Why Smart Women Podcast
Welcome to the Why Smart Women Podcast, hosted by Annie McCubbin. We explore why women sometimes make the wrong choices and offer insightful guidance for better, informed decisions. Through engaging discussions, interviews, and real-life stories, we empower women to harness their intelligence, question their instincts, and navigate life's complexities with confidence. Join us each week to uncover the secrets of smarter decision-making and celebrate the brilliance of women everywhere.
Why Smart Women Podcast
Want the best out of AI? Direct it like you’re Scorsese!
Premium Peptides for Longevity, muscle growth , weight loss
Proudly sponsored by COUP — helping brands cut through the noise with bold, smart marketing. Visit the http://coup.co website or book a meeting with us at. https://go.oncehub.com/RequestMeeting
You are listening to the Why Smart Women Podcast, the podcast that helps smart women work out why we repeatedly make the wrong decisions and how to make better ones. From relationships, career choices, finances, to photo jackets and chaos movies. Every moment of every day, we're making decisions. Let's make some good ones. I'm your host, and you're McCubber, and as a woman of a certain age, I've made my own career of really bad decisions. Not my doctor to find it. And I wish this podcast had been around to save me from myself. This podcast will give you insights into the working of your own brain, which will blow your mind. I acknowledge the traditional owners of the land on which I'm recording, and you are listening on this day. Always was, always will be, Aboriginal land. Well, hello, smart women, and welcome back to the Why Smart Women Podcast. Today is what's the date, David?
SPEAKER_00:It's the 20th. Is it the 20th of November today? Or is it the 19th?
SPEAKER_02:It's the 19th of November. You will be listening to this on the 20th of November. Um I am broadcasting from DY and the Northern Beaches, Sydney, New South Wales, Australia, and it is a pristine, perfect spring day. And as I look at the news in the evening and I look at the reporters all over Europe, I note it's getting cold over there. So sorry about that. Sorry, not sorry. You all need to come to Australia for a holiday. So um last night we were um preparing for an acting class, and um David in the previous couple of hours had had an experience where he'd done a casting for a job, and the casting situation environment was not positive, was it, David?
SPEAKER_00:Oh no. I mean it was probably one of the one of the worst casting conversations that I've had with the uh you know the director and the uh the apparatus uh that I've ever had, and really quite surprised about it. Um I would even say it was kind of insulting, you know, when it was an opportunity to have a a conversation at depth to um understand your perspective on the role. Yeah, that's right. You know, get to get alignment around values, to to to talk about approaches to to actually performing this particular role. Um instead of having that rich um clarifying conversation, it just felt like a um a tick the box, what do you know about us? What can you tell us that we can put to our advantage if you were to join us? Um and um So it was a very disappointing casting.
SPEAKER_02:Oh and I I think the tone of it uh you know when you discussed it later is they were dismissive, rude, disinterested, and also a bit bullying. Would you agree with that? And we don't need that in the arts, do we?
SPEAKER_00:Well, I I I I I The arts are hard enough. I felt that right from the right from the very, very beginning of the casting. Of the casting, um the uh director was intent on keeping me in my place. Yeah. Uh you know, when the when when when the when the the video conference began, I sort of greeted everybody. I'd done a bit of research on everybody, sort of knew their names, got a sense of who they were, uh, and uh and and and said hello. And then the director sort of rested back control and said, Well, you know, let me do the introductions.
SPEAKER_02:Yes, it's very important, David, that you don't overstep the line. Um that's a real status issue, is that's a real power, grab for power.
SPEAKER_00:Grab for power. So yeah, look, it was it it was really disappointing, a bit insulting, and um and uh uh it does not does not make me think that this particular organisation has got a good culture.
SPEAKER_02:Yeah, and I think that's that. And so we we were we were talking about this um prior to starting our acting class that we teach on Tuesday nights, and um we were talking about how we could wreak revenge upon the company and the director. That's what we were talking about.
SPEAKER_00:In in a tongue-in-cheek way.
SPEAKER_02:I mean, of course there's no way I don't mind a bit of it.
SPEAKER_00:And um Well that well that well that was the that was the tone of the conversation. Yeah. You know, you you you uh you were waving swords and shields in the air, we're going to wreak revenge, and I was sort of doing the oh well, you know, you've got to take the good with the bad. But the very lovely centered yoga teacher overheard Annie and then thought it was a good time to give her some um uh a performance conversation.
SPEAKER_02:Okay, so yeah, so we we had this discourse, and she was like, you should forgive. And I was like, Well, should we? Whose rule book is this? I don't want to forgive. And then she was saying, Well, the only person that's damaging is you because you're hanging on to, you know, the the the sort of this the negative emotion. So this, you know, this whole notion of, you know, be the big the person and let it go and and and and and don't hold a grudge, etc. And my point about it was that forgiveness is very, very loaded when it comes to women. Um, you know, we are socialized, are we not? To um smooth the edges and hold the emotional centre and maintain the relationship and also understand where the other person's coming from, right? So anyway, we had this sort of discourse, and also prior to that, which also maybe put me in mind to have an argument, well, a discussion, was that I had done some yoga the week before and the music wasn't working, and she had said to the room, that's because Mercury's in retrograde.
SPEAKER_00:Hmm.
SPEAKER_02:Which is an astrological bit of nonsense anyway. So I was sort of like, mmm, right.
SPEAKER_00:So how did that impact on the music?
SPEAKER_02:Well, David, don't you know anything about Mercury in retrograde and astrology?
SPEAKER_00:I guess I don't.
SPEAKER_02:Well, can I help you out there? When Mercury is in retrograde, it does all sorts of terrible things, one of which is ruins communication lines. Oh, I didn't know that, did you? Okay anyway. So I guess I'd already had a bit of a mmm going when she sort of sort of thrust the idea that I I need to be forgiving. I find the notion of revenge, you know, sort of deeply satisfying having a go back. I don't want people to not be, you know, people should be held accountable for what they do. Absolutely. I don't want to drift around the place like some sort of monk wandering out of a a cave. I don't want that.
SPEAKER_00:No, no, no, and I and I could see Annie had been sort of hooked into this conversation when I heard the question, how old are you? Exactly. How old are you? And the and the yoga teacher provided that information. Then Annie informed her that she was that Annie herself was twice her age and uh and has a different perspective.
SPEAKER_02:I do have a different perspective, and I I as you know, I find these sort of memes around look, of course, there's upsides to forgiveness. I mean, I know that it it certainly Of course if you're hanging on to something that's it's absolutely pointless and that that does clutter up your your brain, of course. I think there's a place for forgiveness. Yeah, but I don't think it should be reflexive. I don't think you should ultimately go, oh, because I think what it is a lot of the time is just women being submissive.
unknown:Yeah.
SPEAKER_02:Just going, I've just got to forgive. No, you don't. You don't have to do anything. I can do what I like. If I want to forgive, I will. I will try and analyse my critical thinking, but I I don't know.
SPEAKER_00:Okay. Well, I mean, what what's what's the damage that's done when when people uh forgive too easily and too soon?
SPEAKER_02:Well, I I I I think people forgive too easily and too soon when they're they want to go towards the old piece at any price. And if you forgive before you've processed what happened, then all that happens is all the the fury and the rage goes underground, right? Also, a lot of this forgiveness nonsense is all is is is definitely um aimed at us from an exterior, you know, there's somebody who's saying to us like it happened to me yesterday, well you should be um, you know, you should move into a forgiving space. And also, um, it can just reinforce a power imbalance. You know, somebody who's been vile, you move quickly into forgiveness. Well, how you know, where's the where's the lesson that they have to learn that they can't do that to you, right?
SPEAKER_00:Yeah, 100%.
SPEAKER_02:So that was that was sort of, you know, that was where I um That's where you were last night. Yeah, that's where I were, and then that's where I was where I were. Um yeah. Look, the thing is that people need to be held account if they've done the held accountable if they've done the wrong thing. And going into some sort of premature forgiveness, um it's is is just is just a way to to go into some sort of conflict avoidance. Okay. It's just avoidance.
SPEAKER_00:So so so forgiving somebody might be right when the process of forgiving them actually makes you feel calmer, you know, not tighter, you know, calmer at the thought of letting it go. Um it could be right when you've made sure that people are accountable for what they have done, you know, even if people's contribution has been imperfect. Um when you are choosing to forgive and not just pretending to forgive or performing forgiveness because that's what's expected of you, and when it genuinely gives you back, you know, energy or space or freedom, then it's the right thing. But it's not the right thing when the process of doing it actually compounds that that feeling of it. Yeah, yeah, yeah, yeah, yeah.
SPEAKER_02:You're you know, th you know, when someone's demanding forgiveness because it's it's proof that you're a really good person, it's just bullshit.
SPEAKER_00:Yeah, yeah.
SPEAKER_02:So that was my thing. So anyway, that happened last night, and then I was telling David about this this morning, and then what I normally do prior to doing the podcast is I ha have my very own opinions, but then what I'll do is I'll have a little bit of a search on AI.
SPEAKER_00:All right.
SPEAKER_02:And just to see what what it says about um you know, whatever it is that I'm talking about.
SPEAKER_00:Yeah, to clarify your thinking, get other perspectives.
SPEAKER_02:Yeah, so I don't just get stuck in my own.
SPEAKER_00:You want to use it as a you want to use it as a kind of a collaborative tool.
SPEAKER_02:I do. Yeah so I I did a bit of I did my usual search on AI, and then I was showing that to David. Um and and what happened was quite interesting. Do you want to describe what happened?
SPEAKER_00:Well, I mean I look I I I noticed Annie's Annie's first prompt. Um, yeah, in into into Chat GPT. Um she wrote, What are the positive and negative aspects of forgiveness? And um and what came out of the out of Chat GPT was um was kind of what you expect, you know.
SPEAKER_02:It's what I it it certainly didn't it's sort of I've been investigating this all this these sort of what I call um social memes for a long time. So I've been investigating, you know, way before Chat GPT or any of the others, I was on to the fact, if I may say, that um that acceptance and gratitude and gratitude had serious downsides. So I've got some pretty well formulated opinions through a lot of research on things like gratitude, acceptance, and forgiveness. So what I'm looking for when I go to um an AI, is it is that what I'm saying? Yeah, yeah, yeah, yeah, yeah. Is I need I want something that doesn't just um give me what I already think.
SPEAKER_00:Yeah.
SPEAKER_02:Right?
SPEAKER_00:Yeah, yeah, yeah, yeah.
SPEAKER_02:I don't there's no point in that. Like I've got some well formulated ideas. It's what I'm doing on the podcast, right?
SPEAKER_00:Yeah, yeah, yeah, yeah, yeah, yeah. And and and and the um I mean there was nothing there was nothing in your first prompt that actually um was groundbreaking or or or paradigm shifting or anything like that.
SPEAKER_01:No, nothing.
SPEAKER_00:And and I think, you know, as as I was watching Annie doing this, I thought um, you know, this is this is this is so common to the way that a lot of people are using AI at the moment. And um I hate AI. You do? Yeah, yeah. That doesn't surprise me at all. I hate it. What what what do you hate about it?
SPEAKER_02:I I just uh I sort of hate its um I I don't like its tone.
SPEAKER_00:Um you don't like its tone.
SPEAKER_02:No, I don't like its tone. I don't like its sort of unbridled optimism. I just don't like it. It's like a really crawly, I guess that's an Australian term for our European and Asian listeners.
SPEAKER_00:So you find it sycophantic?
SPEAKER_02:It's really sycophantic.
SPEAKER_00:Like Annie, you're so smart. What a great idea. No one's not.
SPEAKER_02:It's like one of those employees that always wants to go and get your lunch for you and tell you how fantastic you are, and it's just really crawly and fancy.
SPEAKER_00:I'm just trying to imagine a world in which you would be unhappy with someone saying, Annie, can I get you lunch?
SPEAKER_02:I'm okay about the lunch.
SPEAKER_00:The lunch, yeah.
SPEAKER_02:I'm not okay about someone constantly telling me how great I am. Actually, I quite like that. But can you give that a go by telling me how great I am?
SPEAKER_00:Annie, you're a genius.
SPEAKER_02:Thank you. Do you mean that?
SPEAKER_00:I feel I feel warmed by your being near me. Your your outstanding balance of of perspectives and insights.
SPEAKER_02:Thank you.
SPEAKER_00:You helped me um make sense of this crazy world in which we live.
SPEAKER_02:Yeah.
unknown:Yeah.
SPEAKER_00:Is that doing it for you?
SPEAKER_02:No.
SPEAKER_00:No? Yeah.
SPEAKER_02:You didn't commit to that.
SPEAKER_00:Like I look, I'm I'm I was just paying lip service to it. And funnily enough, that's you know, that's that's the way that AI will work. Um AI's, you know, prime mission is not to um is not to be truthful, um it is to be convincing in the way that it works.
SPEAKER_02:So can we just ignore it? Like let like not have be have it be part of our life.
SPEAKER_00:If you're if you can find yourself a rock to go live under, um then you possibly could ignore it. But whether we like it or not, AI is creeping into, has already crept into so much of the the things that we just do habitually in daily life. Like what? Well, I mean, look, one little one little little test for somebody is to um to go to their applications page, take a screenshot of all the applications that they're currently using, upload that screenshot into a Chat GPT or a copilot or a or a Claude. Are they all LLMs? They're all LLMs. Large language models. Large language models? Gee, you so you say that like an expert. And you can ask it, how many of these are actually using AI in order to deliver the services that I'm currently using? And will it tell you? And it'll tell you. It'll go through, and you'll be surprised how many of the tools that we currently use today have actually got a little bit of AI into them. Now, I I mean Annie's Annie's uh incompetence using AI. Mean? Well, no, look, it's just the truth. It's actually very similar to the way that a lot of um uh uh people in business, leaders, managers, uh are uh are are relating to AI at the moment. You know, I the truth and the data backs up my own um sort of anecdotal experience that a lot of leaders are are performing confidence. You know, they're saying, Oh yeah, I've got this, I'm using it, I'm I'm being more productive. When they're quietly feeling outpaced.
SPEAKER_01:I I don't blame them. Yeah, look, I feel outpaced.
SPEAKER_00:The pressure that is on these people is that I think a lot of boards are saying to the market, yeah, we're acting on AI, we're you know, we're putting things in place, but the evidence actually doesn't match the rhetoric. Um one of the big consulting firms, I think it was oh, was it KPMG or Deloitte? You know, 70% of of boards are saying we're making meaningful steps in um in the in the uh the application of of AI to you know achieving our mission. Um but surveyed only about 30% of the leaders actually feel competent with it.
SPEAKER_02:And and what it's all it's too quick. Everything's happening too quickly. Our pool, we're not our brains, aren't equipped.
SPEAKER_00:Yeah, the w look, 100%. You know, the the the boards are in sort of you know the cycles of uh of of of quarterly or or or half-yearly um uh strategy adjustments, but AR is actually changing every day. And so I mean people are by by you know reality's rules, they are ignorant of what's coming down the pipeline because everything's changing so quickly.
SPEAKER_02:Exactly. It's bad.
SPEAKER_00:And that hesitation bad. You know that that that that hesitation is understandable. You've got very capable people just trying to steady themselves while the ground is shifting between you know beneath their feet. Um and what is emerging is a reality that AI isn't exposing technical gaps first. It's actually exposing leadership gaps.
SPEAKER_02:So it's not exposing gaps. Technical gaps. Okay, so it's not that I'm technically incompetent, it's that I'm not leading.
SPEAKER_00:Um Well, it uh you're you you don't have the competencies that are uh that are required for you to lead.
SPEAKER_02:And you know why do I need to lead AI? Well why can't isn't it isn't it capable, it seems it's pretty bossy.
SPEAKER_00:Uh look, look, look, AI is simply, um, you know, speaking very, very generally, it is a pattern-seeking, pattern-generating machine. It seeks patterns to learn um, you know, what it is that it's dealing with, and then it just continues those patterns as a way of saying, you know, this is what's going to happen. So if I was to say to you, the quick brown fox jumped over the lazy dog. Do you want to try that again? Yeah. If I was to say to you, the quick brown fox jumped over the lazy dog. Now, did the quick brown fox actually jump over the lazy dog, or did you just finish that sentence because that was just predictably, you know, statistically, the right word to finish that sentence? This is the right word. It's the right word. Okay. So that's the way that AI works. Like night and day. Up and down. Black and hot and cold. Yeah. Yeah. So lovely antithesis. It it gets a a it it recognizes a pattern and then then it simply completes the pattern. And so don't expect AI to have any ethics, any values. Um don't expect it to understand context. I mean you're able to finish the quick brown fox jumped over the lazy because you've got that context. If you want AI to do something to do something in which it will give you back what it is that you want, you need to direct it as precisely as a theatre director directs a performer. You have to give it a character and context. You have to get clear on what the intention is.
SPEAKER_02:Is that what you're saying?
SPEAKER_00:Yeah, that's right. Okay. So so a lot of people, you included, talk to AI like it's a machine that that is going to give them an answer. But it's not how the systems behave. They don't answer questions, they imitate answers to questions. They they take shape around the direction that you give them. And the closest, you know, analogy that we have is the director-actor relationship. When you direct an actor, you don't you don't you don't tell it what to do. You don't dictate emotion. I mean, you might if you're a a bad director. A bad director, a high school director. But if you're a professional director and you want to be working with respect with a professional performer, you will have a conversation and you'll set the intention. You'll give the the actor context for the scene. You'll break the work into beats, you'll create a space where the performer can discover something real for themselves.
SPEAKER_02:But the AR's not like an actor's not real. No, no, no, no, it's not. It hasn't got a beating heart.
SPEAKER_00:Look, it it it is it is as real as an actor, insofar as that it has strengths and weaknesses. Um it will do what you want it to do, like you know, like can what's the difference between between an actor, you know, dealing with a difficult text, you know, uh a difficult character like Richard III or or or Porsche? Um an actor that has direction and an actor that has no direction, just instinctively. What's the difference between the performances?
SPEAKER_02:Of of a vague generalized performance.
SPEAKER_00:If you don't give them good direction?
SPEAKER_02:Yeah, and if you give them actors are terrible without direction.
SPEAKER_00:Yeah, actors actors are terrible with bad direction. AI is terrible and embarrassing.
SPEAKER_02:And um Yeah, but how do you know? I don't get it. Like, how do you then know wh which bit which one of the LLMs to go to?
SPEAKER_00:All right. Look, uh I mean that's an interesting question. Which one should I choose? Should I choose ChatGPT? Should I choose Claude? Should I choose Perplexity? Should I choose Gemini or Pope? I mean, okay, so AI is across all of these things. Um my advice, look, there's two schools of thought. One advice is to pick one and stick. So go deep and wide with with open AI and and Chat GPT and custom GPT.
SPEAKER_02:Why would you do that?
SPEAKER_00:Because the deeper that you go with an AI Oh, the more it gets to know the context. Well, the the more it gets to know your context, um, and the more it gets to know uh what you expect of it. Okay? So if if you um wish they'd never bloody invented it. Look, one way of describing what we've done, uh, you know, for many years we've been worried about uh sort of alien intelligence coming in and taking over the planet Earth. What humanity has done collectively is that we have we have we have manufactured that intelligence. You know, we do have an alien intelligence, it's it's in action on the planet at the moment. People are using it and profiting from it.
SPEAKER_02:It can't do everything.
SPEAKER_00:And it can't do everything.
SPEAKER_02:I tell you what, it can't do.
SPEAKER_00:What can't it do?
SPEAKER_02:This morning, when I was on um a walk back from the beach, there was a boy standing next to a car, and it was a pea plate, which is a provisional license here in Australia. And across the road from him were two fire trucks, two fire engines, and standing around the car were five um fire and rescue um people.
SPEAKER_00:Okay, I can picture that.
SPEAKER_02:Yeah, that were trying to break into his car for him with a coat hanger. Oh, good, right, okay. And they were all there, and there they were, and I thank God we're an extraordinarily well-resourced part of the country, aren't we?
SPEAKER_00:Yeah, there's a there's two fire trucks and a and a and a wire coat hanger.
SPEAKER_02:And a wire and there they all were, and he was standing there, and it was quite a jolly little it was a very sweet scene. I really appreciated you know living in Australia at that moment. But there's something that, you know, no AI can't do that.
SPEAKER_00:What can't open a car with a hangover, uh a coat hanger?
SPEAKER_02:No, it can't.
SPEAKER_00:Okay. That's that's tremendously insightful.
SPEAKER_02:Uh I know.
SPEAKER_00:It can't do anything physical. Do you know it will though one day? And here's the other here's the other scary thing that's on the way.
SPEAKER_02:But will it want to? I mean, they were they sort of looked engaged. Will it want to help and they wanted to help and they were awesome.
SPEAKER_00:No, it um okay, so so robotics and AI is is really mind-blowing. So we're now building autonomous robots that can move around the planet and I think could probably manipulate a wire coat hanger into a broken down along the way.
SPEAKER_02:But would they would they ever have the empathy to want to help?
SPEAKER_00:No, of course not. That I mean okay, so this is the point. Um artificial intelligence, you know, for all its for all its strengths, it has no human emotion whatsoever. It has no no no emotion, no ethics, no value. However, you can provide it to them. You can give it to them in the prompt. A lot of people talking about prompt engineering at the moment.
SPEAKER_02:And uh because you changed the prompt for me this morning.
SPEAKER_00:Oh, yeah, yeah, yeah.
SPEAKER_02:And that's how we started off on this. You changed the prompt, didn't you? I had that I had Chat GPT and I thought it was pretty pedestrian.
SPEAKER_00:Yeah, yeah. Because you simply asked the question, you know, what's the difference between And I was hoping it would give me a new thing and it didn't. Yeah, yeah, yeah.
SPEAKER_02:And then what did you do? You changed the prompt.
SPEAKER_00:Well, I mean what I did was I actually applied what our approach is, and that is to treat artificial intelligence like it's a performer. So the first thing that we did was I I cast the character of the AI. I told it that it was playing a psychologically literate, warmly pragmatic, insight curator. A guide who understands humanity.
SPEAKER_02:Now, where did you come up with that language?
SPEAKER_00:Where did I come up with that language?
SPEAKER_02:Yeah, because that's very specific.
SPEAKER_00:Yeah, it look it is very specific.
SPEAKER_02:Psychologically literate, warmly pragmatic, insight curator. Where did you come up with that?
SPEAKER_00:Can I keep going and then I'll tell you where I came up with? Please do. Because what this what this prompt is giving at the start is a character. Right? So this is this is what you are, and you understand human behavior, emotional dynamics, and relational power. You think like a coach, you speak like a storyteller, and you communicate like someone who knows confidence and clarity are acts of self-protection, not self-judgment. So basically, I'm I'm I'm giving the AI a character and a backstory and a perspective. And I give it a worldview. And the worldview is that forgiveness is complex.
SPEAKER_02:You gave it that.
SPEAKER_00:Yes, I gave it that. And so that's the character. Um I told it what the intention was.
SPEAKER_02:Hang on, can we scroll up again? Yeah. Where did you I I'm interested in that language at the top because it's very Okay.
SPEAKER_00:So I I would say that um I've moved beyond beginner level when it comes to working with AI. And when I've got a um a question like the one that you're asking, um, I know that to get a good result, I need to provide it with a character, an intention.
SPEAKER_01:Yeah.
SPEAKER_00:Um I need to, you know, give it the structure that it's doing, I need to give it performance notes, I need to give it constraints, um, you know, rehearsal logic. Yeah. I mean, I'm I'm basically using the same concepts that we use when we are directing actors. Directing actors.
SPEAKER_02:But so you came up with psychologically literate.
SPEAKER_00:Well, actually, I didn't. Oh. I didn't in this instance. Because what I have done is I have built a custom GPT. So again, for those who who don't know, you can you can use the the standard uh open AI chat GPT. You can also build an assistant that will do things for you. And I have built an assistant that helps me to write prompts that work.
SPEAKER_02:So how hard was that? So for our listeners, how how difficult was that process? Because it I you've been locked in this studio as far as I could see for about six weeks working on this, right?
SPEAKER_00:Yeah, yeah, yeah.
SPEAKER_02:From dawn till dusk. Occasionally I throw a glass of water at you and some dry bread. Yes, which is with a delicious dinner at night, though, may I say.
SPEAKER_00:And I'm enormously grateful for this.
SPEAKER_02:And thank you.
SPEAKER_00:Yeah, yeah, yeah, yeah. So um if if uh if if if your knowledge of uh of working with AI says you need to get your prompts right, well, you know, here's the process that you can go through. You can you could go to, let's say we're using ChatGPT, but once again all of them are all of them are going to behave somewhat similarly. Um you'll go to Chat GPT and you'll say, Okay, I want you to help me um write effective prompts, you know, effective prompts that do this, that, and the other. And then it will give you the um the character and the intention and the background. You it it basically gives you the prompt to write the prompts.
SPEAKER_02:Yeah, I get it. So the psychologically literate, warm, warmly pragmatic, insight curator who thinks like a coach, speaks like a storyteller, communicates like someone who knows confidence, it understands these worldviews, forgive forgiveness is complex, empowering in some circumstances, costly in another, the scene. Yeah, it's it's like it is sort of like looking at a play. Yeah. It it is like a play, it's really interesting. I would never have thought that because I've ignored you for the past six weeks because I'm you know, sort of um organically or instinctively bored by the whole thing. But this is interesting.
SPEAKER_00:If you if you just sort of tone down the artificial in artificial intelligence and actually appreciate that what we're working with is unintelligence.
SPEAKER_02:And intel and intelligence.
SPEAKER_00:It is an intelligence.
SPEAKER_02:Yeah.
SPEAKER_00:Um and and to get the best out of unintelligence and intelligence. And intelligence.
SPEAKER_02:You're saying unintelligence.
SPEAKER_00:And intelligence. To get the best out of your out of your LLM, um you want to give it instructions so of the limitless directions that it can go in, it goes in the direction that you want it to go in. So yes. Um I'd be very happy if if if anybody wanted to reach out and say, um, you know, could we have access to your prompt generator? Oh, yeah.
SPEAKER_02:Anybody who needs help with this, because I've got a live-in helper here who knows stuff and sort of uh um can sort of override my antipathy. Um so if if you want any help, do do get um to let us know. Um because we can also, David can also sort of talk to you from anywhere in the world, even though we're in Sydney, New South Wales, Australia, because he is really good at this. I I don't mean to sound surprised, but I have ignored it because uh because you know it's what I'm like.
SPEAKER_00:Yes. Um but I think that you're going to become um more interested as better prompts give you much better uh results in terms of the kind of research that you want, uh the kind of angles that you can take, the way that you can shape arguments and things like that.
SPEAKER_01:Okay, that's that's that's true.
SPEAKER_00:I didn't quite I I I didn't actually finish the two choices clearly spoken in this episode.
SPEAKER_02:Go on.
SPEAKER_00:Oh well I okay.
SPEAKER_02:No, go on, do, go, do it, do it.
SPEAKER_00:Well let me do let me let me let me let me just finish this. Okay. One school of thought is that you get to know one LLM really, really well.
SPEAKER_02:Yeah, is that true?
SPEAKER_00:Yeah, yeah.
SPEAKER_02:Look, I I Which is your favourite then?
SPEAKER_00:Okay, so the LLM that I've got to know really well is the the OpenAI, so the the ChatGPTs. It was the first one that I started with. Um they have continued to evolve and add new features to keep me interested. However, I have gone fairly deep and wide with perplexity, and that's fantastic for research. Um I do have a relationship with Claude, uh, who helps me generate code. Um I do work with Copilot because Copilot is sort of right inside the Microsoft suite, and I do look at how we can get insights out of Excel and you know put them into PowerPoint and those sorts of things. So uh and also um Gemini, which is the uh the Google one.
SPEAKER_02:That's a lot. That's that you just sort of one and you just mentioned six.
SPEAKER_00:That's right. And so this is this is the alternative point of view. Uh go deep and wide with one, or or you can do what I've done, which is to go as deep and wide as you can with one in particular, but find out what the strengths of the other ones are. So it's not that you've just got one actor that you are now you've got six. You've got an ensemble. You've got an ensemble cast. An ensemble cast. For opening nice. That's right. So my my um my research prompts I will generate in Chat GPT. I will put those research prompts into Perplexity, and that will do the research. I might consolidate all of that research into a single document and put that into Claude and then ask Claude to build me a um, you know, an HDML interactive dashboard um that helps us work through a process. So you can go from one to one to one. My suggestion is that you don't do that until you have built a solid relationship with your your foundational LLM. So the challenge here is to find a way to work with AI every day that that feels like it's an extension of your capability, an extension of your leadership. Because working with AI starts with fluency, not expertise, right? You don't need to learn how to code, you just need enough literacy to ask the smart questions, spot bad assumptions, uh, and continue to drive towards excellence. I think that because some people have seen just how quickly AI can work and the um the outputs immediately appear very, very uh convincing and com compelling and credible, um, they start to think that it's a it's a quick process. It can definitely be quicker um when you are automating the research tasks and and you know using the incredible reach and depth that an AI can go to. Um but to think that you're gonna get it in one go is as stupid as a director thinking he's gonna get a great performance out of a cast with one run-through of the play. You need to iterate, you need to spend time on it. Um I would suggest that you spend time thinking about what your workflow is going to be. Um pose the question, do the research. You know, take that research, you know, draw from it a report or some conclusions or some findings or the skeleton structure of a PowerPoint presentation. Um, you know, taking some of the ideas into a into image generation or video generation and and and bring your ideas to life. You can get a workflow that can be very solid, but it's it it is about repetition. Don't worry that you don't feel you're in control of the the whole thing. Just start, set little short-term objectives, like um, you know, I want to I want to improve the quality of my prompts and learn how what you put in your prompt is going to directly impact your outputs. So um, yeah, I mean, that's the most practical advice I can give, you know, in this format around how to how to really start working effectively with artificial intelligence.
SPEAKER_02:We all know that there's shocking bias in AI in terms of hiring decisions in companies, in terms of healthcare. It's distinctly um, you know, puts women at a dis at a disadvantage.
SPEAKER_00:Yeah, yeah. It puts any demographic that is not the mainstream, not the core, at a disadvantage. Because again, think about what I was saying earlier about what what AI fundamentally is. It's uh it you know, it's autocorrect, it's it's predictive, it's statistical. And when searching for a you know the best candidate, it will look at who have been the candidates selected in the past. And it'll use historical data. Historical data taken from a time when gender biases were.
SPEAKER_02:Yeah, so m men are the scientists and women are the nurses.
SPEAKER_00:That's right.
SPEAKER_02:And and and also and also if it goes back, you know, historically in terms of medical research, we we know that medical research is absolutely geared towards men and not about the specifics of women's, you know, symptoms and the difference between the biologically between men and women. We know it's it it's been way more research to do with um to do with men.
SPEAKER_00:That's right. And also and also um, you know, uh people of colour have drawn the short straw as well around medical things.
SPEAKER_01:Oh, terrible.
SPEAKER_00:You know, and uh, you know, uh uh you know, quite specifically around skincare products and and things like that that fail to take into consideration the the different effect of having you know more pigmentation in your skin.
SPEAKER_02:Exactly, exactly.
SPEAKER_00:Yeah, w when you do a um when you do a search, you know, give me an image of um you know leaders discussing strategy, I guarantee that you'll probably see more men around the table than there will be women. And they will probably all have beards and they will probably be all of a of a certain age.
SPEAKER_02:How old is that, like 42?
SPEAKER_00:Oh yeah, yeah, roughly. I mean that's your you know it it gives you the That's your prime age, right? That's right, it gives you the mean. And I can tell you that if there are any women in the photograph of leaders sitting around the table talking strategy, they are probably, you know, about thirty years old, slim, very attractive, and they'll have straight hair. Yeah, there won't be any interesting sizes and shapes.
SPEAKER_02:Um what about will they have hair like me, curly hair?
SPEAKER_00:No. No, no, no, because curly hair um is um is is is wild and creative and non-strategic.
SPEAKER_02:Yeah, yeah, that's right.
SPEAKER_00:That's right. I mean, you know, someone with curly hair.
SPEAKER_02:Like I'm a wild I I I I should be like a wildly artistic, deeply disorganized sort of maverick, shouldn't I, really, with my hair.
SPEAKER_00:Show me a portrait of the HR team celebrating um at the annual general conference. That's where you'll turn up.
SPEAKER_01:W really?
SPEAKER_00:Yeah.
SPEAKER_01:Really?
SPEAKER_00:Yeah, women with long curly hair.
SPEAKER_01:Okay. Right. Okay.
SPEAKER_00:And th they'll they'll probably be um, you know, uh a bit diverse in their look as well.
SPEAKER_02:Yeah. Yeah. Anyway, it's really annoying. I mean, I I do get your point. Uh this actually I I I regret that I've just left you to sort of moulder in here for six years while I ignored you. Yeah. I did keep the food up, didn't I?
SPEAKER_00:You did keep the food up. Look, I uh Annie, what I want you to know is that AI is not a threat to your life, right? It's not a threat to all of the things that you are competent in. In fact, I reckon it's a spotlight for it. Because I think it's the skills that many women innately have. You know, that that that sort of natural emotional intelligence, you know, clarity about what's important, the capacity for nuance, um, being aware of of relational issues. This is exactly what organizations seeking to harness AI actually need. There's nothing in the data that suggests that brusque certainty is the winning trait of the AI era. And it's actually quite the opposite. It's the leaders who can communicate clearly, humbly, stay grounded under pressure and collaborate with with with real boldness, they're the ones who are going to be shaping the future.
SPEAKER_02:Oh, shaping the future.
SPEAKER_00:And I do think that um that they're I think that where women leaders are demonstrating strengths, those are the strengths that are gonna get the most out of AI.
SPEAKER_02:Okay. All right. I forgive you for spending six weeks in here working that out.
SPEAKER_00:And I forgive you for leaving me to moulder.
SPEAKER_02:Mulder. You moulded in here, didn't you?
SPEAKER_00:Okay, call me Fox.
SPEAKER_02:You even you even told me to leave you alone, like frequently. Oh, can you please not interrupt? But it was necessary. Yeah, but it's still quite gendered because I'm also doing stuff. I tell you what, uh when I've put my book that I'm currently writing and said, you know, I've put it in to sort of look at structure, the stuff it gives me back is rubbish. It's so awful. It's like the worst romance novel.
SPEAKER_00:Okay.
SPEAKER_02:People are kissing the downy heads of babies. I ought to shoot myself with dreadful.
SPEAKER_00:So let's let's see if you've been listening. Um when when when AI gives you a really bad bad prompt. No, a really bad output, a really bad result. What do I do? Is it no, no, not what do you do?
SPEAKER_02:Oh, I know why.
SPEAKER_00:Let's assign blame. Is it its fault or is it yours? It's mine. Yes. Yeah, you've really fault. It's your fault, and what could you do instead?
SPEAKER_02:I could give it a better prompt.
SPEAKER_00:You could well, you could give it a better prompt.
SPEAKER_02:I could I could develop my own helper assistant that helps me with good prompts.
SPEAKER_00:Yeah, okay, so now imagine you're directing an action.
SPEAKER_02:I can't do code. Imagine. I'd give it a character.
SPEAKER_00:Okay, okay.
SPEAKER_02:So I'm I'd give it a context and I give it a character and I give it a worldview.
SPEAKER_00:Right, I'm your LLM.
SPEAKER_02:Yeah.
SPEAKER_00:Right? You want to do you want to do some you what do you want to do? Do you want to do some research? Do you want to do you want a sample paragraph? Do you want a poem? Do you want music?
SPEAKER_02:I want to I I want um I want to segue from one incident in my novel. I want an idea for a segue into another.
SPEAKER_00:Okay, alright. Oh, okay, okay. So you're looking for editorial advice?
SPEAKER_02:No.
SPEAKER_00:No?
SPEAKER_02:No. You're looking for it's a narrative leap.
SPEAKER_00:A narrative leap. Okay, so um I'll be your uh I'll be your your avatar, I'll be your LLM. In order to in order to to actually give you what you want, what would I have to be? Well You want a narrative device, yeah?
SPEAKER_02:Yeah. You'd have to be um you'd have to have a good grasp of the language, you'd have to be articulate, you'd have to understand the context of the story that I'm writing.
SPEAKER_00:Okay, okay. Now slow slow down. So what's the role? What am I? Uh so Annie, imagine that you're you're preparing your prompt and you are prompting an actor. I'll be your actor. What are the things that you would tell me if you wanted me to perform what it is that you want me to perform? What would you have to do?
SPEAKER_02:Well, I'd have to cast the the the character.
SPEAKER_00:Yes, yeah. So who am I?
SPEAKER_02:You're a you're literary and you're a storyteller.
SPEAKER_00:Yeah, and what are my what are my values or what's my world view? Um What kind of storyteller am I?
SPEAKER_02:Um w well you're you're not a it's not mechanical. You're a you you um you're an emotional storyteller who takes what's what I you know, you takes people from one state that they're in, the reader, to another state, sort of. Yeah, yeah.
SPEAKER_00:Okay, so I mean it sounds like that I'm a I'm a literary storyteller. Um You're articulate. Yeah who who is articulate and has a feeling for emotion. You know, the emotion is a few years. And and I and I and and I think that transitions should actually have emotional fidelity. Yeah, yeah, yeah.
SPEAKER_02:And then what I want you to do is help me, the author, uncover a whole lot of creative options for for segue between two scenes about a wild birth.
SPEAKER_00:Right here. So that's what I am. Yeah. That's what I'm going to do.
SPEAKER_02:Yeah.
SPEAKER_00:Um is there are there any things that you want me to bear in mind as I'm doing that? You know? Are there are there any things that you don't want me to do? Any things that you want me to leave out?
SPEAKER_02:Uh well, I I don't want well, I don't want that crappy cliche business. I hate all that. I don't want cliches about birth. Quite yo, yo. Okay, so uh I don't want you and I don't want you to overexplain. I don't want it to be overwritten.
SPEAKER_00:Yep, yep.
SPEAKER_02:And I want to sort of, you know, it's sort of intuitively uh, you know, that you should be guided by the im let the imagery guide it, not I just don't want it to be mechanical.
SPEAKER_00:Yeah, yeah, okay, okay. And and looking, you know, we could go further, but I'm starting to get the ingredients of a prompt.
SPEAKER_02:And that's what I would need to put in. Yeah. As and what I was putting in was give me a segue in my book about a wild birth.
SPEAKER_00:Yeah. It's like saying to an actor, just act better.
SPEAKER_02:Ah, yeah. Yeah, that's the worst note.
SPEAKER_00:Yeah, do it do a monologue in an interesting and um and and vaguely comedic way.
SPEAKER_02:Yeah, that's a be funnier.
SPEAKER_00:Yeah, be funny.
SPEAKER_02:Can you be funny? Yeah, I've had that.
SPEAKER_00:It's a bit like that casting that I went to yesterday.
SPEAKER_02:Yeah.
SPEAKER_00:You know, it's just like be this, and it's like a be what? Be what, exactly.
SPEAKER_01:Terrible prompts.
SPEAKER_00:And look, this is this is actually a a point not to be missed. Yeah. That the same communication skills that a director uses with an actor, yeah, that a um a prompt engineer or a leader or a manager uses when prompting an LLM, they are exactly the same skills. Dressed up differently, but they are the same skills that that we employ when we are simply d drawing a performance out of another human being.
SPEAKER_02:You you you have not wasted the past six weeks I commend you on your analysis because I mean I never ever would have come to that conclusion that you know you can if you're prompting AI, you've got to treat the AI like a director teachers and actor artists would never have come up with that. And I can I can actually see, I don't mean to be daggy about it, but I can actually see how that could help me.
SPEAKER_00:Yeah. And look, look, look, the best evidence that something works is something working for you. Yeah. So Annie, yeah, I am going to encourage you to start experimenting with AI and directing AI in my absence.
SPEAKER_02:Well, where are you going?
SPEAKER_00:I'm gonna go moulder somewhere else.
SPEAKER_02:Go moulder somewhere else. I'd quite like to go to the maildives.
SPEAKER_00:Muldering in the maldives.
SPEAKER_02:Muldering in the maldives. I'd quite like that. Anyway, that's very good. That's very helpful. Thank you, David. You're welcome, Annie. Well done.
SPEAKER_00:I hope I hope that was useful.
SPEAKER_02:Well done. Let us know if you want any help, David. It's very helpful.
SPEAKER_00:There is a um there is a uh a webs a web page that that includes a lot of this uh in in more detail and time to go through it in your own time if you'd like. And uh we'll put that in the show notes. Harry will put it in the show notes. It's I mean it's simply ku.college forward stroke directing AI. Directing AI at ku.college.
SPEAKER_02:Um I'm finishing now.
SPEAKER_00:Okay, okay. And I will be quiet.
SPEAKER_02:Good. Um, thank you so much, smart women, for tuning in. Um I hope that was helpful. I actually found it helpful. That's for sure. So I hope you did too. So, wherever you are in the world today, stay safe, stay well, and keep your critical thinking hats on. See you later.
SPEAKER_01:Bye.
SPEAKER_02:Thanks for tuning in to Wise Smart Women with me, Annie McCubbin. I hope today's episode has ignited your curiosity and left you feeling inspired by my anti-motivational style. Join me next time as we continue to unravel the fascinating layers of our brains and develop ways to sort out the facts from the fiction and the over 6,000 thoughts we have in the course of every day. Remember, intelligence isn't enough. You can be as smart as paint, but it's not just about what you know, it's about how you think. And in all this talk of whether or not you can trust your gut, if you ever feel unsafe, whether it's in the street, at work, at a car park, in a bar, or in your own home, please, please respect that gut feeling. Staying safe needs to be our primary objective. We can build better lives, but we have to stay safe to do that. And don't forget to subscribe, rate, and review the podcast, and share it with your fellow smart women and allies. Together we're hopefully reshaping the narrative around women and making better decisions. So until next time, stay sharp, stay steady, and keep your critical thinking out shiny. This is Annie McCubbin signing off from White Smart Women. See you later. This episode was produced by Harrison Hess. It was executive produced and written by me, Annie McCubbin.
