In this episode of Humans WithAI, David Brown is joined by Joanna Shilton for a wide-ranging conversation about how AI is changing work, creativity, bias, image generation, critical thinking and the way people use technology in daily life.

Joanna reflects on what she has learned since hosting Women WithAI, including why AI should be treated as a tool rather than a source of truth. The conversation explores whether AI gives women a competitive advantage, how it may affect entry-level roles in PR, marketing and communications, and why younger workers could lose important opportunities if AI replaces early-career training.

David and Joanna also discuss AI bias, generated images, ownership, social media algorithms, the limits of large language models, and the need for humans to stay actively involved in judgement, editing and decision-making.

At the heart of the episode is a simple question: if AI reflects the information we give it, what responsibility do we have for what comes next?

AI isn't sentient. It's not a person. It doesn't have the experiences we have. It's only learning from what it's been given.

It's only learning from what has been pushed into it. And as much as we think it's now, you know, it is very clever, and no one really understands how it works.

It will take different viewpoints that different people have fed into it, but it's not necessarily going to give you the answer. I think that answer is still within you.

Well, hello. Welcome to Humans with AI.

I'm your host, David, and today I have Joanna Shilton with me, who, as you will know if you've listened to any of the past podcasts, probably was in season two was the host of Women with AI. Good morning, Joanna.

Good morning, David. How are you?

Yeah, I'm really good. It's been. We took, what, nearly five months off? Yeah, I guess between kind of the last season. That's kind of how I'm thinking about it.

Like, season one was creatives, and then season two was all the different channels. And season three, we've kind of just bringing it back together a little bit because it got a bit excessive there for a while.

Yeah, I think you're a bit snowed under. We had a lot going, a lot of shows.

I mean, I loved it, I enjoyed it, and it was easy for me because I just, as you said at the beginning, you'll just do the talking and ask the questions, which, yeah, we know, isn't always the easy bit, but you were obviously doing all the editing, all the putting it together, all the pushing it out there, all the clipping and all the rest of it. You should probably get an AI agent to do that, David.

I should, like, get a studio or something that can do it for me. Yeah, yeah.

Well, that's what you were busy doing, so, yeah, congratulations to you as well, because you are doing phenomenally well now with the studio.

Yeah. Thank you. Yeah, we. It. That was part of it is that I was, you know, desperately trying to get the studio up and running and.

And to be honest, I just had burnout on talking about AI. I just kind of. It. It just had reached a limit, and it's all I ever seem to talk about with anybody.

And I was just like, I just need to focus on other things.

Anyway, we're back now and it's five months later, and I've had a little time to reflect, and I know you probably have had some time to reflect as well. So how are you feeling about AI These Days.

I mean, I'm probably feeling. I'm feeling about the same about AI as I was before.

I think it's interesting seeing how other people are feeling about it and how other people are kind of accepting it, because I think as well, there's loads of stats out there, and obviously I'm going to generalize here, and I haven't got the stats to have, but not as many people are using AI as we think they are. It's only because we're interested in it. We are using it. You've been doing the podcast.

I mean, I, for one, I wouldn't have been as interested if you hadn't given me the opportunity to be the host of Women with AI. And I think it's really interesting because obviously I've gone down that route or I was going down that route.

And I do still, you know, I do a lot of my kind of senses are kind of, you know, attuned to what's happening with women. And I think the fact that you have women in business, women in tech, women in stem, there is still.

There's a big, huge chunk of people or that are kind of, you know, as a society, we really need to be looking at women with AI and how they're portrayed and how they are. AI is talking about them just like any demographic.

But I think I'm still surprised that people think that this magic wand has been invented and now they can just use it and it will do everything they need it to do. I think that isn't what it's here to do. It is like, I think the third person I spoke to ever, Julia, and The women with AI1AI is like an intern.

I think it's someone that you bounce ideas off. Not saying that you bounce ideas off an intern.

You normally bounce ideas off your friends, your peers, people that, you know might be more clever than you or kind of doing the things that you want to do.

And I think it's a bit like the advice that I would give anyone or that I once received that only take advice from someone that you respect or that you kind of think, well, actually, especially say you're looking for relationship advice.

Do you go and ask someone that's in a terrible relationship and you always think, goodness me, I don't know why they're with that other person or what they're doing with their life or why they're being treated like that. Would you take advice from them? Maybe you do. Maybe they're your friend and you.

You advise them, they advise you, but you take it With a pinch of salt. And I think it's. It's searching out and finding the advice you want. But I think also AI can sometimes just give you the answers you want.

So I think it's quite tricky to sometimes know. And I think that's where anyone that doesn't really understand anything is going to get tripped up and caught by it. But I think it's.

Everyone's got to kind of keep an awareness that AI isn't sentient. It's not a person. It doesn't have the experiences we have. It's only learning from what it's been given.

It's only learning from what has been pushed into it. And as much as we think it's now, you know, it is very clever. And no one really understands how it works.

It will take different viewpoints that different people have fed into it, but it's not necessarily going to give you the answer. I think that answer is still within you. But AI is a tool.

But that's the same as a human, isn't it? Well, yeah, we take input from all different places and then we spit out something that we think is whatever, and it could be right or wrong or.

Exactly. That's why I say AI isn't the answer. I think it's. What are we looking for? What's the question?

Well, yeah, we're back to Hitchhiker. The answer is 42. We know that already. We just don't know what the question is. But, yeah, it.

But, yeah, no, and this is what I've been saying all from the very beginning of when I started thinking about this is, you know, AI is just like a human in that respect. Right? I mean, that's all we do. We learn from other people. We read from books. We see what we see on TV. We. We watch idiots on YouTube.

We, you know, we see. We get information from our friends, from our family, from newspapers and everything.

And then we munch that in our brain and then we spit out some kind of thing that we think is, you know, sort of original. It's exactly the same thing that AI does. And I. I'm not. It seems just blatantly clear to me that it is exactly the same as a human in that respect.

And people swear up and down that it's not, but it is at its most basic, we can only make decisions off the information that we have. The thing is, AI has a lot more information than we do, and so it comes to slightly different conclusions than we do. But that's.

Maybe that's an okay thing. So What I was going to ask is.

So going back to when we first started talking a couple of years ago about this, I had two kind of questions about particularly AI's impact with women.

One was obviously that a lot of the jobs that were seemingly being taken by AI or the AI was having a huge impact on were roles that were traditionally kind of more populated by women than men.

So if you think like marketing, pr, a lot of the creative work and stuff like that, particularly writing and those sorts of things, were very heavily female anyway. And that seemed to be the big industries where AI was making massive inroads.

So I'd like to get your thoughts on where you think that is and how that's played out over the past couple of years.

And then the other one was really, I was wondering and wanted to explore the idea that women could use AI as a competitive advantage in business and stuff like that.

So it may actually give women more of an advantage, again, maybe in those roles, or if they wanted to be more, just more competitive in the marketplace, maybe against men or other women or whatever. And would that, you know, again, would that be helpful or could it be used and could it be helpful for women?

And yeah, I'm interested to know what you think about that as well. So what do you think?

Go,.

I'll turn my microphone off for 10 minutes and you can just rant.

I think I, I think you've got to use it. So women, you've got to use it. Men, you've got to use it. Men probably are using it more than women.

They were the stats, you know, a couple of years ago. Again, I'd need to go back and look at that. And I think it does give you an advantage.

And I know in equal parts, you know, men and women that I work with that have both said, no, I don't want to use it. I think it's going to stifle my creativity. And I don't want to fall down a rabbit hole of getting lazy or of not learning.

But I think that's where, again, it's your choice. You've got to see it as a tool, you're using it. How else do you learn? You do the research, you, you know, nowadays you Google it.

In the old days, you get out an encyclopedia or, you know, you trawl through the newspapers. Now you can just ask ChatGPT or AI, you know, whatever your other AI is available, Claude, to, to go through and kind of give you those stats.

But again, you've got to then fact check it, which they're still doing in schools. I think that's still a big part of teaching, you know, younger people that you don't just believe everything it gives you.

Just like you don't believe everything that a person, you.

And I think, yeah, there are a lot of advantages to women, but I think it's still, it all comes down from the top and what it's being fed to begin with. So it's that constant battle. And I think if you're coming up against something that is very heavily against you, you might not want to use it.

And I'm, you know, thinking back to some of those conversations I had where it was about bias and it was about, you know, searching for a picture of a professor, it's a man, search for a picture of a cleaner, it's a woman. And you know, and it's not just kind of the gender thing, it's the kind of, you know, the race question as well.

And kind of, you know, is black person, is a white person, you know, what, what ethnicity are they? And it's again, it's the tool that you have to use.

So I think there are a lot of advantages to be had, but you've got to be willing and open to want to use it. And I, you know, I hope that people are using it. I use it, I use it in my day to day life. We're allowed to use it at work.

We use Copilot, that's the one that has been kind of signed off and approved. And there are regular weekly kind of sessions on how to use it and what it can do for you.

And it's quite interesting because the ones that I've joined, it has been mainly that are on there, but whether that's whether there's just, you know, a high percentage of women that are working where I am. But.

But you work in a comms team.

Yeah.

Which is exactly what I was saying, right, is that those roles traditionally have been highly populated by women. And so again, do you know what I mean? So are you seeing any?

I mean, I've talked to a few people in prison that I know who a few years ago were vehemently of the, of the opinion that PR was a personal game. Right? It was. You needed personal contacts, you needed, you know, those sorts of things.

It needed somebody to be able to do it and that it, you know, the AI wouldn't really affect their business too much.

And then I've now come back to them a couple years later and they said basically they don't hire interns anymore because they don't need them and they're actually more trouble than they're. They didn't say that. That's unfair.

That was the subtext of what they said is that it takes more time and effort to work with an intern than it does for their people to just use AI to do some of those basic low level tasks that the interns would do.

Yeah. Because the intern think about it, wants to progress and go, they're going to move on and got to start again. Whereas with AI you're just.

But there's. Yeah. So they, they had essentially. I, I won't quite.

Well, I'll, I'll try and get this person to come back on and talk about it and see, see if they'll talk about it openly.

But I think it was about 14 interns that they weren't going to do this year that they would have done normally and they just didn't need to do it because their, their normal, you know, sort of PR agents and whatever could just handle the business and it was easier for them to do it themselves. The. And this again gets back to okay, well that seems okay a little bit maybe. But then you just get a constant erosion. Right.

Because those interns aren't going to get trained up.

They're then, you know, how do they progress to then becoming actual employees that are like a full time employee that, you know, is a PR consultant or whatever. And then the problem is, is that now you don't have any of those coming through the pipeline. So where do you get the next batch?

Well, you don't need them either because three or four years down the AI is going to be so much better that it can actually do the job of a normal PR consultant. And you know, so this is a long term thing and this is what I said from the very beginning. This was the canary in the coal mine is right. This is.

We can see the writing on the wall.

Yeah.

Have you seen this in your. Like, has it made a difference in.

Maybe not necessarily in your team, but in even with friends of yours and their networks and the stuff that they're doing. Have. Has anybody commented on it yet? Does it seem to be having an impact?

I think people are.

I mean, because I do have a lot of female friends and the ones I speak to know that I've done the podcast and so they'll always be sort of telling me if they've seen some news or have I managed to get in touch with Hannah Fry yet, if she's listening, we'd love to have you on, but it's that kind of. Yeah, it's the young people coming through. What do they want to do?

I've got a friend whose daughter at the moment is saying that she wants to join the police force. And I'm like, what that? I mean, that would never have even crossed my mind.

And I do sort of wonder if people are, you know, younger people are going to start thinking of different roles because, as you say, if we don't take the interns on.

And I, when I worked in publishing, sort of 15, 20 years ago, I had a male intern and he wanted to get into publishing and kind of, you know, whether it was writing or editing or doing the kind of sales side and who's going to train them, what they're going to do. But I think that's a massive question hanging over all of us.

And especially since COVID when young people are working at home and they're not even going into an office, so they're not even getting that exposure to other humans.

So, so many people now we have people that are working, you know, for the council, for the comms team, that, you know, unless you're being told that you have to come in to an office every day, which is really rare now, you're basically sitting at home and you're just looking at a computer. You might as well be talking to an AI bot or something, you're not getting that information from other humans. All you're getting it from is AI.

So I think that's. That's kind of one worry that's kind of off the topic that you asked me.

But I don't really know what these younger people are going to do or, you know, is it looking for other opportunities? Is it, you know, training up Claude, becoming a coder and then like building your own company, doing something.

But what is it that you're going to do? I think it's, you know, what do people want? How do you tap into that? What are they being taught at school? But yeah, but I still. I don't.

I still think that AI is affecting all of us. I think, you know, the beginning is, you know, are women going to be outsizingly impacted?

And yes, you know, in certain areas, then definitely 100% that we are.

Whether that comes down to just the differences in women not wanting to use it and wanting to prove themselves a bit more and kind of doing the sort of fact checking and the attention to detail, well, that's not quite right.

I do have examples where I can think of situations where someone's asked AI and they've just found out and given it to me and it's like, well, that's not right. That isn't right, is it? But they haven't bothered. They just thought, oh, AI is correct. It's given me this. I'm just going to use it.

But how do you know that? Because that's something that probably people have to teach you.

Well, in this. I saw an article and it was talk. They've been talking about it and they've been saying that essentially we're losing.

We as a society are losing the ability to critically think and exactly that and to understand what's right and what's wrong. Because everything we get is from the AI and so we haven't learned anything from anywhere else. Do you know what I mean? So we just take it as. That's.

That's correct.

Yeah. But you know, in schools they're actually saying, okay, you know this question, but then actually go off and research it yourself.

Oh, actually it's completely wrong. That wasn't right. That isn't how you do it. That isn't how you do it. Whereas we're just like, oh my God, we've been given this amazing tool.

It's doing this. And then that's the thing that. I think you shared an article with me yesterday and it was. I can't remember which one it was now because.

Yeah, I don't know. I love it when we're sharing what we've seen and whether it's an email or if it's Instagram.

And I just said before we started, I'm trying to find that link. Can't find it. Yeah. How people are using it and what it's saying and if it's lying to you, but you don't know it's lying to you.

Yeah.

And you're just accepting it. But I was asking, I was asking ChatGPT the other day. I'm. I'm trying to buy some more shares in my house. I've got a shared ownership house.

I want to buy some more. It's like, how much will it cost me to spy? 15% More or to buy another? Sort of 10% to go up to 50 or 70. But it didn't do the maths right.

It gave it to me. I was like, this sounds amazing. Yes, of course I can afford all this money. And then it was like, well, no, you've only done. This is 10%. This is.

You haven't added it together. But you have to have that knowledge already of a subject of what you're asking it to go back. So I think it's really.

You can't just learn direct from AI.

No. And I think. And you know, the key is in the name. It's a language model, not a mathematics model.

Yeah.

So I think a lot of people don't, but we forget that. Right. Because we just get used to interacting with it and we think that it's. But it's not maths. It's not a maths engine. It's a.

It's a language engine. So it's going to give you things that sound, you know, correct, but maybe not.

There's. I can't. Oh, gosh. What's her name? I think it's Pharrell Hegatey. I need to check it out.

But she a comedian and she posts things and she always does this one about LinkedIn and it's just everyone going, oh, I'm thrilled. I'm thrilled, David. Oh, I'm thrill. I just got divorced. And this is what it taught me about project management, you know, all these things.

And it's just as soon as you start looking.

And, you know, I write a lot of LinkedIn posts and I realized the other day, actually, that my presence on LinkedIn since stopping doing women with AI, was kind of bothering to talk about things. I think to myself, I'm on LinkedIn every day, but I'm doing it for the program that I work for. I'm not actually posting things.

You know, it might be what I think, what I think the program should be saying. So I'm doing that. But whenever I write something and I run it by my, you know, my team members and say, I'm posting this, it's okay.

They're all like, why is everyone always thrilled on LinkedIn? Why is everyone delighted? Well, okay, you come up with some other words to use, please. Thrilled. Delighted.

But I bet. I bet I can guess who called you out on that as well. But we won't mention your name. I know exactly who that is,.

But I was. So I totally forgot what the question was now.

That's all right. Yeah. Hello to Gemma. So how do you use AI differently? Because you said you use it every day.

So we'll just go on to this and we do have a couple of things to talk about, actually. Sorry for anyone listening. This is how our brains work normally. And sometimes we edit this out, but we're live today.

The article, yeah, the article that I sent, I said I did send to one was about the bikini picture. That there was this battle between two sort of creators on social Media, which we'll get to that in a minute. The other one was about.

Yeah, it was about AI lying and, and starting to. What we're starting to see is that some AI is, is beginning to go off and do tasks seemingly on its own. And it will, it will go.

And there are some unintended consequences, like it actively lies. It makes stuff up to seemingly forward its own goal, whatever that is.

Or if you ask it to do something, it takes unusual steps that people wouldn't expect. It's deleting people's emails, it's doing all sorts of stuff behind the scenes. And I think we've seen this from the very beginning.

I mean, even when ChatGPT sort of went public in 20, you know, late 22, there was some articles where, you know, they were just giving it simple games to play, like hide and seek and stuff like that. Or one car had to, you know, chase another car kind of around a track.

And what it would do is it would like, move one of the cones out of the way and then go outside the track and then put the cone back so the other vehicle wouldn't see it because it was outside of the track. So it did that on its own. And it worked out that that was a.

The way not to get caught was to actually just go outside of the play, you know, the playing field. So cheating, basically, it's always done that. And this gets, I guess, comes back around to how people are starting to use it.

And I think as some people are really pushing the boundaries of what, you know, the AI can do and is able to do, we're starting to see unexpected behavior because again, we think it's going to act like a human, but it doesn't act like a human. But it kind of is.

What you said at the very beginning, that is exactly it. People lie, people cheat, trust no one. Because that's the thing. If you can. If someone can get around something, they'll get around it.

And that is what it's probably learning. Because also if it's been. If it's only learning on what people are feeding it, you might say, well, this is my challenge. What's the quick way?

What's the easy way? Do I have to like this? I mean, we all know the law is an ass.

All these laws that everyone's abiding by in every single country that have been made up hundreds and hundreds of years ago, and we're still following, going, no, we've got to do this to the letter of the law. Why has anyone actually looked at it and real. Have things changed? Have things moved along? You know, women are now treated differently.

You're treated differently depending on what you believe in. What. Hang on, why doesn't someone go back to the drawing board and kind of work out if it's actually fair for everyone?

You know, equality, equity, looking at all those kind of things. That's the sort of training you have to do in the workplace now. And I think AI is, it's, it's learned from humans, so it is acting like a human.

And that's the bit that's still, I think, why are we, why are we surprised by this? And also when it's trying to, I think, you know, you did that podcast ages ago with one and say, you know, what if I tried to switch you off?

Or when it was thinking, you know, if it's going to take over the world, will it switch itself off? No, because we're almost feeding it that information or that kind of subtext that, well, no, why would I turn myself off?

Why are they asking me to do that?

And it's that kind of, you know, the overthinking, the sort of thinking around the problem, thinking outside the box, but trying to circumvent, you know, a way of doing it or making it easier. Because that, that's why AI has been invented, isn't it, to make our lives easier. So AI is just trying to, to do that, to make its own life easier.

It can quite happily come back and go, oh, yes, David, I did that for you. I've sourced out all that and I've only deleted the emails that have got spam in it. And then actually, oh, no, I did, I did email.

I did delete everything. Sorry or no, I deleted the other emails. But again, it's acting like a human. It's apologizing because it knows that that's how we'll react.

So maybe we should just switch it all off and start again. Or use what we've learned so far. No, I'm not saying burn it to the ground.

Maybe it'll be even worse. You never know again.

And that's why, you know, why women, you know, impacted or affected differently, who's.

Who's building it, who's training it, who's getting involved, which sort of closed Masonic club is like pulling all the strings and not letting the women in, you know, can the women, like, form their own? I mean, there's so many spin off. Sidebar. Yeah,.

Yeah, I, and I, I think, you know, I have, I have huge issues with that whole argument about women aren't building. No one's building it. No, it's just the people building it don't have any impact. It doesn't matter whether you're male or female.

That, that's not the, that's not it. It's not the way it works.

So just because a guy is building it doesn't mean it's going to give results that are skewed positively towards men and negatively towards women, or vice versa. What we're seeing is a reflection of the historical data. That's all we're seeing.

So it again, it only knows stuff if in a sort of rear view mirror kind of way. So we are being confronted with what the data actually shows.

But that good as what it's been of the data that's been fed into it.

Yeah, but that's totally neutral. They just, it goes out and reads all the books. It goes out and reads all the scientific papers published.

And if they call the women that did the scientific discoveries that were written out of history and it's all attributed to the men.

Exactly.

That's not, that's not.

But you can't fake it. You can't make it up. And you can't just go, but you can't make it up. This, this is my point. You can't, you can't give it, you.

Can give it anything you like.

Yeah, but you give it all the stuff that's there. But then what you can't tell it is, you can't then say make up a history that didn't exist.

No, but I'm saying that it's not given it.

That's what, that's what if you say that women have been written out or I'm saying people say women have been written out of history, you know, like who, who.

So, so let's take the CEO. Yeah, yeah, let's take, let's take the CEO thing for images.

Okay?

So you go, you request us, you know, you say, give me some images of CEOs and it gives you 99 men and one woman. Yeah, that's historically accurate.

Yeah, that is historically accurate. But I'm talking about things like, so.

You can't complain about that.

Where women weren't, women weren't allowed to go to university. Women weren't allowed to play golf, for God's sake. Women weren't allowed to do this thing.

So it never meant that they couldn't do it, they just didn't have the opportunity to do it. But some of them did do it but were then ignored.

And what they discovered or what they wrote, whether it was art or music or a scientific discovery, was all attributed to a man. So that. Is that a fact? Is that the truth? Because that. That is what happened. But that doesn't mean that women weren't as actively involved.

But that is the history that I was learning.

But that's why there's all these books and that's why it's being taught in schools now that, you know, some of those facts weren't right and they were fabricated, and then that's where we're sort of back to square one.

But again, I don't think that has anything to do with the engineers that are developing the algorithms, because they just happen to be trained the data.

They've had the opportunity or the advantage or the pushing to get there, or the opportunity to get there. So, no, I think it's, you know,.

So we get in that situation. So we get in that situation. So now when you search for CEO.

Yeah.

What some people are saying is, is that it should give. It should give more women. So it should give a variety of women of different races and men of different races and everything else. Even though.

Even though they have traditionally been white men, let's say in the west, where that goes.

If you just say a CEO, because. Yeah, there are more CEOs named John than there are female CEOs.

Exactly.

So let's change that. If you're asking for a picture, are you trying to inspire young women? Are you trying to portray a woman who is a CEO?

Then you ask for a photo or an image of a female CEO. So that's.

But then you specifically ask for one.

But it depends. What do you want out of it as well? Why are you searching for this?

But this is what I'm getting at, though. So this is the crux of the issue. If you just search for. Show me a picture of a CEO.

Yeah.

What it should show you is something based on the data that it has.

There shouldn't be some secret rule behind it that says if they ask for a picture of a CEO, make sure that it's ethnically diverse and gender diverse. Because it's making it up.

Yeah.

So that's now not truth, it's fiction.

Yeah. But then you ask it to generate the image that you want. You don't just accept the first. You never accept the first offer.

Yeah, but again, so. But this is getting back to you, and this gets back into AI lying to you, because the. The thing that came out of it. Do you remember there was a.

There was a thing where people Asked for pictures of the founding fathers in the U.S. yeah. And it like had them as Indians and, you know, black and all this stuff.

And it's like somebody in their infinite wisdom had programmed the AI to give diverse results when it presented images, but then that meant that those images were totally wrong.

Yeah, but then also.

And so it was making stuff up,.

Because I think that's the way thing it. So that's how we found all these things out. But. But what do we want to do? And if you don't like what you see, then you change it, don't you?

If you're not happy, no way. Something is, you change it.

Yeah, but you need to specify. Again, you need to specify what you want. But I'm getting to the core of it, is what do we want the AI to do?

Do we want the AI to present some fictional future that we think is the correct future, or do we want it to actually give that? We. We want it to represent what has been the fact so far?

I think it's more valuable if it presents the data that it has, because then that enables us to see what's been happening and then to react to it.

Because if it just starts presenting and you just search for, give me a picture of a CEO and it shows you all sorts of men and women and different races and everything, then you start to think, oh, there's loads of women CEOs and there's loads of black CEOs and loads of this, when actually there aren't. It's still all white men named John.

Yeah.

No, so do you know what I mean?

Yeah.

And. And so it's kind of. Some people, like, want to have their cake and eat it too. And it's like, okay, but you want it to do that.

But then you don't want it to tell you and don't want it to lie to you about. No, some other topic.

Yeah. So what I'm saying is it's been federal AIs got that information and then what do you do with it? I think again, it's.

People are so fixated on just saying, well, it's giving me this, it giving me that. But what have you asked it? But also, why have you asked it? What is your end game? What is our end game? What do we want from it?

There was a comment you made last week in the conversation with Lena. You were saying, you know, are you wasting time, not spending a lot of time doing social media?

Should you be spending a lot of time using AI to kind of direct what you're Doing rather than doing it yourself. But then I thought the answer to that is but what is the end game?

And you're saying, you know, will that other person win if they've done this differently to me. But, but one what, because what is it that you want? And that might be different to the guy down the road that's doing something differently?

Because what is both of your end game?

And I think that AI is, you know, it's really held up a magnifying glass to what we do believe or what we have been told or what some people think the truth is or what the truth was and how we're finding out that that isn't maybe the truth.

But I think you're right, it's, it's got all this information and knowledge and it's what we do with it now that counts because I think that was the kind of the, the easy game at the beginning.

Oh, give me a picture of a cleaner, give me a picture of a professor, give me a picture of a primary school teacher or, you know, a chef in a high end restaurant or a cook in a kitchen. And the difference it was giving you. And I think that's what then has spurred people on to go, well, actually I don't want to be viewed like that.

I don't want to do that. I want to do this or let you know, this is my chance. There's a niche here, I can make a difference.

So I think again, it's the fact that it is this tool, it's learning from what we do with it and it's what we do with it that counts. So it's not giving out. It shouldn't be predicting our future or telling us what we want. Yes, it's a language model, a large language model.

All it's doing is predicting. So all it's doing is predicting on what it's got.

So this is where we need to come in and change the narrative if we want to, and we want to change what it's doing. But I don't think AI in any shape or form should be telling us what our future is or what we're going to achieve or what we're going to do.

I think that's still up to us as the humans of this planet to decide. What do you think?

Which is we ask it a question, it shows us an answer. We don't like that answer because that's historically the way it statistically has been.

Then it's our job as a society to change the data that's coming in so that the data is more correct and will correct itself over time. Yeah, but that's unfortunately not what's happening in a lot of circumstances. They're just saying, oh, you can't say that. Say something different.

And they're putting in all these rules. So it's, they're, you know, it's not actually doing what it's supposed to do.

And the thing that, this is where your bit gets correct is then it's who are the people that are putting those rules in? And are they men or are they women? Are they, Are they. Do they have a political agenda?

Yeah. Are they religious? Do they have religious? Yeah. What do they believe in exactly?

Why do they think that what they believe is what everyone else should believe.

Is, Is any better? Yeah, because everybody thinks what they believe is the, the way.

Yeah. And. Yeah, and everybody just sees what, you know, it's a bit like the algorithm as well, isn't it?

You know, I, you know, I see a completely different algorithm to the ones that, you know, that I see next to me in the office is going to say, and whether that's because they're male or female, but we can be looking at the same picture and you see different comments. So again, AI is just, you know, it's fanning our, you know, our ego. It's giving us what. Oh, yes, I am right. This is what I think.

Look, all these other people agree with me. It's just not showing you the other hundreds of thousands of people that don't. But, oh, my God, totally to go off topic, it's like the manosphere.

Have you watched that yet? I decided I didn't want to waste a whole hour and a half of my life in one chunk watching it.

And I've seen all the clips and I've started it, but again, it's like, that is what those people think and then we're all reacting to it.

And Louis Theroux is phenomenal and what he does and how he speaks to these people and it's, you know, it's shone a light on a lot of, on something that a lot of people probably weren't aware of or some people were aware of in the edges. But a lot of it is, you know, you're pulling people in because you may be offering something that they're not getting.

Maybe all these young kids that aren't being giving intern opportunities and actually giving time or opportunities to spend with humans, they're just spending hours and hours scrolling through bots and chatbots and AI and stuff that's made up and Sort of. I don't know. It's also these young brains. It's the social media rule, the one in Australia where they're banning it for under 16s.

Is that the right thing to do? You know, you constantly take it away from people. Is it like a smoking thing? Do you do it, you know, from like a generation.

Generational point and then stop it? You know, the brains that are developing, should they be able to use AI? But again, it's something that you.

I think you have to be learned, have to be taught about it in schools. You have to learn it.

Yeah, yeah. Personally, I haven't seen the manosphere thing. I am. I'm probably an outlier. Luther does my head in. He absolutely just.

I. I can't bear to watch his stuff because he is so, like, wet and just.

But he's letting me. He's letting the subject come through. That's the thing. He's not putting his.

His hot. I don't know, just his whole thing. Every time I see something of his, I'm just like. It just makes me cringe just watching him.

That's not what all his followers think.

I know, I know, but this is.

Why we have this. But this is why we were all.

I'm not judging you for watching him. I'm just saying I haven't watched the manosphere thing.

I was so distraught and disgusted and just appalled and getting angrier by the second.

Yeah, but I think what. But also laughing, I think with the manosphere thing. And again, I haven't seen the. The specific documentary yet, so I've.

I've heard bits and pieces about it from other places, but I think in general about this whole manosphere thing. And I know it's about some of the big influencers and stuff like that, right. And. And it. Their effect on boys and that sort of thing.

But I think what you're seeing is very similar to what we were seeing with women's rights a while back. And you had these really extreme women that were pushing, really. That were pushing a really hardcore agenda and they were outside of what even most.

Most women probably considered normal.

But I don't know.

And I think you're getting a similar thing now because the. The pendulum has swung and I think there's loads of stories and evidence out about that that the pendulum has swung massively towards women.

Women are getting better jobs. They're making more than men now. They're getting better in education. There's more people in.

More women are graduating university, more women are going to university. All these Things. Right. So the pendulum is swung towards the female side now.

So now you're getting a little bit of pushback from the male side and I think.

But it's the pushback to dominate women and to push them down because really we all, we all rise by lifting each other.

Yeah, yeah, but that was. Again. But I think historically that was. We're just seeing a reflection of what happened.

You need to watch it because.

As well.

I'll just agree with you there. We'll agree to disagree. That's a very interesting take on it, though, David. I like it.

I know what some of them say, well, we're going to get in really dangerous territory. Yeah.

My mum and dad, I mean, my nieces to be able watch it. And they're very young.

But I, I also. But there is also a certain, for me, a certain irony to it in that women want to be equal, so they're becoming more equal now.

You're going to have to fight equally with men. So we're not giving you all the advantages. We're not being polite, we're not holding doors. We're not doing all that stuff. We're not protecting you.

We're not doing all the stuff that we used to do because you want to be equal. So now you're equal. Yeah, but so now it's a, it's a. It's an equal fight. And so we're not giving you those protections that women got in the past.

And so maybe that's as well.

Violence against women and girls. Maybe stop doing that as well. Stop using your strength to, to beat someone up.

But again, I would hold a door open for a man or a woman and I think it's. Yeah, I found as well, you know, that, oh, the man should walk on the outside of the street or by the. I think it's the stronger person would do that.

I'm now finding with, you know, with my mum, I want to walk on the outside of her. I want to be the one that's protecting her. And I think that's just, you know, you do it for the person that you love.

You do it for the person you care for. You do it. And it's that kind of. It's. It's that equality and equity again.

It's, you know, what's, what's the right thing to do and how are you helping someone or not helping them? And do.

We want, you know, everyone gets a good feeling, hopefully from helping someone, but obviously there are those, the bad humans, the bad actors, the people that get a kick out of not helping someone and not hurt. Not, you know, hurting someone instead. So, I don't know, we've gone completely. I was just talking about the life in general.

Yeah, that's fine.

But yeah, we'll go.

Whatever.

What was the other one you wanted to say? It was you'd sent me.

Oh, so, right, let's.

Yeah, let's bikini.

But again, let's bring it back around.

Could be a man and a pair of budgie smugglers, I think. Again, don't focus.

Okay, but that's not the story. It could be, but that's not the story. The story is, is that there was a. A guy who was just on the Internet, he was running a campaign or something.

He asked. He went to Nano Banana and he asked for an image of a woman in a bikini on a beach.

And it gave him an image and he put it in his marketing material and off he went. He, like, didn't even think about it. He didn't ask for anyone specific. He just said, give me a woman in a bikini. It generated this image.

Well, it turned out that it was an image of another female content creator influencer. And so she sent him a nasty note saying, hey, you can't use my image. And he goes, look, I didn't use your image on purpose.

I just asked Google for it. And she's like, well, I don't. Google wouldn't come out with my image, blah, blah, blah. And then he sent the query.

This was all in public as well, I think on social media. So they were like commenting back and forth. And he said, look, here's the exact query that I used. You can test it and see what it gives you.

And what they discovered is, is that that exact query gives a picture of her every single time.

Wow. Are they all using or the same.

Yeah, Nana Banana.

Oh, yeah.

Which is Google. That's Google's image thing and image and video. And so that was. The really interesting point, is that we. We also think that AI is random.

Yeah, but it's normal.

And normally it is. No, but normally it is. And it's. It actually has.

Part of the algorithm is it has a randomness factor, because in natural speech, if randomness was set to zero and you asked AI a question, you would always get exactly the same answer with exactly the same words in exactly the same order. But that's not how people are. And so what they worked out in the beginning is that it didn't feel very human if it wasn't a little bit random.

So they. There's a randomness like coefficient that they put in. So that if you ask AI the same question, you'll get a slightly different answer every single.

The words will be different, but the theory will be the same. But the words will be different.

Yeah.

And so what you would expect is that if you ask for. If you put that same query and said, give me a picture of a woman in a yellow bikini on a beach, you would get a different woman every time.

But what's happening is you're getting exactly the same woman every time.

Yeah.

And so he was kind of able to prove his point. She's still like, okay, well what's going on here? So what they were able to discover is that Google has been training its AI. It's Nano Banana and.

And Gemini and all the other ones on all of your data from YouTube and all your emails and all that sort of stuff, which we've kind of known for ages. But it did kind of come out.

Yeah, it's like you never read.

It's in the terms of service.

You know, you get your new iPhone or you scroll, scroll to the bottom, click, accept. You're never going to read it. We all just accept it.

Did you see the south park about that?

No.

They basically, there's a whole south park where I don't remember who it was.

It was like Kyle or something, or maybe it was Cartman and was sitting there and basically an ambulance pulls up outside and they're like, we're here to take your kidney. And he's like, what do you mean? And he goes, well, you signed in the Apple terms and conditions.

You signed that you were like, we could come and take your kidney, like at any time. And he's like, what? I didn't agree to that. And you're like, no, you agreed to the terms and conditions.

And they're like, take him to the hospital and do like a kidney transplant. Take his kidney. Yeah, exactly.

But, yeah, the whole point of it was nobody reads, you know, what's actually in the terms and conditions and anything could be in there. But.

But that. Can we change.

It could be. It could be a. It could be a man in Budgie Smugglers. Right. But it wasn't. It was this influencer.

And I think the thing is there were so many images of her and so, so much traffic.

She's the ultimate image.

No, the algorithm did. Yeah, the algorithm did. It's not someone searching for it probably, it's the data.

Yeah, No, I want her in a. Actually a bigger bikini, a smaller bikini. I want her to be a different Proportion taller, smaller, whatever. Everyone's just accepted it.

So it's gone, oh, great, I gave it the right answer. I'll just keep giving this to everybody.

Which is how the algorithm works. So it's like, is that good or is that bad? And if you give it a thumbs up, it goes, it learns from that. And goes, yeah, okay, that's cool.

So somebody saw that, give me a picture of a woman in a. Well, I'm saying yellow, I have no idea what color, but, you know, whatever. And they go, yeah, okay, that's good.

Or they copy it, or they ask to download it. And then the system goes, okay, they like that. That's a good response. And then it trained it.

But I think again, because there were so many images of her and so many likes and so many reshares and whatever that it learned from that.

Yeah.

And so now it's using that single image. So.

But what's again, problem? What's the question? And did someone say, did someone check?

Because I think that's the other thing when you, you know, like to get around also, again, as a student or something, to get around plagiarism, you ask Chat to be teacher or whatever to write something. But then you're supposed to go in and go, can you check that this isn't plagiarism? Can you check this hasn't been used anywhere else. So did here.

Or they. Whoever was asking for this image. But why would they even think that? Because they're thinking, I've asked AI for this picture.

Did they need to go in and say, is this a picture of a real woman? Or can you show me where you got this from? And then go, well, actually, can you make it so it's unidentifiable and it's not a real person?

But then that's the same as, you know, you have to get photo permissions. When you go to an event, if you're taking pictures at a conference, you know, you ask people, can I, you know, if we're taking pictures, do you mind?

You know, I went to one recently and you had to wear a different colored lanyard if you didn't want your photo on there. And again, I think it's just coming back to that common sense. What are we using it for? Why are we using it? What does it mean?

If you take someone else's piece of art and then you just sort of scribble on top of it or put a blob of different colored paint, does that become your painting? This is one for Lena to answer, you Know, where does ownership stop and begin?

Yeah.

Yeah. What's the question? The answer's mostly no.

It was just, we were just talking about it and this is, this is what we were talking about.

Yeah, I love it.

And, and I think. So getting back to that, I think taking a lot of this into consideration, this gets back to.

And again, I feel like I'm going to talk about this in every single conversation. So I apologize to everybody ahead of time. I'll try and say it in a different way every time, but I'm not sure how well I'll do.

But we prefer it when you say it the same.

But it gets, it gets back to how we use AI, right? So I don't use AI anymore to create just a brand new image. I always give it something.

So, for example, I have my Life by Misadventure podcast and what I've decided to do with the. The YouTube thumbnails is I take a photo of a person that they've given me which their profile or whatever picture they want.

I take it and then I have it make it into a cartoon style. So I'm not saying create a new image for me. I'm saying here's a picture, make this look like a cartoon.

I like your profile pic, by the way, on this podcast.

Thank you. Or I'm giving it something and saying, do something with this. Right? So I'm giving it information and then asking the AI to help me with it.

So I'll write an article and then I'll put it in AI and say, hey, can you help me improve this? Have I forgot something? Or can you make this, you know, how would I improve this? Whatever.

So I've stopped asking it for original content, if you want to think about it that way. And I give it something and say, do something with this.

Yeah.

Now. Yes.

Yeah. That goes back to the question you asked me. Sorry. Which I completely failed to answer, which is how do I use. How do I use it? And I use it how you do.

Like you're saying. And I think that's. But I think that's because we're set, you know, well, relatively sensible people. But I do.

I will put in what I want to say and then I might say, how would you polish this? How would you make it sound more professional? I mean. But again, I don't always use what it. Exactly what it. I don't always use what it.

The answer it gives me back. And sometimes it gives you back four. And I think, well, I'm choosing and I'm actually amalgamating. Three of those. Still my words.

But that's what you want.

Yeah.

And that. So it's help. It's an editor at that point. It's an editor. And it's what Lena talked about.

I think it was with Tom Morley, when she talked to Tom Morley, because, you know, he's an artist and he uses AI quite a lot to generate hundreds of different variations of an idea that he has. And then he curates that.

That list down to the ones that he wants, and then he can go and physically paint those things, but he uses AI for the ideation process and it becomes curation. And. And that's kind of what you're talking about is. And that's what I do.

So again, I'm asking it, or if I ask it for something original, I might say I'm trying to plan a social media campaign. I need a list of topics and I want to do X number of posts over X amount of time. So I might say I want to. I want 52 posts over a year.

So this is a weekly post. I want to generate a list of topics. Here's the subject that I want to generate them on. Here's my top 20 that I know off the top of my head that.

That need to be included in the list. What else can you think of? Yeah, right. And then it will go and it will say, oh, that's a great idea. I love it. It always tells me it's a great idea.

And I know it shouldn't do that, but it does, and I love it. And then it says, oh, that's a great idea. These are some good ones. Here's a nice order with some extra ones that I've done.

And then it creates sort of a. A logical flow to it. It adds some extra stuff into it. And then I look at that list and I go, yeah, I don't like that one. I don't like that one.

But that's a good idea. I see where it's going with that. And then. But you do have to interact with it and you have to, you know, kind of.

It's there to help you, not to replace you. I think if. Yeah, no, yeah, if we can remember that. I think that's where we're at. So it is still very much a tool.

Yeah, I'm making it mutual. So I go back, I don't know if you do as well. And I might go, this is the final email I decided on. Or this is the final post.

This is the final wording for this quote. This is the final bit. Because otherwise it's never going to learn what I actually want or what. What I actually use.

Exactly. Yes. Yes, exactly. I do that as well. So I. I take all of that and then I get to the end and I go, here's what I used.

Yeah, yeah.

And then it usually comes back and goes, that's really good. If you'd like to tighten it up. And I'm like, off.

I'm like, no.

Yeah. I'm like, no, I'm keeping it this way.

When it to. Comes. Goes.

Okay. But. But that's how we train it for our own voice. And, you know, so. Yeah, so I think we agree on that. And I think, you know, that's this.

And I'll keep, you know, talking about that because I think it's important. Right. I'm conscious. I know. We had an hour. We started a little bit late, so probably. Any final thoughts before we go?

Well, just. I've really enjoyed this. I do. I. I never know what we're going to talk about.

We sometimes have a little disagreement, but we generally always come back at the end. Yeah, well, not a disagreement. There's different viewpoints, but that's what makes it interesting, and that's what you need.

That's why the conversation needs to keep happening, and that's why people need to keep listening and need to tell us what they think. And because that's the only way, we're only going to learn if we do it collectively. Because you can't have one person, one entity, one anything.

You know, it's got to be all the voices.

Exactly. I will. I will endeavor to watch the manosphere thing.

Oh, God. That means I'll have to watch it.

As well and suffer through it. And then I know my wife's wanted to watch it, so we can just.

Well, and your son, get him on it as well. See what he thinks.

I'll. We'll. We'll watch it together. And then. And then probably what I'll do is I'll. This isn't the right show to talk about that, but I will go on my Life.

Life by Misadventure, which is my other podcast. I'll probably cover it on there.

Cool.

And I will probably, for that one, I might try. I don't know how I'll approach it, but I might try and find a. Like a.

Maybe a clinical psychologist or a psychologist or someone to get on actually to talk about it, to have an expert opinion. Or I'll just get a couple of my male friends and my son on and we'll just go on and have a massive rant and just talk about.

It, find some of the comedy clips. So the clips. I'll send you some after this.

Well, I have a couple of friends who I think would. Would probably give some thoughtful, you know, we could have a thoughtful discussion about it on one hand and it would also be funny on the other.

So I might aim for that because it could be a really heavy, you know, kind of subject. And so it feels like it needs some inappropriate humor to go with it.

It does, because part of it is. It's really interesting. And it's. It is that whole bonding between father and mother, or mother. Sorry, mother and son.

But I think it goes back as well to what we were talking about earlier. You know, is the pendulum swinging the other way? There has been a lot of coverage. Like our boys now, they don't know how to express their feelings.

They don't know what to do. They're being made to feel differently. They're being, you know, they're being pushed out. It's like the whole Barbie thing.

Barbie and Ken, you know, Ken suddenly in a world where, you know, he's grown up, where Barbie is in charge, and then suddenly he comes into the real world and, oh, no, men are in charge. Or the patriarchy, or is it just horses? But. And it's like, but what? But again, what's fair? No one should have the upper hand.

Everyone should be playing to their strengths, their own strengths. And I'm not just talking about physical body strength. You kind of, you know, that go.

And then that opens up, you know, the competing, you know, you can, you know, you should science. What backset. You can only compete in sports if you're with people that are physically the same as you. So there's.

There's all these other questions and things that.

Oh, that's risky. But also it's a risky stance to take.

Wow.

Men and women are equal.

We're going into teenage brains because. Let's talk about that for a minute. Because that's why they're looking at social media and AI and the effects of effects that it has on that.

Because I've got, you know, friends that are going through that. Their daughters are early teens. Why are they being so horrible? What is the difference? Why aren't the boys reacting like that?

And that's because it all goes back to how we're programmed biologically to cut the other. Cut those apron strings. So girl. Teenage girls are going to act completely differently.

Teenage boys are full of testosterone and really angry that's why they're all sent to war and why they're so efficient on the front line. And girls are, you know, being set up to kind of, you know, to, to start new families and go off and do their own things.

Yeah, get a clinical psychologist, get any kind of psychologist. Let's get people that know about brains.

I think that we'll finish, maybe we'll finish on this because again we're starting to go down a whole different conversation here right at the end. It is the humans fit. It is humans fit and we use AI so we're technically humans with AI.

I think the thing that we've missed and that society misses a lot in, in the, in this sort of last 30, 40 years is that men and women are different. And that's okay because we're supposed to be different. Yes, because we do different things and we bring different things to the table.

And this isn't going to be a popular opinion either and I don't care.

People can twist is that family should be a man and a woman, two parents who raise a kid or kids because the man brings the, the male figure brings one side of the personality and the female brings the other side.

Yeah.

And you need both of those to, to form a well rounded person and any, anything outside of that there's loads of single dad families now as well as single mom families. But you in those to have a balanced answer you miss yet you, you haven't got the balance right.

And that's what we're seeing in society is we're seeing the side effects of single parent families. And I think that's a huge contributor to where we are at the minute.

And so you know, but, but no one wants to admit that you need both people because then that's stigmatizes people who, you know, parents who are single.

Again, that's to their facts. That's the science and it's, but then.

It's implying that, that maybe someone is better than the other and it's not. We're just different and we bring different things to the table and like this whole thing of boys can't express their emotions. No, that's not true.

Boys can express their emotions. Boys don't express their emotions like girls.

And we're being told that we don't express our emotions by girls because girls don't understand how we express our emotions.

And in the same way we don't.

Like being forced to express our emotions like girls.

Yeah, but in the same way boys don't understand the way that girls are expressing their emotions.

Exactly.

They react. You're going, well, that they're saying, well, that isn't right. But it's not. It's just not how you.

Exactly. Yeah.

Anyway, I love that. Just to bring it back to AI, that's exactly what my goddess says. And that was how I read. That was how I originally got into AI.

I mean, apart from you. It was scary smart and it's about your masculine brain, your feminine brain and bringing those together.

So, yeah, brilliant. Thanks, chair.

Thank you, David.

Always a pleasure. And let's stop this live broadcast. I have no idea what this is going to do or if anybody even saw it. What's going to happen to any of the files.

Because you would think that running a studio, I would be people asking live.

Comments and they're not like a teams chat.

Well, there is and I've had the chat open. I just don't know if anyone's out there actually watching it. So maybe next time when I do one, I'll.

I'll get my son to monitor it and kind of see what it looks like and. And just put in a comment as a test. Or maybe I'll just do a live broadcast at some point. But anyway, thank you. Have a good one.

I know you need to go and you've got cost to do and stuff, so have a good day and I will see you soon.

Thanks. Take care.

Bye. Bye, everybody.