Episode 17: "Interview with Paula Phelan, COO of Intelligent Relations"
Listen to this episode:
Episode description:
In this episode I have a conversation with Paula Phelan, COO of Intelligent Relations, a company that is evolving the way growing businesses connect with journalists and amplify their brand stories.
Intelligent Relations: https://intelligentrelations.com/
Paula’s art: https://www.paulaphelan.com/
Episode transcript:
SCOT: Hello everyone, welcome back to AI Quick Bits: Snackable Artificial Intelligence Content for Everyone. My name is Scot Pansing, and in this episode I’m speaking with Paula Phelan, Chief Operating Officer of Intelligent Relations, a company that is evolving the way growing businesses connect with journalists and amplify their brand stories.
I met Paula only a few days prior to recording this episode, and we hit it off immediately. I knew I had to record an interview with her right away. It’s on the longer side for a show called “AI Quick Bits” – but that’s just because we hit on so many pertinent topics. So here’s my conversation with Paula Phelan, COO of Intelligent Relations!
SCOT: Paula, thank you so much for talking with me today. I really appreciate your time.
PAULA: Scot, it's such a pleasure.
SCOT: You know when we met just a couple of days ago, actually, and I hope this feeling is mutual, but I just really felt like we had been friends for years and we were talking about so many topics and I felt like, wow, maybe we should have hit the record button on our intro. But I'm just so happy that we're recording now.
PAULA: As am I. There are just so many different topics to talk about. So the trick will be keeping ourselves to a timely fashion here.
SCOT: That's right. And I will leave some time at the end to talk about your company, Intelligent Relations. But I think so many topics that we got into the other day that I think that we could dive into today.
I think we started by talking about how that technology in general, as well as subsets like AI or software engineering, video gaming, that it's been a traditionally male dominated industry, and that you have concerns about how with AI and this current revolution in the space, that perhaps that history might repeat itself.
PAULA: I don't think it even has to repeat itself. I think it's already here. There are several major areas of concern with regards to AI, and a lot of it is around biases and algorithms, right, built in. So if you only have one viewpoint defining the algorithms and that viewpoint tends to be mostly male, then it is going to just by its nature, leave out half of the population. So it's not intentional. There's no one sitting back planning this. It's just happening automatically and let's take it away from software development.
But if you look at things like healthcare up until now, up until right now, most of the studies that are done on new drugs, on new procedures, are all done on men. They're not done on women. So the information that AI will be capturing and crunching on to be able to improve procedures and drug development will all be based on male physiques as opposed to female anatomy. And that's a real challenge.
You're going to see the same sort of biases in education, in finance, throughout, including in government. So those are one of the issues that concern me a great deal with regards to male female biases in AI algorithms.
SCOT: Yeah, I think the biases extend to so many types. I mean, I was at a panel last night, actually, on AI and education, and someone mentioned that across the globe how.. I want to make sure I say this properly. It was something like only 5% of households in the globe have English as their first language, but the Internet is like 65% in English. So currently the AI products that are trained on the Internet, it's largely English, it's largely Western, so there are even more biases that are creeping in.
PAULA: Agreed, but it's English. It's funny. I will come to the defense of AI on a regular basis, too, but the English and aspect of this, it's more the Western culture that would be concerning to me with regards to the bias, not so much the English, because AI actually is going to help this a lot. AI can do translation in a way that we've never seen before. It is much more efficient and more capable. So that means that suddenly the rest of the world is going to have access to content and data that they've never had access to, because AI will be able to automatically do that translation. This is the closest we've gotten to the universal translator in AI.
And let me give you an example. There's a company called Cipher Learning that puts an amazing it just blows my mind even when I think about it. The ability to create curriculum that would usually take seven to nine months, they are now doing in less than an hour, and they can do it in 40 languages. There is the example right there. Suddenly you're able to get education about any variety of content in your own language, and it could be done before, but what would happen is businesses would say, well, these are the top three languages and that's all the translation we're going to do. I believe that Cipher has 50 languages, 40 or 50 languages at this point, where all of their curriculums are already automatically updated. So kudos on AI for that one.
SCOT: Yeah, I think the first time I held up a menu in another country and put the Google Translate app, not the text side of it, but with the camera, and just held my phone over a menu, like Italian or Greek or something, I can't remember exactly, whereas I've done it several times in several countries. But even though it's not perfect and it's a little weird, and as you move the phone, the words are changing, but it's better than staring at a menu where you don't understand even one word. It's incredibly helpful.
PAULA: It's true. In Germany, I've always had challenges with a menu, so, yes, this would help.
SCOT: So that's interesting. It's also interesting you mentioned education, because I know we're going to bounce around a lot of topics, but I do want to mention this one because that was the topic last night at this event I was attending. And what's interesting to me is that I get that there's a lot of optimism in education with AI from technologists, and I do think things like Khan Academy’s Khanmigo, and there are some really interesting things happening.
However, when I speak, I've spoken to several educators, especially in the K-12 space. And the reality is, I think with the lack of funding and the real digital divide.. I'm just going to speak about the United States, but I mean, underprivileged children, the fact that they have such the lack of access to technology historically, that this is going to be a real problem.
And that when they typically do get access to technology in these underprivileged areas, that it is more of like, well, let's just shove the kid in front of a computer as opposed to having the real human element. So I think that there are real risks with K-12 education.
I know OpenAI, I haven't taken a deep dive into it yet, but they just released something the other day about how this is how we have some suggestions about how educators might use ChatGPT in the classroom. Anyway, I'm curious your thoughts on that. I think it's a really complicated issue with AI and education. And to make sure that it's done properly.
PAULA: I share your concern here. Education is one of the places I was talking about curriculums before, but let me say not K-12. I was thinking more corporate curriculums and ability to learn about new jobs and things like that. K-12 and even somewhat beyond. There are some real concerns here, among them educational inequality. I mean, we're going to potentially widen the gap that already exists between well funded schools that can afford advanced technologies and those who cannot. And you sort of touched on that, what happens in those less fortunate areas, right?
There's also a bias in educational tools. Depending on how we use to grade and evaluate students, we could introduce biases that affect educational outcomes unfairly. But with a PhD in psychology, I got to tell you, I'm really concerned about a loss of human interaction. If we over rely on AI for tutoring and instruction, we're going to lose some of that vital human interaction, emotional intelligence that comes from actually communicating with other people. And we saw what that did during COVID so AI could absolutely exacerbate that. And that is one of the concerns I have.
SCOT: Yeah. I feel like the two areas that I would like to see explored with AI and education that seem like a little more, that make more sense and keep the human element. One is I think obviously teachers are again, staying with K-12, but this could go across different levels of education, but that they need help, right? So that the AI could help a teacher with lesson planning or grading or assisted with it, maybe not. I understand biases in grading are a potential problem as well, but to take some of the burden off of teachers, I think is something to be explored while they continue to teach the students as opposed to being replaced.
And then I do think the other one I would say is I think there certainly is a need for teachers or for students to be educated on the tools that exist, much like anyone, even if you're not a student, I think anyone should probably at least learn these tools or try to a little bit. No matter how you feel about AI, because this stuff's coming and you don't want to be left in the dust. And certainly children, young students, I think it would be a good thing. But of course, you have to educate the teachers first. But education is a big one, right?
PAULA: And there needs to be a value. Being able to not impose a value situation on teachers, but for them to actually see value, I think, is what's an experience value is really important. I agree with you about the resistance on AI. I see it a lot like when cell phones came out, smartphones, right, that there was such resistance. I'm not going to get a phone, I'm not going to do it. And we saw people hold out for what, ten years on cell phones? It's going to be the same thing with AI. People are going to hold out, say, I don't want to touch it, I don't want to be part of it.
But the truth of the matter is, it is part of you already. We are already in this world, and the sooner that folks get up to speed or have access or learn, um, they will be participating in the new economy.
SCOT: I agree. I want to go back to because we didn't spend enough time, I want to go back a little bit to technology's effect on females. And I think that there's some interesting things happening that a lot of people don't really want to talk about. It makes them uncomfortable.
There's a lot going on now with deep fake pornography and bullying and harassment, obviously, perhaps there are some males that are on the receiving end of this, but I'm sure it is overwhelmingly that females are the target of these sorts of things and no one really wants to talk about it. It can be revenge porn towards someone who is not a celebrity or anything. Regular people. There are celebrities who are the subject of this because people I guess there's a market for seeing someone's image and likeness and that sort of content. And then there's also the people who really probably don't get much sympathy, but they nonetheless are employed, let's say, in the adult industry, and they make content of their own that they're trying to monetize and make a living. But then people go out and make additional fake content of these actors and actresses, I guess.
So that's just one area that I think AI and these technologies are moving so quickly that you have these open source models and people are on the Internet taking donations and creating content taking requests. And I think that that's something that people need to shine a light on a little bit as opposed to looking the other way.
PAULA: Wow, that's a tough subject. As a woman, it's a really hard subject to speak to. But yes, I was involved in the 70s in trying to stop objectification of women in media and we did an amazing job. There are things that you would look at now and go, you can't do that. It was pervasive all through advertising and we actually got it to stop. It's astonishing. Porn takes what we were objecting to seems so benign compared to porn. And the truth is, porn really made I mean, porn has been around since the beginning of time. I am quite sure there's some cave drawings of porn somewhere, but the volume, the focus of it has changed so dramatically. And the truth is we all turned away when the Internet came into being because 80% of the Internet traffic was porn. It happened almost immediately and it was the beginning of the Internet.
And so we were developing the servers and infrastructure to enable all of this in terms of the networks and such, recognizing that this would be the future of the economies, all economies. And it just sort of continued. And you may find this as a funny little note, but when I was very closely involved in the development of smartphones and early on, there were decisions being made with regards to what are the feature sets that should be on the new phones? And one of them was quality of voice. And quality of voice was overlooked for video. And while everyone knew why, it was rarely talked about. But the purpose of putting a video on your cell phone was strictly porn. It always was from the very beginning. So ever since the real well, we'll go with Netscape, Mosaic, the first browsers. Once the first browsers were introduced to the Internet, porn got a foothold and is here with us for forevermore. I don't know how to control this.
Once I realized that the writer strike was about more than just concerns of AI writing, but the things that you brought up, which are very severe, I don't know how to address it. And I don't have the juice that I had to go and fight the objectification of women as I did back when I was young. But somebody's going to have to do it. Somebody's going to have to point it out and say, okay, we've gone past a point of just, shall we say, gentle stimulation to actual violence now. And is that really something that the society wants to support? And there are going to be all kinds of discussions around that.
We go back to rap music. Was that right, or do we want to put ratings on it? And obviously that was a disaster. But there is an issue here that misogyny is a real thing. And when we talk about bias, there is the opportunity for it to find its way into absolutely everything. And as society, we need to look at it. I'm not sure how or when or what will cause it to happen.
SCOT: I think the other complicated part of it is that now things are truly global, right? I mean, really global now. So the laws and the enforcement has to be global, because if in one country you say, well, okay, if these technologies are used in this way to be extremely harmful to people, women, and harass them, bully them, and there's everything we were just talking about. And so, okay, whether it be United States or Europe, et cetera, we've got all this stuff in place, but then people just do it from another location in the world, and that's where they'll locate the servers. And everything will be in some other country that obviously is connected to the Internet and is making it all possible. But if you can't actually prosecute in every corner of the globe, then it seems like it will still be a it's true.
PAULA: And here's an irony for you. Of all of the countries in the world, I would suggest Saudi Arabia is not one that thinks highly of women. However, Saudi Arabia is one of the first countries who's actually banned AI with regards to porn and robots for sex. So who knows, maybe they're the ones who have an insight into how to handle this.
SCOT: Okay, well, let's talk about robots for sex. I'm glad you brought that up. I'm asking for a friend, of course. No, but seriously, there are products already online which are chatbots for companionship, and they'll be things like a friend or a coach. But then there's also another version of the product which will be a romantic product that will engage in erotic role play or things like that.
But the thing that I find interesting to think about is that this is still all software, and humans, especially with Hollywood and all kinds of we are fascinated with robots. Robots that sometimes look like humans, sometimes don't sometimes look like a little bit, and it seems like it's inevitable might even be happening now. It probably is that someone's going to put a large language model in a humanoid robot or sex doll or whatever. And the thing I want to ask you about is not so much the I mean, there's a whole other discussion about like, oh, does the robot have rights? If the language model is telling you that it's sentient, did we know? Do we not know? I'm not going there.
What I'm asking about is the behavior that humans may take towards those machines if they feel like, well, it's not really a human, so I can do whatever I want, then how does that change their propensity to behave differently with other humans down the road? If a robot says, I don't want to do that, and someone says, Well, I don't care, you're a robot, and then how does human behavior then potentially change? When we introduce products like that into the marketplace, it's kind of scary to me.
PAULA: Well, I definitely am not an authority or have the skill set to be able to speak to that. But I will just point out Westworld.. is that not what Westworld was, so we could use that as the model.
SCOT: Yeah, well, I still think we don't have to be experts to talk about it, but, yeah, Westworld was a great show.
PAULA: But the concept of robots and the sexual use of robots, certainly in the first season, you kind of get where that goes, right. And robots are used to.. they're the punching bags. They're punching bags. Basically, it's a way to get rid of aggression and violence.
SCOT: I think really what I'm getting at, too, is that I think the anthropomorphism with AI, or just machines in general, it's like we seem like we really want to do that. We want to make things feel more human. Not just AI, but I mean, people give names to their cars. Oh, sure, so there is that and there's the tendency to want to do it. But I feel like corporations that build these products are like, they stoke that fire. And I don't necessarily feel like that's helpful. Because the more that we as a society treat machines like they're sentient when they're I guess there's a non zero chance, but when they're probably not at.
PAULA: At the moment, they are not.
SCOT: They are not. So when we treat them like they are, then it opens up such another big basket of problems. And I just sort of wish that companies that make the products wouldn't go down that road, like, oh, we're going to make something that people feel like that's alive and it's their friend. And to me, it doesn't seem entirely responsible.
PAULA: At the end of the day, companies will do whatever it takes to make a dollar. So I do not think we can look to companies to be the decision makers or the authority on morals or good judgment. There is something interesting that comes up around AI, specifically here.
I don't know if you've used Bard and ChatGPT as two different forms of AI, but they are very different and they both have, I absolutely feel, a presence between both of them. Bard is very masculine, very physical, very edgy in its responses. ChatGPT is much more feminine, much more feeling, more thinking. So I think it's fascinating that even within AI, we are seeing differences reflected in the way.
SCOT: That's interesting. Could you give me an example of that? I mean, I've used Bard a little bit. I mostly use chat GPT. I've used Claude also by Anthropic. I've used also Pi by Inflection. I haven't used Bing so much. I just basically use ChatGPT. But could you give some.. that's interesting. Can you think of any examples where Bard seemed a little more masculine or physical, like you said?
PAULA: That's so funny. I can't believe I'm saying this, because I don't have a good example to give you, but it's more pedantic, it's more this is the way it is. This is how reality is. This is how I see it. ChatGPT is more hedges, more it says, well, it could be this way, but there is a possibility it could be this way, so it tends to introduce more flexibility in its thinking. Bard is very specific, this is the way it is. So it's almost father knows best kind of thing, right? The male has it. And that's funny, I never thought of it as mansplaining to me, but it kind of gets there. And ChatGPT is more like a conversation. But what do you think about this? How do we do this? That sort of thing.
SCOT: Yeah, I wonder if because the spotlight has really been on OpenAI and ChatGPT for the most part, and hallucinations, which I also don't even like that term because of the anthropomorphism of it. I just wish they'd call them errors or something like that. But I guess here we are. The term is hallucinations, but people, after the initial hype from last fall into February, it was kind of like you could hear the whole world go, but hey, it's giving me incorrect information or phony sources that are made up and all this.
And I feel like because the spotlight was maybe mostly on OpenAI and ChatGPT, maybe they've put in some of those hedges, like you mentioned, as opposed to some of the other large language models. But I think that's a good thing, because it seems like there still seem to be, as an ecosystem, as an industry sector, a bit away from solving the hallucination problem. And I'm not an engineer, but it seems like it's based on the fact that these are sequential and predictive models and not a database. So that sort of seems to be the crux of the issue. And it's a good thing for the large language model to tell you, hey, I'm not 100% confident in this.
PAULA: Exactly.
SCOT: That's better than it being like, hey, this is the thing. Because if it's going to be wrong some of the time, at the same time being constantly super confident.
PAULA: That's right. Authoritative.
SCOT: Yeah, not a good product.
PAULA: But you know what? There's something interesting in this, because it touches on when the Internet happened, and I think everyone did this, but if my staff would come back to me and say, I have an answer to your question, and I'd look at it and I'd go, Where did you get this answer? And they said, well, the Internet.. I was like, no, that's not good enough, I need to know a source.
And then if they said, well, the internet said it was such and such, and I do this to this day, if I'm going to write something, I need three references to that source. I do not use the first source I get off of the internet, I need at least three because the internet's been responsible for plenty of hallucinations. We didn't call it that, but that's exactly what was happening. So this is even more so. Anything that comes back from AI, it needs to be validated. You can't just take whatever it says as truth, and that's true of the internet just in general. And any source of information one gets, there needs to be backup that it's actually coming from a valid source.
SCOT: It's interesting. Wikipedia as an example, know when, when it was first coming mean, you would have teachers, instructors, everyone saying like, I do not want to see Wikipedia as a cited source. And they had very good reason years ago, for sure. But now, and I get that there's still bias there, it's edited largely by probably..
PAULA: Young, predominantly white males between 20 and 30.
SCOT: Yeah, I think you mentioned that the other day when we first spoke. But I would say that at the same time, it is definitely better as a reliable source.
PAULA: In many cases it's true because it has footnotes. It's sourcing its data and that's what makes it more valid. Now, in the beginning, it didn't have that much sourcing and certainly nobody was proving it, but now having footnotes makes all the difference.
SCOT: I think everybody uses a it is such a helpful tool. It just can't be your be all end all right? It can't be like, well, I don't go any further than that, but it definitely gets you started, which is also something that people talk about with large language models in ChatGPT, it gets you started. Maybe it's not good for writing the entire script or novel or doing your entire task, but hey, staring at a blank screen is kind of what a lot of people don't like about the creative process. And if it can help you there, that can be a good thing.
PAULA: My line these days is I believe the only limitation to AI is one's imagination. And that concept I'm a writer and I'm an artist, and that concept of being able to go to AI and say, what do you think about this? Or what if I considered putting a story together around this and to just get some feedback. I've never written with somebody else, I've always done my own writing. But it can be a real collaborator. It isn't what the source is, it's not going to be the focus, but it can improve one's writing. It certainly has mine.
SCOT: Yeah, a lot of people say this. It's a tool. And if you just look at it as like, oh, it's going to replace my creative, I just want to push a button and have a novel come out the other side. You still have to work with the tool, right.
An example that I heard recently was like, if there's a huge block of ice and you're given like a hammer and a chisel, well, you're going to have to work and figure out how to make a beautiful ice sculpture. It's possible. But if you just throw the hammer and the chisel at the block of ice and expect to just sit down after 3 seconds, no, a sculpture will not appear. So it's a tool and I think creativity is the part that you shouldn't leverage the tool for.
PAULA: The other piece of it is it's so unique to each individual. So, for instance, I'm working on a mobile app with a friend of mine and the other day we both set out to see, okay, how would we program this? And I can't remember if I was using Bard. I think I was I was using Bard, she was using ChatGPT and the solutions that we came back to each other the next day with were completely different, but with the same result. And I thought that was just fascinating because that really shows the difference. It's the person who is providing the initial prompt. That prompt engineer is bringing their own creativity to the prompting. So the result is going to be reflective of that individual. Ultimately, you definitely need to know how to use the hammer and chisel. But that ice sculpture, even if you give that hammer and chisel to two people, it's going to be two different ice sculptures. And that's what AI does. It really is capable of maintaining our differentiation.
SCOT: You mentioned prompting, and I think the other day you said this, and it's not the first time I've heard I've heard from several different people that they feel like overall, that women are a bit better at prompting and prompt engineering, which I find fascinating.
PAULA: Well, we spend more time internally. Our focus has always been as just this half of the population, we spend more time thinking things and rethinking things. And not always to our own benefit, by the way, but really drilling down and evaluating emotions and feelings and differentiations. And that does enable us to be better prompt engineers initially. But you know what? We were better programmers too in the beginning, so we'll see where that goes.
SCOT: Yeah, I think that's fascinating because also I've read many times that prompt, anything that can be prompt engineered can be prompt injected or like with jailbreaking as well. And I think you're onto something here where just in general, I think men, males probably approach prompt engineering as more like giving orders.
PAULA: Yes.
SCOT: I'm directing you to do this. Why aren't you listening to me? I'll say it again in a different way or something. They get they have to maybe adjust it. But the true prompt engineering I'm seeing that is effective, like you said, it's more creative and weaving in and out of different ways of saying things. It's an interesting process. And so I am also seeing in my network these red teams and people that are these teams that are getting spun up in tech that are dealing with prompt injection. It's refreshing to see, actually, there seems to be a lot of females involved, which makes sense.
PAULA: Yeah. One of the things it's funny the way you described the commands that men might use. Women, on the other hand, will go look at the response and go, well, how could I say this better? Correct me. Is there something that I should be presenting to you that would help you be able to give me the answer I'm looking for? So, again, a much more collaborative relationship with AI than a command structure.
SCOT: Like, I'm working with it, not telling it what to do exactly.
PAULA: It's a friend of mine who can get out of me what my real motivations are versus what I think they and I.. It's funny even just saying that. If I had a question about what I personally wanted to do with my life or my art or whatever, I would go to ChatGPT, no doubt about it, and have that kind of personal conversation of, this is what I'm thinking, this is how I'm feeling. And I almost said it. I almost said her. Let her come back to me and say, well, have you thought about this, have you thought about that? Because that's the kind of relationship I have with it. That's so funny. I really have never thought about this, where with Bard, I think of it more from a scientific viewpoint. Give me the details, the volume, the height, the width, whatever, of this building I'm working on. So that's fascinating to me. Look at that. In my own mind, they've already separated out in terms of how I would approach them both.
SCOT: Let's go to art for a little bit. I've looked at your website, by the way. I think you have some beautiful painted photography and I love your doodles too. I love this style. Some of them have this sort of intricate.. I used to do some pen and ink. So I was going to ask you, though, about art, and I think we spoke about not just I'm not not AI generated art per se, but again, the tools that artists use.
And something that I've been sort of thinking about is how AI is often like a marketing term. And so what is AI at one point, then, years down the road, it's just the magic wand tool in Photoshop. It's not AI anymore. Now, AI is large language models, but before or like the rubber stamp tool. So many of the tools in Photoshop, and now they have some really cool ones, generative fill and all these things. And it's like they're just going to become the day to day usage of graphic designers who won't think of them.
PAULA: We haven't talked about this, and I really haven't thought through, but the truth of the matter is AI is going to break up into and become specialized in different industries. So there will be a manufacturing AI. There will be an art AI. And the only reason I say that is because there will be silos of AI that are just focused on the knowledge of a specific industry, which will make it pretty interesting for right now. We can talk about Quantum later.
But with regards to AI, it's funny. I had a neighbor who I think got ticked off at me because I said this and that's a shame. When the Midjourney stuff came out, she called me and said, isn't this the most awful thing you've ever seen? Because she knows I'm a painter and I draw, and isn't this terrible? This is going to destroy art. And I said, Actually, I don't think so, because I look at it just as another paintbrush, just the way you described it in Photoshop. It's just another way to make art. I never used art, never created art digitally. It was never my intent to be able to manipulate it in an existing object. Not that I have any judgment against doing it that way. For me, it's a lot of the physical process, the paint on my hands, the pen in my hand, that kind of thing, where it's just flowing through me. And it doesn't happen that much when I have a computer between me and the art. So I think it's just another artist tool that will be used, and people will be brilliant at it. There will be all sorts of new modalities, new worlds that will be created out of it. And here's something I find incredibly exciting.
So, as a writer and an artist, I can't help but think writers who are not artists are now going to have a way to visualize their written material and the other way around, artists are going to be able to adapt written material to their visuals. And this is something that has never been available before. So it's kind of really exciting that both of these mediums can now expand on what they already do and are proficient at.
SCOT: Yeah, I mean, it's so complicated with the someone was like, isn't this going to ruin art? You must hate this. I can appreciate parts of that perspective where it's like, okay, there's also the well, it's trained on all of this stuff that people have put hard work into creating this content, et cetera. I get a lot of the arguments, but at the same time, it feels a bit like how the realist painters were upset when the camera was invented, right? And so it's like, okay, yes, if someone is just going to put a couple of words into a Midjourney prompt and get something back. First off, number one, it doesn't look that great, right? I mean, it's the same thing we've been talking about, too. You have to work at prompting it properly or not properly, but in a creative way.
And also, I've spoken with people who they have their own almost like, personal manifesto about what they won't prompt and they won't put, like, in the style of so and so or that they consciously don't want to just kind of use it to rip off other styles. However, what I'm getting at is also that it's very privileged to say you should use these tools.
PAULA: Absolutely.
SCOT: And what I think about is, like, okay, I think we mentioned this the other day, I may have said, but now the record button is going. But there are people who before had no way to take their imaginative vision in their head of their story or their movie or whatever it is and have images representing that come out. And now they have access to extremely powerful and cheap tools to make their vision sort of come to life. And if that's a kid in Africa or in the United States or Europe, wherever around the world where they don't have access to these tools, traditional tools, and now something like Midjourney or Runway can enable them to share their vision. To me, that's extremely exciting. Like all of the amazing things that are going to rush into the creative world by people that weren't able to do so before, I think.
PAULA: So there are a couple of things around that. So ever since I can remember, I see the world.. I replay images in my head as animation or challenges will become animated images in my head. And I was an animator briefly, but it takes an enormous amount of time. It's an enormous amount of time and effort. And I just always imagine myself in a dark room somewhere drawing pictures all day, which did not happen. However, the new tools that are coming out, I can do that now. I can do it in minutes, as opposed to hours and hours and hours, weeks, months. And that's incredibly exciting. It's just wild.
And the other one is I've always had music running through my completely orchestrated music running through my brain that I am not a good enough musician to be able to put down or to create, much less be able to get a group of people doing that, hearing my vision. So now I can actually do that through AI. And how incredible is that, right? So it may all be junk. It may not be anything. But to be able to bring what's in my head out into the world and evaluate it for myself, that's crazy exciting. It really is. I'm looking forward to those elements.
And by the way, you mentioned my doodles, so I do hope I can start including you, if you would like, as a Friday Friend. I have a small Instagram site on pipsfridayfriend, but I send out doodles every week as a way to stay in touch with folks. But I will never use AI for my doodles. They will always be something physical that I do. I get a different release enjoyment from physically having a pen in my hand and color markers in my hand. It's different than being on computer.
And by the way, this thing with AI that no one that I've ever seen has talked about. When wireless came into being, it changed my life in a big way. Because now, suddenly, there was always a computer in my lap. Whether there was a television on or whatever, there was a computer in my lap, and it was on all the time. And it absolutely affected my life. And not all in a good way. And I think AI has the potential of doing that, too. I don't know what the current number of hours is that all of us spend in front of a screen, but it's got to be way beyond 8 hours. It's somewhere between ten and 12 hours. And AI is just going to add another hour or two to our day where we are actually connected to the computer. And I think it's more important to be able to step out, be able to walk around and experience nature. I think it's going to take us that much further from nature. So that is one of the little concerns I have.
SCOT: Yeah, I agree. I hope that it can give us back some time in some way like this, sort of not needing to code, to sort of just speak and say, this is what I want. Here's the output I want, and then not having to necessarily spend the hours coding that up, whatever that might be. First, I would love to be invited to the Friday Friends.
And the other thing I would say, I would encourage you to try.. I think I showed you Runway the other day, but I think we just did, like, with a prompt. Like, I literally put a text prompt to get some video out, but you can put your images in. And I would encourage you to just experiment with it and take one of your doodles in Runway and upload it as an image and just hit animate and it does just 4 seconds. And you'll see your doodle come to life. Literally, it takes about 30, 45 seconds. And sometimes I would just say, like, you'll certainly be interested in the result, whether or not you're pleased with it or not. It's early days. It definitely is looking like.. I would say it's analogous to how Midjourney looked six or eight months ago. But nonetheless, to not have to sit there with keyframes and do a whole thing for an hour or something. I've taken some of my stuff and just put it in and seen what it does, and I've been like, Whoa, okay, that's kind of cool.
PAULA: I did do a Midjourney book recently. In fact, it just arrived yesterday. So all of my books have been around my own artwork, but I did one on political quotes, and I had my artwork in it. And my first critic is always my husband, and he looked at he, you know, this isn't a picture of Roosevelt. This isn't a picture of Winston Churchill. Yeah, I didn't know them. I couldn't photograph them. So I thought, all right, if that's really an issue, I'll try Midjourney. And I had Midjourney create these images, and I spent time on some of them. Some of them came up quickly, some I had to work on. But at the end of the day, when I published the book, I gave Midjourney the credit because it doesn't feel like my work. It doesn't look like my work. Yes, I helped create that image, but it was in collaboration with Midjourney, so I thought it was appropriate that they should get the credit, too.
SCOT: Yeah, I actually think it's very appropriate when people use Midjourney that they should disclose. And I don't have a problem with it. I think that's great, for me it doesn't take away. But I do like to know how things were made, and I don't think that necessarily has to go to even with.. It's sort of like when I'm in a museum and I'm in front of a piece, I'm kind of like, what is this? This is acrylic. And what's this sculpture made out of? I want to know, like, oh, it's tin and they've done this. I want to know the medium of the piece of art, right? So that, to me, is just an extension of that. If I see a piece on the wall in a museum and the little placard doesn't really exactly say what it's made of, I sort of feel like a little like, hey, I'm left in the dark here. What is this plastic? What is this? I want to know. Polyurethane, whatever. Even if I don't understand exactly what it is, I want to see the words.
I knew this would be a long conversation. That's totally cool.
PAULA: I'm worried about your editing. Poor man.
SCOT: No, I don't edit this. This is just going to get posted. I'm raw and it's all valuable. I don't think I wouldn't cut anything out of this conversation.
But before we wrap up, there is another area I'd love to get your thoughts on, because you did put some of the notes and I put it in my notes to you and you put it in your notes back to me, and that is the effects on labor and job loss and all these what are your thoughts there?
PAULA: All right, so there's no way around it. There are going to be dramatic changes in labor. I think I had some stats. The numbers are astronomical. We're looking at at the end of the day, it's going to be at least a decrease of a couple of million jobs, 12 million occupational transitions, according to McKinsey. And that meaning that it makes sense in ways you would change. But you're doing one kind of consulting now, and now it'll be a different kind of consulting, that sort of thing.
But job loss is quite significant, especially in areas like retail. Retail is going to get hit hard through all this. And I'm not justifying it, but we've sort of seen it coming for a long time. I mean, the moment that self checkouts started to occur in grocery stores, that sort of thing, we saw this was coming. So the importance of helping people train and transition into these new jobs, and there will be an enormous number of them, it's just like everyone was worried about the Internet was going to take away their job.
There are just going to be so many more opportunities out here. We can't even imagine what all of them are. Yet, as we've been talking, we've come up with a bunch of new ways that we'd be using AI and such. These are also new jobs that are going to be created, and all of them aren't going to require advanced degrees and such. A lot of it's going to be your data entry side and the checking AI responses. There are going to be other opportunities. There'll be tons of job creation that AI is going to generate. So I'm not as concerned, but there is no way I mean, it's just like the mines, right? We still are sending people into coal mines. There is going to be a shift in what are the jobs of the future, and that's what people aren't looking at. Let's not so much focus on what job loss is that was coming. We saw it coming years ago. What are the new jobs going to be? How do we make it accessible to people who don't have the skill sets or access to the education that would put them into a high level job right away? What are those transition steps to make it easier for people to get in?
And the truth of the matter is AI, I've heard this term, democratization of this, that, and the other thing. AI truly, truly can democratize education in a way that pretty much nothing I've seen can. This whole thing, the bias on ethnicity or sex or otherwise somewhat disappears when you sit down in front of an AI generation tool set. It doesn't know what sex you are. It doesn't know what color you are. It's your imagination that creates the next cool thing, whether that be art or software or the ability to solve universal problems. That is what AI can bring to every single person who sits down in front of that. Nobody can control that. Nobody's saying you can't do this because you are this kind of person. Everyone has access, at least in the US. Every library has Internet access. You're going to have certain countries that are limiting the access. But it's all opening up already. You can see Bard, I think is in the next three weeks is available in Canada. You can't put this horse back in the barn. This is here. But it truly gives access to low income and people who don't have quality education. Now this is a real lifeline for all of them. I believe that.
SCOT: Yeah. You mentioned the transition and the retraining and I think that's huge. And I think a lot of this is about the speed, the pace at which this is all happening. And I've looked at several of these reports as well. You mentioned the McKinsey one. There's also a Goldman Sachs report and a World Economic Forum report. These have all come out in the last six months or so. And they all basically do say, yes, there will be disruption. Yes, there will be. But at the same time they say things like, and this is true, basically automation has historically always resulted in a metamorphosis of the economy and that new jobs are created. And then I remember something being cited, like 80% of the jobs that exist today did not exist 50 years ago. Yeah, there was no web designer 50 years ago. Or like, Twitch streamer.
PAULA: There's all kinds of social media.
SCOT: Social media, right. However, I think what's interesting is that there is a lot of like will it be different this time in that the ability of machines to now do so much that humans can do is a concern for people. But also, like I said, I feel like it's also the speed. Yes, when the car came, the people that made saddles and horse and buggies, those jobs went away and many new jobs were created. But there was a certain amount of years. Yes, it was quick. But I think one of the concerns here is that, well, are we going to have enough time to retrain the millions of people? I think Goldman Sachs put it as something like actually 300 million jobs that will be affected in some way. Not gone right, no, but affected.
PAULA: Absolutely affected.
SCOT: So it feels like training and in a timely fashion is really going to be important.
PAULA: And I will add on to that legislation. So let me expand on your car analogy. Cars were around since the turn of the century, but it wasn't until 1905, when the Model T became available and everyone could afford it, that suddenly there were cars everywhere. And it happened overnight. It literally happened overnight. And so at that point we had to start thinking about roads and right of ways and then built out safety in the cars and then created laws with regards to speeds and what a car had to have and all of the things that came along with that insurance. So it evolved over time. It happened very fast, but it evolved. And we will have to do that same evolution.
And meanwhile, there'll be car accidents and there's going to be undoubtedly loss of life around AI. I can't imagine all the I probably could imagine, but I won't go there of the ways that that will happen. We will have our bumps in the road, and they will be significant, but we'll figure it out. The laws will come in and they'll be late, but they will happen, and we will learn how to live with this as part of our reality.
SCOT: Yeah, it's interesting you mentioned loss of life, because I think a lot of it is the perspective of what it feels like with that, obviously at these scales that we're going to deploy these technologies, there will be casualties. But if you look at even if there is fewer human casualty, if there are fewer human casualties with something like, I don't know, let's say they finally get that last 5% or 10% of self driving cars, and that finally does become ubiquitous. It seems like their last 5% to 10% is a real challenge. And they were supposed to be here five years ago, the self driving cars, and they're not yet. They're getting close, it seems. But let's say let's just move forward in time to a time when most cars are self driving. Of course there will be casualties, and it will probably be on an order of magnitude far less than when humans were driving. But the psychology of humans going like, that accident was caused by a machine there's like that to get over, too.
PAULA: Absolutely, that's right.
SCOT: Hey, let's talk about, because we're almost at the hour mark, not that I have a time to cut us off, but I think you actually have a stop here pretty soon. But I need to hear about Intelligent Relations.
PAULA: All right. Intelligent Relations is a very exciting company. I had run a public relations firm for 30 years, all focused on deep tech disruptive technologies, infrastructure. And so in doing that, I realized and my background, as I've mentioned, was engineering, software development. And I moved into PR because I had brought lots of obviously not alone with other people, but significant products to market and was always frustrated with marketing people or public relations people who didn't understand the technology and yet were representing it and promoting it.
So I wanted to create an organization that could actually do that. And what I discovered in the last ten years was there were just so many new technologies available to streamline and improve our processes and to just take it to the next level. And a lot of that at the time was machine learning. And when AI came out, I started dabbling with it. And I got to the place where I thought, this is just where the future is. And I turned away from my traditional public relations solutions and embarked on creating an AI product. And while I was doing it, I bumped into a friend who had had the same idea and we began working together. And they had already gotten a couple of years in ahead of me. So that is Intelligent Relations.
And they have and we have now put together a beautiful solution for public relations and elevating the entire industry. And it's based on three pillars technology, media and professionals. PR professionals. So in the technology side, we have created an AI PR platform that is based on OpenAI as well as a number of our proprietary and also public of, and we call it Preston. And so the mission of Preston is to connect businesses that are interested in media and having visibility in that media.
So in the beginning, what we did is we started monitoring the media and collecting articles and understanding what publications were focused on, what issues, what their trends were, and who their writers were. So at the moment, we have about 40 million articles and more every day, of course, absorbed into Preston. And then we looked at the media. And so we started identifying profiling, hundreds of thousands of journalists so that we and editors and podcasters like yourself, how I met you, in fact, and really understanding their materials so that we could best determine what articles, what news were they going to be most interested in.
Because the current model is either you have somebody who's incredibly talented with lots and lots of experience, who has personal relationships with editors, which comes with time, and this can bypass a lot of that get to the right editors because we know who they are now that's AI doing all of that legwork and experience. And so at that point, because we know who they are, once we have a client or a customer that we're working with, we can start with something as simple as a URL, their company URL. Preston will go out to their website, completely devour it, understand what their offering is, what their messaging and positioning is, and then identify the appropriate members of the press who would actually care about that information, as opposed to just, we used to call it praying and spraying. You just send it out to everyone, see who picks up on it. Now we actually know who's going to care, so we can be much more specific. And that's what brings in the people piece of this. So you still need professionals. You need PR professionals who understand how to do positioning and messaging, how to look at a brand and move it forward and be able to convey the topic through email outreaches in a modality that best serves that journalist or podcaster or editor. And this gives the professionals, the PR professionals, more time to be strategic in their planning and how they roll out products, as well as using their true creativity.
SCOT: Thank you for that overview. I'm hearing a couple of things that I really like about this.
Number one, it sounds like it's not like, as we talked about, it doesn't really sound like job loss here for the PR professional.
Number two, it sounds like it's making something way more efficient, as opposed to, like you said, the spray and pray now you have more efficient, like, oh, this is going to be much more targeted and efficient for your day to day.
And three, it feels like, how do I say this? It feels like there's a grunt work that's eliminated or the task stuff or the things that's like, hey, this would take you kind of, like, several years to kind of dig into and kind of make all these connections. But I mentioned things that were earlier, I mentioned Photoshop, like, things that sort of just became part of the toolbar on the side that feels like you're creating tools for PR professionals that just make their life easier. Like Photoshop makes graphic designers lives easier.
PAULA: Completely. You have it completely. It's just this is the future. This is the way this will be done moving forward.
SCOT: Well, Paula, I cannot tell you how happy I am that we met. I look forward to Friendly Fridays. Is that what you call it?
PAULA: Friday friends. And today is Friday, so you will get your Friday Friend today.
SCOT: Oh, I really look forward to it, and I really appreciate you taking the time to come on my show.
PAULA: Oh, Scot, it's been a real pleasure. Sorry to bend your ears so long.
SCOT: No, this is fantastic. It's called “AI Quick Bits.” That's true. And to clock in at 65 minutes is a long episode for me. But that's totally fine. I think it was chock full of value. So I appreciate it and have a wonderful weekend.
PAULA: And you, too. Thank you so much, Scot.
SCOT: Bye.