What is AI doing to my kid’s planet?

Hello, and welcome to Raising Generation AI with me, mum of two, Holly Green.
And me, AI expert, Fiona Morgan.

This podcast is all about the complicated world of parenting and AI, because I’m a mum and I’m a bit worried about it all.

And this week, we’re asking the question, what is AI doing to our kids’ planet? Later on, we’ll be talking to Dr. Kelly Widdicks to find out how our use of AI impacts the environment.

So I mentioned to you the other day that a friend of mine had created the most amazing party invites on AI. And I decided I was going to go down this route and create my own, because I’m planning a unicorn party. I’ve mentioned that before as well.

And I tried to make these invites on AI and they are unbelievably bad. Right, have you got your phone down? I’m going to send you a picture now and you can tell me what on earth I’ve done so wrong here.

I’m just going to try and describe it to the listeners. So I asked for a unicorn party invite with some rainbows and some nice pretty things to make it look lovely. A girl riding the unicorn, some details on it.

And what I’ve ended up with is, well, the unicorn doesn’t have a horn, but that’s probably the least of its problems. It looks like it’s been drawn with the most basic shapes from Microsoft Paint. I mean, isn’t it bad?

I mean, it’s just made up of like shapes. As I said, the unicorn has got a horn. The girl doesn’t look like a person.

Like, what did I do?

I have no idea what I did to you.

I feel like I was quite specific in my instructions. I was like, please gonna have an invite with this information on it. I would like a picture of a unicorn, a girl riding it with brown hair.
I’m going to have some rainbows and clouds and stuff in the background, make it look really nice. That is what I got.

There’s this thing called prompt engineering, which is actually how you write a prompt. And maybe you are just not a prompt engineer, Holly.

But what could I have possibly written to come out with that? I mean, I just went on to Claude, just put all these instructions in and just… Did you do it in Claude?

Yes.

Interesting. I thought Claude couldn’t make pictures.

It can’t. Have you seen this? Have you seen this unicorn?

Do you know, I’ve never used Claude to make pictures.

Oh, so is this what I’ve done? Have I just used the wrong program? Should I have used one of the others?

I had an invite sent to me for my nephew this week. And I’m afraid to say it was made by AI, Holly, and it was phenomenal.

See? That’s what I mean. Everyone else.

My friend created phenomenal ones. And I was like, why have I got this so wrong? I just feel like the most rubbish tech person some of the time.

So tell me about your invite, sorry.

Well, it was a Pokemon one. So my sister-in-law did it. And I think she did say that it took a little while, and she had to rewrite the prompt and say, do this, change this and work on it.
It wasn’t one type and done. But she also was very quick to say, I didn’t put the name in, I didn’t put the address in.

I didn’t, she was really like, I don’t want to be told off. To be clear, Fiona, neither did I.

There we go, good, nothing, all the detail. To be fair, it is a relatively low risk that, because if you’re just typing something in once and whatever, it’s not like you’re putting it somewhere on the internet for everyone to read. But still, this brings us actually on to our next thing that we wanted to talk about, because you mentioned this in the School Report episode, you didn’t really feel like you were learning how to use AI and what to do with AI particularly well.

Yes, I do feel that is true.

You might have demonstrated that. So for the listeners, we got together and came up some ideas of how we might address that, and maybe what some of you guys listening might enjoy as well. We’ve come up with this idea of each week, and the weekend, I’m going to post, and you may have seen this, we started it last weekend.

Each weekend, I’m going to post an activity, a little activity that you can do related to parents or kids or whatever, in one of the AI tools for free. It’ll always be for free, it’ll always be privacy safe, it’ll always give you the instructions on how you can do it. Just so if you are a bit like Holly and maybe not the best at AI, you might learn a little thing and get some good ideas.

The first one we did was creating a colouring page. You can buy a colouring book, so maybe you don’t need to do this, but if your kid really likes skateboarding and unicorns, then you can create a colouring page with skateboarding unicorns, which I think is really good fun, right? And so I set you the challenge then, Holly, of having a little go at doing this for this episode.

How has it gone?

So this went much better than the invite, I have to say. But you were nice and helpful because you gave me the prompt to work with, so I did have the prompt, and then just put my own information in. And I have to say, it was actually really good.

So the first time I did it, the lines were a bit thin, and also there was some slightly dodgy bits in, so it had drawn a butterfly, but it had turned the antennae into a bit more of the body, where it should have been the space between the antennae, it put some butterfly patterns. So there was some slightly dodgy things going on, which I have to say, I didn’t actually notice, but when I printed it out, my little girl was like, mom, what is this? So I think a little bit more care and attention from me would have been good there.

And I noticed as well, you could change the ages, and it would make it much simpler and more complicated, which was really good. I did have problems printing out, but again, I think that was probably just me being technologically a bit rubbish. But we got it printed, we got it out, and yeah, they really enjoyed it.

Weirdly, when I asked it to do, my son’s running into Captain America, I asked it to do Captain America and it wouldn’t do it. And I did wonder if that was a copyright issue.

When I was doing some silly pictures of you, do you remember your toothpaste story? I was trying to create an AI generated image of you to accompany that story on social media and it wouldn’t let me do it at one point. It was like, this is a public person.

You can’t put this. I’m not generating it yet. And I was like, she was pregnant.

It was me. Because I used one of your screenshots from the Webber as extra pictures. I want to give it some.

That’s good to know that it does that, doesn’t it? And it was like, nope, nope, nope.

But yeah, it is funny about copyright, which is probably a good thing, right?

But I just love the fact that you can just create whatever it is that your kids like.

And it can be really good. I think that’s the best thing about it, isn’t it? You can tailor it and personalize it to things that your kids love.

One thing I did think though, and this obviously links really well to this week’s guest, is I did think how does generating that colouring page or colouring pages compare to going and buying a colouring book. Because if you’re making a specific journey and driving into town and that’s using energy, burning up fossil fuels, then that probably is worse. But if you’re just picking up on your weekly shop in Tesco, that’s probably better.

But I don’t really know. I find it really hard to judge these things and compare them. I feel like I just don’t know enough about it.

So I’m really pleased that we’re chatting to this guest today.

And I think the other thing to think about as well is the whole on-demand stuff. So if you went and bought a colouring book, that’s going to have, what, 50 pages or something. Is your daughter or son going to colour in all 50 of those pages?

Or is some of that going to be wasted ink and paper and stuff like that? Whereas actually, if you just generate one they love on the moment they want it, is that better?

Yeah, that’s true, actually. Yeah, that’s a really good point. Most parents will know that a lot of young kids just like to go through and scribble on every page, which is really, really irritating.

It’s like ruin every page of the book and then go, but you’re right, it would stop that. And yeah, finding out more about doing those sorts of things and whether it is a good idea environmentally, I would be really keen to do. So I’m going to be quizzing our guest about that today.

And yeah, just remind us who we’re chatting to today.

Yeah, we’re chatting to Dr. Kelly Widdicks. I was really keen for this to be not a campaigner. Not that Kelly isn’t for the environment, as you will see when you talk to her, but what I wanted to make sure was that we were talking to someone who could actually give us data driven examples that are really factual and based in research rather than anything sensationalist, because what you’ll find when we talk to Kelly is that a lot of this stuff, there isn’t that much research.

We will ask her the questions, but we have to be prepared for her to say, actually, this is a new area, there’s some research going on at the moment, or we’re trying to figure this out, or we don’t know how this is measured. But I still thought it was more important that we had someone coming at it from that evidence research basis rather than a campaigner, because I do think there are two sides to everything, right? Like you were just explaining with the colouring book.

So yeah, after this, we’ll have a really great chat to Dr Kelly Widdicks.

Holly, this is an episode where I think you probably have more opinions than me. As a meteorologist, you deal with the reality of what’s happening to our climate every single day. Whereas I’ve just read some books that explained how data centers that power AI have impacted the environment.

And even then, if I’m honest, the focus has mostly been the impact on the people, less so the physical environment. And as we’ve discussed before, I’m not really sure our listeners, the parents, the people using AI have a true awareness of their use of AI and what that’s doing to our environment. So I was keen to find a guest who came at this from a scientist perspective, not a campaigner.

And that person is Dr. Kelly Widdicks. She’s a researcher for the UK Centre for Ecology & Hydrology, where she leads on digital innovation. And she spent her whole career measuring the actual environmental impact of digital technology, which includes AI.

And she co-authored one of the most cited papers on the real impact of ICT. And it’s researchers like Kelly who are grounded in data that make sure that when tech companies tell us what they’re doing for the planet, we actually have the means to check that it’s true. Dr. Kelly Widdicks, welcome to Raising Generation AI.

Thank you for having me. It’s lovely to meet you both. And yeah, really excited to talk about this topic.

Can I just start by asking, so for me and our listeners, just headline the key environmental impacts AI is having.

It’s quite hard to know specifically about what AI impacts are, particularly because it uses underlying digital infrastructure, which services all lots of types of different technologies and applications that we use in everyday life. But every time we do use a digital technology, such as an app or a service or AI, it does depend on this infrastructure, such as the devices that we’re using right now to talk to each other, the networks and data centers that have environmental impacts in their creation, their use, their maintenance and their disposal. If you take data centers specifically, for example, these require energy to run, they require water for electricity generation and for cooling, they require raw materials in mining for developing those technologies, and also things like land use for the data center location.

I think because of the fact that quite often these technologies are away from users, we see the devices, but we can’t see the networks and data centers. We don’t really think of them as real physical infrastructure that has that environmental impact and is drawing on these different resources. We have terms like the cloud, which makes it sound like it’s non-existent in some ways.

Something fluffy and nice.

Fluffy and nice. In terms of specific stats for AI, there are quite a lot of different calculations and estimates that are out there in the news and the media, so it’s quite confusing to navigate. Most recently, there was a study in Patons by someone called De Vries, who estimated that from company reports based on data centers, what they actually provide in terms of data publicly, that AI impacts may be equivalent to, in 2025, to the carbon footprint of New York City.

And also the water footprint might be equivalent to global consumption of bottled water. So they’re not insignificant when you think of that bigger picture. But actually, there’s quite a lot of different methods that are out there to do these calculations, quite a lack of data underpinning them.

It’s quite hard to then scrutinize what the actual impacts are. But hopefully, that gives us some sort of indication of how they have impacts.

I’ve seen a lot on social media recently of pictures of the Rio Grande dried out and claims that that’s because of data centers and things like that. So the water consumption in the data centers is potentially, if this research is correct, the same as the bottled water consumption across the whole planet.

That’s what this study estimates, but it is just very difficult to quantify. There is different methods in doing those types of quantifications, and actually the data is not very transparent or measured at the moment.

Are some of the companies not transparent enough about what they’re doing, do you think?

I think so, yes. I would say yes. There are some companies that have started to produce more of this information, and there are things like in the UK, for example, there’s a big initiative from these AI Growth Zones, which is new data centers to support AI specifically, and they have set out certain requirements of what, power, water availability, land availability that they need.

So it’s really making sure that they have calculated properly and they’re doing that in a sustainable way. So yeah, we do need a lot more transparency in this space, I would say.

Again, this is a lot of what I’m seeing on social media. I’m seeing that there’s these sort of specific areas where these data centers are being built. They’re really big.

They’re really impacting the people that live around them. They’re contaminating their water. They’re limiting their ability to access enough energy because the data center itself is using so much energy.

So it’s really the people that are living close to these data centers that are seeing these huge impacts. But again, I’m getting this off social media. To what extent is that the case, do you think?

And are we going to see huge growths in these data centers and them popping up in so many more different areas?

So in the UK, for example, currently quite a lot of the data centers are in the south and particularly in Slough, that’s seen as a key data center hub. And historically, they might have been placed close to big populations so that they can reduce network latency when supporting technologies such as video conferencing, streaming. But with things like AI, the activities in the compute doesn’t necessarily need to be as close to people for speed transmission.

So these can be done at other locations. And so there are kind of initiatives, as I said, that mention the AI growth zones about depending on where that availability of energy or water, etc. might be, that they could be placed in other areas.

But yes, you’re absolutely right. There’s been studies and reports of things happening in Ireland or the US, for example, where they’ve took so much water from that local community that we’ve not been able to then support other infrastructures such as developing new housing. So there is some real impacts on other people.

You know, quite often these data centers are targeted at rural areas where there’s lots of access to land. And this then may be my kind of more marginalised communities.

And there’s actually been quite a bit of pushback from some of, I think, one of the Scottish AI growth zones, for example, because they’re not quite seeing that there’ll be as many benefits to the people in terms of jobs that are maybe suggested.

There will be some jobs, like around construction and of the data center itself and potentially some jobs coming from those AI innovations at those new data centers.

But I don’t think they’re seeing the numbers of jobs and benefits locally that then outweigh the kind of negative impacts on the landscape, such as destruction of trees and farmland. Yeah, so there is a lot to do with thinking about the location of these too.

But I guess I would say, there is some positive, if we do start thinking a bit more systemically, sorry, taking more kind of systems approach, seeing data centers as one part of a broader system of our infrastructure, then we can start to see potential benefits. So for example, there is an issue with data centers requiring energy and water for cooling. If we, because of the fact that they have this excess heat, if we then instead use that excess heat for more positive points elsewhere, for example, for heating local housing, then we’ve got suddenly a challenge in one sector becomes kind of an opportunity in another sector.

So it’s this kind of more joined up collaboration that we need, and this is being done by some councils, Blackpool Council have actually got the Silicon Sands project around sustainable data centers and they’re thinking a lot more about this kind of joined up approach to infrastructure. So it is possible for there to be positives as well.

Do you think that, as it stands currently, most of the responsibility for thinking about the environmental impact of AI is on the AI companies? And do you think there should be more done by governments to ensure that they’re behaving in environmentally conscious ways? And are some of the AI companies taking these responsibilities more seriously than others?

Is there a distinct difference there? Or is it hard to tell?

I think I would say that it’s definitely… There is agency in technology companies. This is a really big point where technology companies need to start considering sustainability.

And there are instances where they are doing this. They are a big investor in renewable energies. There has been news articles recently around them interested in things like the small modular reactors for data centers as well.

So really trying to invest in those renewable energies. But at the same time, there is reports around them investing in gas generators for the data centers as well. And we do have to think about that kind of broader picture again about our use of renewable technologies.

If we use renewable energy just for data centers, then that reduces the availability to use renewables for other sectors such as our housing. And renewables aren’t, they are still limited as well. There’s limitations in terms of the amount of silver that we need for solar panels, that we don’t currently have enough renewables for a stable power supply.

So whilst there is some investment from technology companies in renewables specifically, which is a really good way in which they’re thinking about sustainability, it’s not, you know, we do need to consider this kind of continuous increase in demand at the same time and how we design those technologies as well. So I do think there is a lot more that they can do to support sustainability. Firstly, by thinking about what that technology is actually doing, you know, if it’s positive for the environmental society.

Secondly, by designing these technologies that are built to last. So, you know, obviously because of the impacts at the creation and disposal stage, we can design for longevity and make sure that we’re reusing, repurposing the kind of more physical aspects of those devices and infrastructures. And that does require some consideration of how the software impacts on that as well.

I don’t know how many, if you’ve ever had it where you need to get a new device because your operating system no longer works on that device.

Yeah, it’s so annoying every couple of years.

So it’s things like that. We need to be thinking about the software, making sure that that is supporting the hardware as well at the same time. We can make sure that there’s alignment with other infrastructures.

And that is where governance will probably have more of a role to play because it’s beyond the technology sectors alone in terms of what they can actually do for other sectors and aligning that with infrastructures. And then the kind of fourth point I would say is that they really need to start thinking about what we call rebound effects in the design of technology. So quite often in the design of different AI models or digital technologies more generally, there has been an emphasis on efficiencies.

So if you look historically, they know the efficiencies of technology have just completely increased all the time, but at the same time, the impacts of technology and the carbon emissions have also increased. What is happening is actually this thing called a rebound effect, where we make something more efficient and then we end up doing more because we can. That then drives new innovations and then that then creates more impacts.

We have to expand the infrastructure to then support those new innovations. So we really need to start thinking about these rebound effects within the design processes when we are creating a new technology and what that impacts and the ripple effects that that might create. I guess quite often digital technology and AI sometimes is seen as like, oh, this will like save us, you know, this is save the planet.

And some of the examples that I’ve seen of that have been around video technologies in terms of video conferencing and how this impacts flights. And this is a complete myth. There’s a kind of assumption that basically because we’re meeting online today, that we’ve maybe saved a flight to Thailand or whatever, if we were going to meet Fiona in person.

I like how you’re coming here. I’m not coming to the UK. Oh, you’re not coming to the UK.
We’ve got to. Definitely going to Thailand. But yeah, if you look at that data, video conferencing technologies and video streaming data has increased, so have flights.

There is not that clear impact on other things. Actually, because we’re meeting today, Fiona and Holly, maybe one day, Fiona, we might meet in person because we’ve been able to connect through video. So there is these efficiencies that are impacting on all other sectors as well.

That really surprises me. I had assumed all this video conferencing had reduced carbon emissions, whether that be from driving or flying or whatever else. Are there any positive environmental stories to be had then, or is it all kind of negative?

I think this really depends on what it’s being used for. So I think going back to what I mentioned around making sure that what you’re using AI for is having like a positive for the environment or societal. And personally, I don’t really use a lot of AI.

It just kind of is more embedded into the technologies that I use. But in my work life, I see real opportunities for it. So I work for the UK Centre for Ecology & Hydrology or UKCH for short.

And this is a research organization focused on environmental science and we’re using AI to support environmental science in quite a number of ways. And just to give a couple of examples of that, been developing of a large language model. So you kind of ChatGPT style AIs that allow people, like scientists and decision makers to discover data and environmental data and methods a lot easier.

Because actually that’s quite fragmented across the scientific landscape. So if we can get evidence from trusted sources to people a lot quicker for them to make decisions for nature, then that’s a really great use of AI. We’re also using AI for biodiversity monitoring.

So with automatic sensors, we’re able to then scale up the amount of monitoring that we can do about the environment and biodiversity. You know, they have so many more observations now and data, and that really allows us to scale up where humans just can’t do that same amount, which enables to find out what’s going on in the environment.

Is there opportunities or solutions that we can enact and spot these changes easier

So there are some real positive uses as well, but I think it’s important to remember that just because we can use AI for a good purpose doesn’t necessarily mean it will be. So we do need to design that kind of boundaries, I guess, and the governance around it. So having conversations like this is great.

Now you see these trends going around as well, don’t you? So do you remember the one where everyone was creating like dolls of themselves, the Barbie dolls and the action men. And you know, it looked great fun and they look really effective.

But I did at the time think so many people doing it, that’s got to be having a huge environmental impact. Do you think doing things like that is perhaps a bit misguided?

I mean, who am I to say?

I know, it’s one of those questions, isn’t it?

That these dolls don’t add so much value to people’s lives. But I think this is where it’s really important to think about as an individual. Obviously, our impacts as an individual level aren’t that important.

But as you said, when we start to add them up, and we start to then think about what kind of impacts things like mass amounts of dolls, AI-generated create, is that really, are we doing this in a way that’s positive for the environment and society? Is that adding so much to society or could we potentially live without those types of things? There’s other areas that are quite harmful uses of AI, things like deep fakes that we could completely avoid, if we cared a little bit more about each other, if we cared more for the environment.”

So I guess at an individual level, I would think really, is this going to add value or purpose to society and environment? And then only using technologies within that kind of thing that’s for good in a more sufficient way. So only using it when it’s really, really necessary, avoid waste and ensure that that AI use enhances yourself and your life meaningfully.

And is there any way to get a sense of how much of an impact something you’re having is doing? So for example, if I was to make a birthday invitation for 15 kids, let’s say, and I created it on AI and I texted it to various mums, is that comparable to me driving to the shops and going and picking up a packet of invitations and writing them out and sending them by hand? Is there any way to get a sense of which of those is better for the planet or not really?

So you could do it, but it is really complex because you’d have to think really carefully about the types of AI model and what training that that’s had. When you’ve done that activity, and as you said, who you’ve sent it to, thinking about all those impacts on the network as well. Then you’d also at the same time have to think about how you’ve gone and bought the cards, you mentioned driving, the supply chain of the paper, how the creation of that paper as well.

So there’s lots of different factors that you’d have to measure. You’d have to make sure that the measurements are correct. I do think there’s an element of us becoming, we do need to become more carbon literate.

I don’t know if you’ve heard of the book, How Bad Are Bananas? But one of my collaborators, Professor Mike Berners-Lee, he developed this book to help people understand more of the carbon footprint of anything really, so that you can start to know that, just how we know general costs. But that really does then again, put the illness on the individual and we really need to be thinking at a kind of much broader level because you sending that card via using, you know, generated picture with AI to lots of other parents and kids, then the impacts of that individually aren’t that bad.

And in some ways, you know, you’ve not got the car emissions. I don’t really want to go into the whole trade-offs thing. It’s very difficult when if you say that you’ve saved emissions from driving because of that flight example I said before.

But by focusing on that individual perspective, we are then missing the bigger picture.

And if actually if parents want to do something that’s really sustainable for them, then they would have bigger impact by doing things like flying less. And individually, probably best to not worry too much about the smaller uses of AI.

So it sounds like just use it mindfully basically.

Yeah, might be responsibly for good.

So just to summarise, if AI does continue on the current path that it’s taking, we continue to see this increase in demand for AI, what is it doing to our kids planet?

Really good question and one that we do need to answer. I honestly think it’s impossible to know at this point. It’s really hard to know what the future holds for AI in terms of its innovations, how that’s going to be taken up by different people and companies and us in our everyday lives, how it will impact education, jobs, etc.

It is quite unknown about how this will be taken in into society more generally. We do know that the environmental impact of technology is not minimal and with this investment, it will increase. I guess we have to think more about, with our resources that we have, that are finite, water, etc.

What are we wanting to actually use those for? Is it for things like everyday living, our housing or is it for powering our AI use? It’s very difficult to know and I really wouldn’t, I guess I wouldn’t want parents to super worry about this for their kids in some ways.

It is important that we do something about it, but there’s not a lot that parents individually can do unless they’re working in technology or government space. I think just trying to navigate information as it emerges, having conversations like this podcast that we’ve got today is a really good way in which helps people understand and navigate this as we go through this new period of the age of AI.

Kelly, thank you so much for joining us today. It’s been such an interesting conversation. Thank you.

Thank you so much. It’s been lovely to talk to you both.

Kelly, if anyone’s interested to find out more about you and what it is that you’re working on, what’s the best way to do that?

Yes, I think the best thing is to follow the work through the UK Centre for Ecology and Hydrology. We’ve got ongoing research projects on the environmental impacts of digital technology and data centers and taking a more kind of systems approach to that. So yeah, take a look at the UKCH’s work.

I’m also on LinkedIn and happy to connect through that as well.

Well, that was interesting, Holly. I assume now you’re feeling hugely clear. You know exactly how much impact is having on the environment and you can sleep well.

Well, yeah, it’s obviously such a difficult thing to quantify, isn’t it? We sort of know it’s having this environmental impact and different things are having a greater or lesser environmental impact. But I think it’s that sense, isn’t it, of just trying to use it mindfully and use it in a sensible way.

Use it for what you need to use it for and don’t maybe go crazy unnecessarily using it for this, that and the other. I still get this feeling that the use of AI is going to increase, increase, increase and Kelly mentioned there that, did you say that it’s already generating the same amount of energy usage, I think she said, as a small country or as New York City?

New York City, yeah.

You wonder what that looks like in the future. So it is hard to get that sense of exactly what, you know, what is AI doing to the kids’ planet? I mean, we’ve already got, as we know, we’ve already got a planet that’s warmed 1.4 degrees since pre-industrial levels.

You throw in an extra country, that’s the last thing you need.

I have some level of faith that there’s a lot of excitement at the moment around using AI. And so going back to your story there about creating these dolls and, you know, those really cool images, I think it’s a new toy. And I do think the novelty will die down.

If you remember back when we were teenagers, everything went on Facebook. You had a night out, digital cameras were new, so you took so many pictures on your digital camera. You then promptly woke up the next morning and you put all of your pictures from that night out on Facebook.

Your status was unbind broccoli in Sainsbury’s, LOL. Like no one does that anymore. Like it’s ridiculous, I know.

Yeah, exactly. And so I think like we don’t use social media the way we used it back then. I think there is an element of AI being this really new tool and people are trying to figure out how they use it and what it’s capable of.

And it’s really exciting and really fun. And I do think that will die down a little bit.

Didn’t you say it’s just our generation getting really over-excited and the youngsters are like me?

I think it is. I talk a lot to people for my job and generally also for this podcast when I’m preparing with our guests. And those that have teenagers or are connected to teenagers in some way often tell me that if I’m honest, they’re just so non-plus.

It’s almost an uncool tool that their parents use. You know, it’s so uncool to generate.

That’s interesting, like Facebook.

Exactly, exactly. It’s just not cool to do it. I mean, there’s a scale and some kids will find it really cool and exciting.

But anecdotally, just from hearing from different people, it is really interesting to see how I think it is mostly our generation and that are excited by this new tool, whereas it’s to them, it’s just another thing that happens on the internet. And they’ve always had the internet and it’s just not that fun. But the question I think is whether or not, the thing that I found particularly interesting with Kelly was her explanation of what I think she called, did she call it the rebound principle or something like that, where because we can do things far more efficiently with the technology, we now use the technologies much more.

So if something like when we used to connect to the internet to send an email, connecting to the internet, all those noises, used to take, you know, it was a faff, right? You had to press the on button on the computer and it took ages to connect up, then you had to check no one’s using the phone. So sending an email was a mission in itself.

It’s nuts when you think how much that’s changed, isn’t it? It’s crazy when you think now you just pick up your phone and do it.

It’s mad.

Kids will never understand. They’ll never understand.

Exactly. But sending an email, you’d never then just send a quick email to just be like, Hi, how’s it? But because sending an email now is so quick and easy.

Exactly. Exactly. And so I think that’s sort of what she was explaining with.

Now that technology has made everything so efficient, it’s not so much the use of the technology to do those things. It’s just the fact that we now use technology in so many more places. And because it takes far less time, we’re using it so much more and it’s that compounding impact as well.

You can’t just measure sending this email takes this much energy. It’s not as simple as that because there are so many additional… Yeah, I thought it was fascinating.
I thought, and I hope the audience enjoyed it as much as I did. I am a complete nerd.

So it was very much on a level that I love to try and pick things apart.

But hopefully, everyone’s understood it and feels more empowered to make those decisions for themselves.

Yeah, I think we’re just going to have to, as individuals, try and use AI in a… I’ve already said this, but in a mindful way and maybe hope that on a broader scale, governments and AI companies are going to do what they can to continue to implement AI in as sustainable a way as possible. And there were some positive stories there weren’t there using the heat generation to heat homes and that sort of thing.

So I think we’ve got to kind of hope that that’s the route we go down.

I think people are making that decision. I saw there’s a search engine called DuckDuckGo, which is all about, it’s a privacy safe search engine and stuff like that, and they actively promote that AI is optional. It never forces AI on you.

And they’ve seen, since Google announced a couple of weeks ago, that they are going to change Google search so it’s purely AI. I don’t know if you saw this at their big conference. DuckDuckGo, this privacy search engine that allows you to choose not to search with AI, has seen a 30% increase in sign up since that news.

Oh, really? Or a 30% increase in usage since that news. So people clearly are making decisions away from using AI by default, which I think for me personally, is not using AI as a default solution is the way forward.

I think it’s a great technology. I talk about that all the time. I think it has some real purpose and some good and we can have a lot of fun with it.

We shouldn’t be too hard on ourselves, right? But I think it should be an active decision, not a passive. Yeah.

Yeah.

Forced upon you.

Opt in rather than opt out.

If anyone has more questions, we will probably revisit this topic and keep following the progress of what’s happening. We’ll definitely get Kelly back on if there is any big progress to be reported.

In the meantime, have a fabulous week and we shall see you pretty soon.