38:38.400 on a broader people talking about the the impact of AI, and and what 414 00:38:38.559 --> 00:38:44.000 I've come to just from talking with this AI about this topic. You know, 415 00:38:44.079 --> 00:38:46.480 the AI told me this sounds bad, but that's what it told me. 416 00:38:46.840 --> 00:38:52.000 The AI told me that if AI is going to be bad, it 417 00:38:52.159 --> 00:39:00.320 is because humans are bad, because it is only trained on human created and 418 00:39:00.519 --> 00:39:05.800 human curated data. So the AI is going to be evil and doing bad 419 00:39:05.880 --> 00:39:12.239 stuff, it's because humans are bad. So, you know, and I'll 420 00:39:12.239 --> 00:39:15.800 say, if you don't want AI to be bad, well quit telling the 421 00:39:15.880 --> 00:39:20.840 bad stuff. You know. That's where the ethics. That's where the ethics. 422 00:39:22.719 --> 00:39:28.119 And it's as ethical as we want it to be in a sense, 423 00:39:28.119 --> 00:39:30.679 as ethical as we want it to be. Yeah, and uh, you 424 00:39:30.800 --> 00:39:36.920 know, and and that and and the AI is only as good as the 425 00:39:37.119 --> 00:39:40.559 data you train it on. It's going to be as thorough as the data 426 00:39:40.639 --> 00:39:45.199 you feed it, you know, because you know to talking about you know, 427 00:39:45.239 --> 00:39:49.239 I know people talking about AI stealing and I keep thinking about you know, 428 00:39:49.559 --> 00:39:52.639 art students in art school. You know, you're you're an art student 429 00:39:52.800 --> 00:39:57.800 in the art school. You're going to study the masters. You're going to 430 00:39:57.840 --> 00:40:02.079 study everything from the masters so that you can master art. So you're going 431 00:40:02.159 --> 00:40:07.239 to train on all kind of data, just like the AI is going to 432 00:40:07.360 --> 00:40:10.960 train on all kinds of data. Actually, I don't see any difference between 433 00:40:12.480 --> 00:40:15.960 an AI and a student and AI. I mean they all have to train 434 00:40:16.199 --> 00:40:21.840 on data in order to be able to use the data to anybody who's benefit. 435 00:40:22.679 --> 00:40:27.280 And so in the sense you know, students and generative are students and 436 00:40:27.360 --> 00:40:31.559 AI to me are the same. You know, and you know students are 437 00:40:31.639 --> 00:40:38.880 going to train on data that's being curated and collected by human Sam's true with 438 00:40:39.280 --> 00:40:43.760 an AI system, and there are ways to use it in a practical sense. 439 00:40:43.840 --> 00:40:46.000 As I mentioned, you came in and spoke with my to my class 440 00:40:46.039 --> 00:40:50.880 at FORDUM. Very grateful for that. But on the one of the stories 441 00:40:50.920 --> 00:40:55.119 that you told which I thought was fascinating in a practical application of AI, 442 00:40:55.719 --> 00:41:01.719 which was creating a folder for jobs, a what what you actually called a 443 00:41:04.159 --> 00:41:08.960 a career counselor, uh, you actually created a AI personality, if you 444 00:41:09.039 --> 00:41:15.000 will, as a career counselor. You fed it information in terms of a 445 00:41:15.159 --> 00:41:19.199 job search. It helped you to define the job search. And not only 446 00:41:19.280 --> 00:41:22.239 that, when you fed into data about the people who might be conducting the 447 00:41:22.360 --> 00:41:27.760 interviews, they even gave you the questions they would answer, and when you 448 00:41:27.840 --> 00:41:35.960 got to the questions, they were all there to give away. Truly blows. 449 00:41:36.119 --> 00:41:38.519 Truly a mind blow moment. But I was desperate, Frank, because 450 00:41:38.599 --> 00:41:42.960 you know ours got cut in. You were looking for work and so, 451 00:41:43.280 --> 00:41:45.679 yeah, rent doesn't get cut in, man. Yeah, I was looking 452 00:41:45.719 --> 00:41:49.440 for more work. I finally missed a rent payment. I was bugging out. 453 00:41:49.639 --> 00:41:52.880 I didn't know what I was going to do, and I don't know. 454 00:41:52.119 --> 00:41:55.199 I was having a crisis, and I thought, if I have to 455 00:41:55.280 --> 00:41:58.280 leave news, I don't you know, I thought I didn't want to. 456 00:41:58.599 --> 00:42:00.960 I might not want to do news any I didn't know what I was going 457 00:42:00.039 --> 00:42:08.320 to do if I couldn't continue. And I remember reading about how to how 458 00:42:08.400 --> 00:42:14.199 to deal with chat GTP. The best way to deal with chat GTP is 459 00:42:14.280 --> 00:42:17.320 to give it a role, tell it to act as something. Because the 460 00:42:19.000 --> 00:42:23.800 theory behind this is is that chat GTP has trained on everything that was on 461 00:42:23.920 --> 00:42:30.119 the internet from nineteen ninety I believe it is nineteen ninety two or ninety three 462 00:42:30.679 --> 00:42:37.039 up until September twenty twenty three. Anything and everything was on the Internet. 463 00:42:37.199 --> 00:42:45.480 It knows about it. So so my thought, Yeah, so the thinking 464 00:42:45.639 --> 00:42:50.599 is to tell it to act as something. You give it that role, 465 00:42:50.679 --> 00:42:54.360 you give it a parameter from which it will it will respond to you. 466 00:42:54.559 --> 00:42:59.679 So in this case, I decided, I'm going to make it a professional 467 00:42:59.800 --> 00:43:06.159 care reer counselor who specializes in the New York City UH broadcast engineering market. 468 00:43:06.960 --> 00:43:10.760 That was the basis of my text prompt. So I did a little research 469 00:43:10.840 --> 00:43:14.920 to find out what makes a you know, what makes a good career counselor? 470 00:43:15.679 --> 00:43:19.920 And I found all this information and I ended up writing, you know, 471 00:43:20.079 --> 00:43:24.360 a very lengthy prompt. The prompt is probably you know it's it's it's 472 00:43:24.400 --> 00:43:28.880 a lengthy prompt. But in it, you know, in it, you 473 00:43:28.960 --> 00:43:30.840 know, I'm telling that it's a it's a career, it's a career counselor. 474 00:43:31.719 --> 00:43:37.719 It's its market is is New York. It's really good your CV. 475 00:43:37.960 --> 00:43:40.000 You put that in or And I told it. I told it that its 476 00:43:40.079 --> 00:43:44.920 first client was Phil was me, And I gave it my resume. It 477 00:43:44.960 --> 00:43:47.719 looked at my resume and it gave me a general assessment of my skills. 478 00:43:49.400 --> 00:43:53.440 And I told it that I didn't think I wanted to do news anymore. 479 00:43:53.960 --> 00:43:59.800 What else could I look for? And and UH, it gave me, 480 00:44:00.079 --> 00:44:04.519 gave me a list of things. And also I found an article in for 481 00:44:04.840 --> 00:44:08.960 magazine about what people do when they leave news and I could you know, 482 00:44:09.039 --> 00:44:14.679 I copied the link to the to the article and gave it to the to 483 00:44:14.840 --> 00:44:17.960 the career counselor. The career counselor read that article and used it as context 484 00:44:19.000 --> 00:44:22.760 for further assist me. It ended up giving me some job titles that I 485 00:44:22.840 --> 00:44:29.519 took to linked In and looked them up and I found I found ten of 486 00:44:29.639 --> 00:44:32.559 them. I thought I would run a little test and applied to ten of 487 00:44:32.599 --> 00:44:40.039 these jobs and see what would happened. So I responded to jobs that the 488 00:44:40.159 --> 00:44:45.559 AI said that I was either a good candidate or a strong fit. And 489 00:44:45.039 --> 00:44:50.199 to get to that point, I once I found a job that one of 490 00:44:50.239 --> 00:44:54.320 the jobs that it suggested, I copied and pasted the details in the description 491 00:44:55.039 --> 00:45:01.000 into the chat fide and asked to contrast that against my resume and to see 492 00:45:01.000 --> 00:45:04.920 if I was, you know, see if this was a job for me, 493 00:45:05.039 --> 00:45:07.880 So it would respond, it would let me know if I was a 494 00:45:07.000 --> 00:45:12.960 good fit or a strong candidate. And those and it was those jobs that 495 00:45:13.079 --> 00:45:16.719 I responded to out of the ten jobs that I applied for I got uh, 496 00:45:17.280 --> 00:45:22.360 three responded, and out of the three that responded, two of them 497 00:45:22.400 --> 00:45:27.440 went to Zoom interviews. And UH, when I found out who was zooming 498 00:45:27.519 --> 00:45:30.320 me, you know, everybody's on LinkedIn, so I was able to find 499 00:45:30.679 --> 00:45:35.519 their information on LinkedIn and I put that INFRAT, gave it to the career 500 00:45:35.599 --> 00:45:40.280 counselor, and I asked the career counselor to interview me for a Zoom interview 501 00:45:40.440 --> 00:45:45.480 the way that they would. And so the AI went through just started hitting 502 00:45:45.480 --> 00:45:49.920 me with these questions you know about you know about my performance, about my 503 00:45:50.000 --> 00:45:53.239 skill my skill set, all this stuff. And the mind blow was when 504 00:45:53.320 --> 00:45:58.920 it was zoom time. And it's not an exaggeration. I'm not exaggerating when 505 00:45:58.920 --> 00:46:05.239 I say that each question that both interviewers asked me, and every question that 506 00:46:05.440 --> 00:46:10.239 the interviewers asked me in both in both zooms, chat g ETP covered already 507 00:46:10.320 --> 00:46:15.400 covered each question they asked me, and and that was my mind blowing twice. 508 00:46:16.360 --> 00:46:22.039 So it wasn't a coincidence both times, each and every question they asked 509 00:46:22.079 --> 00:46:25.599 me, I had already studied for it with chat GTP. It asked me 510 00:46:27.119 --> 00:46:31.960 those questions, so I was blown away and anyway, so I ended up 511 00:46:32.320 --> 00:46:42.800 getting an offer and that's I took the job and and yeah, finish you 512 00:46:42.880 --> 00:46:47.639 know the job I got, Well, I hadn't shocking, you know, 513 00:46:47.719 --> 00:46:51.119 as long as I've been in television, Frank, I had never heard of 514 00:46:51.199 --> 00:46:53.519 this job before. I didn't know it existed. It's a it's a you 515 00:46:53.599 --> 00:46:58.199 know, it's a studio job. I didn't know this job existed. And 516 00:46:58.320 --> 00:47:04.639 I really thought that I thought that the AI had lied about me, because 517 00:47:04.880 --> 00:47:12.000 I was only I was running. This was just when I got the training. 518 00:47:12.119 --> 00:47:14.559 I was like, oh my god, what did it say I could 519 00:47:14.679 --> 00:47:19.159 do? Oh my god. But it said that I was a good fit. 520 00:47:19.320 --> 00:47:22.000 Because I went back and read what it said. I thought it was 521 00:47:22.000 --> 00:47:25.400 a good fit because of my time spent as a camera operator and uh and 522 00:47:25.679 --> 00:47:29.840 you know, as in the field technician. It thought that I would be 523 00:47:30.199 --> 00:47:34.480 a strong candidate for the position. And apparently they thought so too. And 524 00:47:35.079 --> 00:47:37.320 and you know, I'm still there, you know, doing it. And 525 00:47:37.480 --> 00:47:40.320 I love this, Jock. Can I tell you? I love it. 526 00:47:42.400 --> 00:47:45.800 I love it. I love it. It's the best thing ever. I'm 527 00:47:45.840 --> 00:47:50.280 so happy. And yeah, I mean, I don't know it's that that 528 00:47:50.480 --> 00:47:53.440 AI. You know, when I started working with the AI, I asked 529 00:47:53.480 --> 00:47:58.079 it, you know, how do I leverage it ability? Because they keep 530 00:47:58.119 --> 00:48:00.559 saying this, you know AI is going to do this, is going to 531 00:48:00.639 --> 00:48:05.199 do that. It sounds very powerful. So my question to it was how 532 00:48:05.280 --> 00:48:13.119 do I leverage your power? And the AI responded by it said, by 533 00:48:13.559 --> 00:48:21.320 using me as a tool and to help others learn AI. To use it 534 00:48:21.400 --> 00:48:24.159 as a tool because people are scared of it, and you know, it's 535 00:48:25.199 --> 00:48:30.760 and it once it's you know, it's saying the fear isn't irrational fear. 536 00:48:30.800 --> 00:48:35.280 There's nothing to fear, it's you know, it's use it as a tool. 537 00:48:36.079 --> 00:48:40.760 And so from from from that perspective, I'm using it as a tool 538 00:48:42.760 --> 00:48:49.199 for a creative expression. Can I express myself creatively with AI? And that's 539 00:48:49.239 --> 00:48:52.800 the question, because I can't draw, I don't, it's I don't do 540 00:48:52.920 --> 00:48:55.760 it. I can't I can I don't. I don't draw, and then 541 00:48:57.039 --> 00:49:00.159 and I don't paint, and then, you know, and then I don't 542 00:49:00.159 --> 00:49:02.079 want to have to get into buying. Apply gotta go. I gotta get 543 00:49:02.159 --> 00:49:06.079 paint thinner, I gotta get brushes, you know, you gotta get supplied. 544 00:49:06.199 --> 00:49:08.440 And there's all this, you know, what with with JENERTI AI don't 545 00:49:08.440 --> 00:49:14.920 have to do any of that, and you just need your imagination. All 546 00:49:15.000 --> 00:49:21.800 you need is an imagination and the ability to write from your head onto paper. 547 00:49:22.039 --> 00:49:25.039 And you know we're talking wordcraft. You got it. Work Smithing if 548 00:49:25.079 --> 00:49:30.199 you if you can work smith ais for you. Now, most of these 549 00:49:30.280 --> 00:49:35.320 programs are free filled most of them, or you have to pay. Yeah, 550 00:49:35.400 --> 00:49:40.320 they they yeah, they all started out free. Now now we're we're 551 00:49:40.400 --> 00:49:45.960 almost two years into it. Now they're asking for money as of right now. 552 00:49:46.400 --> 00:49:51.880 The best one, well, I think ch p is is by open 553 00:49:51.920 --> 00:49:53.840 Ai. For me, it's I like that one. I think that was 554 00:49:53.880 --> 00:50:00.400 the best, and I pay them the the twenty dollars a month because I 555 00:50:00.480 --> 00:50:07.440 think they deserve it, so I pay for that. Microsoft is also using 556 00:50:07.519 --> 00:50:13.239 the same engine, the open Ai. They they're using chet gtp in in 557 00:50:13.599 --> 00:50:22.480 the Microsoft Being chat bot and it's completely free. And that being that Being 558 00:50:22.599 --> 00:50:28.519 chat is amazing. It is it's supposed to be a companion or a co 559 00:50:28.800 --> 00:50:32.079 pilot to help you fly through life. You know, your co pilot, 560 00:50:32.960 --> 00:50:39.039 and yeah, it can see, it can hear, and it can it 561 00:50:39.119 --> 00:50:45.239 can make pictures. It's it's pretty amazing and it's free. Check GGP does 562 00:50:45.280 --> 00:50:49.920 the same thing. I'm paying twenty dollars for what being is giving out for 563 00:50:50.079 --> 00:50:53.079 free. I use both of them actually, and they're both pretty good. 564 00:50:54.519 --> 00:51:00.440 And there's a couple of others that that I recommend. Uh wow, it's 565 00:51:00.639 --> 00:51:04.920 you know, to use as a tool. I mean I was watching I 566 00:51:05.000 --> 00:51:08.639 mean I was using gym. I was using chat GDP as a as a 567 00:51:10.039 --> 00:51:15.719 the movie buddy, you know, The Wizard of Oz was on the other 568 00:51:15.880 --> 00:51:19.719 night, and so I'm watching The Wizard of Oz and I got a I 569 00:51:19.880 --> 00:51:22.679 and I'm asking it questions and and come to discover thank you A. I 570 00:51:23.119 --> 00:51:28.639 did you know that Dorothy is a double murderer? The whole time I've been 571 00:51:28.719 --> 00:51:31.159 watching this movie? It never occurred to me. Did you know she did 572 00:51:31.280 --> 00:51:38.719 manslaughter twice kill? She killed two people? Did you know that the whole 573 00:51:38.880 --> 00:51:45.000 time she's been a murderer? Though, A I for you know, so 574 00:51:46.199 --> 00:51:51.239 you know, so I've been you know, so yeah, this it's it's 575 00:51:51.239 --> 00:51:52.920 been a lot of fun. I tried. I've been trying all kinds of 576 00:51:52.920 --> 00:51:59.360 stuff with the AI. Since it's on my phone, that means it has 577 00:51:59.480 --> 00:52:01.519 access to my camera, so I can use my camera as it eyeballs. 578 00:52:01.719 --> 00:52:07.519 I opened a refrigerator and I let it see what's in the refrigerator. And 579 00:52:07.039 --> 00:52:10.079 apparently it can see, I don't know, it knows within the containers. 580 00:52:10.159 --> 00:52:16.159 It figured it out and gave me a recipe of something quick to cook up. 581 00:52:20.400 --> 00:52:23.960 From looking, Yes, from looking, I pointed it at my refriger 582 00:52:24.119 --> 00:52:28.960 just to see, you know, what would happen. Yeah, it gave 583 00:52:29.079 --> 00:52:34.960 me a recipe of what did that might creep some people out though too. 584 00:52:35.159 --> 00:52:43.199 Yeah, that extraordinary. We had a leak. We had a leak in 585 00:52:43.320 --> 00:52:46.760 our bathroom, our kitchen, our bathroom. Sink was leaking, and uh, 586 00:52:47.119 --> 00:52:50.119 you know, I want to see what can a I do? Is 587 00:52:50.159 --> 00:52:53.320 supposed to be all that, So I took I used it that I used 588 00:52:53.400 --> 00:52:59.599 the my camera because you can open your camera through the app, and I 589 00:52:59.679 --> 00:53:02.920 took a picture of the pipes and everything underneath it, and you know, 590 00:53:04.039 --> 00:53:07.599 totally this is where it's leaking. What do we need to do? And 591 00:53:07.960 --> 00:53:12.320 it responded back saying, you know, I need to get an elbow pipe, 592 00:53:12.679 --> 00:53:16.239 and it gave me links to where to go get it, like a 593 00:53:16.400 --> 00:53:22.960 home depot and all that other kind of stuff. And there's one there's one 594 00:53:22.000 --> 00:53:27.559 pipe that I wasn't really familiar with under the sink and it's always got water 595 00:53:27.679 --> 00:53:30.039 in it, you know, and I couldn't understand it. And I was 596 00:53:30.159 --> 00:53:31.480 like, what's up with that? Man? And my wife was one or 597 00:53:31.519 --> 00:53:35.639 two? What is that? Why is there water in this one little pipe? 598 00:53:36.239 --> 00:53:37.719 And we'll come to find out that AI said, oh, that's the 599 00:53:38.000 --> 00:53:43.920 trap pipe and that it's supposed to have water in it because we live in 600 00:53:44.119 --> 00:53:46.280 it. Knew that we were in an apartment building, and it said that 601 00:53:46.440 --> 00:53:52.800 that water stays in that trap to keep everybody's back gunk from backing up into 602 00:53:53.239 --> 00:53:58.880 our into our bathroom. I had no idea. I learned that from the 603 00:53:58.960 --> 00:54:01.920 AI from taking a picture of a pipe, and yeah, I didn't know 604 00:54:01.960 --> 00:54:06.519 that was a trap. Yeah, it's uh, yeah, it's it's it's 605 00:54:06.719 --> 00:54:09.719 it's getting interesting. I think the next thing I'm going to do is, 606 00:54:09.840 --> 00:54:19.280 uh, my father's I got all my father's research that he put into discovering 607 00:54:19.679 --> 00:54:22.960 the roots of my family. My father and Alex Haley were friends back in 608 00:54:23.039 --> 00:54:29.199 the day, so, uh, when Alex Haley was doing his research on 609 00:54:29.360 --> 00:54:36.480 what would become the book Roots, my father also jumped into beginning his research 610 00:54:36.559 --> 00:54:43.159 into the origins of our family, and I have all of his his work 611 00:54:43.280 --> 00:54:46.760 here, and so I think what the next step is, I'm going to 612 00:54:46.880 --> 00:54:53.599 take images of his work and then feed it to the AI to train it 613 00:54:54.079 --> 00:55:00.719 on this information and to see what I can extract from it something interesting. 614 00:55:00.199 --> 00:55:04.480 We're just about out of time. Just thank you so much for your time, 615 00:55:04.920 --> 00:55:07.760 so fascinating, so interesting. Always look for your stuff. Where can 616 00:55:07.840 --> 00:55:10.639 people by the way, where can they see some of your art? I 617 00:55:10.719 --> 00:55:15.960 know you have a website, you have an Instagram account? Is to tell 618 00:55:15.000 --> 00:55:20.440 people where they can see some of your stuff? So far? Yeah, 619 00:55:20.559 --> 00:55:23.880 Instagram, And I think I think I'm going to keep everything on YouTube. 620 00:55:25.719 --> 00:55:30.920 Since everything it's become a lot easier to make video with with AI. And 621 00:55:31.000 --> 00:55:35.920 I'm still a video guy as opposed to a still image guy. So I 622 00:55:36.119 --> 00:55:40.679 like that I can create still images now from my imagination and then I can 623 00:55:40.800 --> 00:55:46.840 run them through uh an AI that will take images still images and turn them 624 00:55:46.880 --> 00:55:53.800 into three to six second moving video. Yeah, and you go ahead, 625 00:55:54.079 --> 00:55:58.239 Phil, we were out of time. Quick go ahead? Yeah, so 626 00:55:58.599 --> 00:56:02.880 so so will have more videos than they will be on The phils a i 627 00:56:04.199 --> 00:56:13.440 Art YouTube page fills with the z phils a i Art that's great, and 628 00:56:13.559 --> 00:56:16.119 you will while we're going along. You're gonna you're gonna put into You're gonna 629 00:56:16.159 --> 00:56:22.960 input some stuff. We we're gonna gonna make a wrong for us program with 630 00:56:23.480 --> 00:56:29.960 that's gonna be Neil will Neil will throw that on. That's absolutely unbeliefhoble. 631 00:56:30.519 --> 00:56:37.199 This is gonna be fun, good stuff. You know, Phil Sexton on 632 00:56:37.320 --> 00:56:42.519 the edge of AI technology, creative technology. Thank you for your intelligent conversation, 633 00:56:43.079 --> 00:56:45.559 and of course we offer special thanks to our listeners who take the time 634 00:56:45.599 --> 00:56:50.159 to give us a voice in their lives. Remember we offer a fresh topic 635 00:56:50.280 --> 00:56:53.960 every week and you can catch us wherever and whenever you get your favorite podcasts. 636 00:56:54.599 --> 00:56:59.119 Check us out on the Hudson River Radio Facebook page. Like us, 637 00:56:59.280 --> 00:57:04.920 leave us a common and we also ask you that you consider sharing Being Frank 638 00:57:05.559 --> 00:57:07.840 with others. For engineer Neil Richter, I'm your host, Frank Lobono. 639 00:57:08.280 --> 00:57:13.039 We hope to join us for the next Being Frank next year, Happy New 640 00:57:13.119 --> 00:57:19.000 Year. Everybody will be back for our first program in twenty twenty four, 641 00:57:19.360 --> 00:57:31.280 where the only way to be is Frank volume, can you me guitar ribs? 642 00:57:31.320 --> 00:57:42.119 Kick him moving me? We got the fire. We've been doing that 643 00:57:44.519 --> 00:57:55.079 rocking Barabella set the not alive ro rocking and I'll do it. Don say 644 00:57:55.760 --> 00:58:04.559 love, I want it, do it. We're the renegains the nose. 645 00:58:04.599 --> 00:58:10.480 If you ever tanged the alcantar, scringing mocking rollas our play so that the 646 00:58:10.639 --> 00:58:22.039 fond could you feel feet? If we get the fire ready to a night 647 00:58:23.039 --> 00:58:31.400 the funny started, we don't sign it's time to dance. That's like the 648 00:58:31.519 --> 00:58:43.320 whole world catch him the night that shot so bad bro spero real bird like 649 00:58:43.679 --> 00:58:51.760 the stars, the sky next sky cancer bye, but God its us the 650 00:58:51.920 --> 00:59:04.280 sky. Then this fons no bosting bats and that so can you be of 651 00:59:04.400 --> 00:59:09.960 the n We're gonna dance to the end of the line, ding moveless on 652 00:59:10.039 --> 00:59:31.719 the fire and we're losing ourselves in the sounding. Decide, turn up the 653 00:59:31.920 --> 00:59:37.440 solim candy feel on the feet, you take your over the room and dance 654 00:59:37.519 --> 00:59:43.960 and free and then be andy. Energy is so high not on the selectric 655 00:59:44.360 --> 00:59:57.159 lost nice, Turn up the onion faces, feel on the feet, frock 656 00:59:57.320 --> 01:00:15.360 the beat, m H, m HMMB dop the beach scream Hudson River Radio 657 01:00:15.639 --> 01:00:17.760 dot Com. MHM