Can ‘small’ AI have a huge impact in sub Saharan Africa?

Can ‘small’ AI have a huge impact in sub Saharan Africa?
Artificial intelligence is usually fed with large datasets allowing the technology to create what should be well balanced algorithms – but what happens when large data sets aren’t available or the impact of the AI could have a very significant impact on people’s lives? Researchers at Tufts University in the United States have started a program focusing on what they refer to as "small" AI - applications that have significantly positive impacts on individual lives and livelihoods. These applications are specifically designed, using models that have been trained on specialised datasets. Bhaskar Chakravorti, Dean of Global Business at the Fletcher School at Tufts and Abidemi Adisa, the author of the report, explain how farmers in Kenya can benefit significantly from AI. 

Recycling e-waste in Kenya
A UN report recently said electronic waste is rising five times faster than documented e-waste recycling. In 2022, the waste was up over eighty per cent from 2010 and on track to rise by a third by the year 2030. We looked at the problem in Brazil a while back, but Africa has a growing e-waste problem too – and the stuff isn’t recycled in the same way as in developed countries. But could a recycling plant in Nairobi be helping – studio expert Wairimu Gitahi has more. 

The programme is presented by Gareth Mitchell and the studio expert is Wairimu Gitahi.   

More on this week's stories:
Improving the lives of smallholders in rural Kenya with AI
WEEE Centre, Nairobi

Transcript

00:00:00 Gareth Mitchell 

Hello, it's Gareth. Welcome along to Somewhere on Earth. It's Tuesday the 7th of May 2024. We have voices today from the United States and Nairobi, and listeners all over the world. Our latest analytics tell us that you people are out there in 111 countries and territories including in Colombo district in São Paulo, in Central Java and Mexico City. And what's that? 107 other places? Aren't we doing well? Here we go. 

00:00:38 Gareth Mitchell 

All right. And we're going to have a little bit of gossip and generally just brilliant tech geeky company from Wairimu Gitahi today. Hello, Wairimu. How are you as I live and breathe. I'm all right. Good. Thank you. Thanks for asking.  

00:00:47 Wairimu Gitahi 

Hey, guys. 

00:00:49 Wairimu Gitahi 

How are you? 

00:00:55 Gareth Mitchell 

So, what do we need to know as we jump in at the beginning? I always like to get a little vignette from our our kind of expert contributors. 

00:01:02 Wairimu Gitahi 

Yeah, I I mean, I just felt I need to tell you something that I've been thinking about in relation to AI. 

00:01:08 Wairimu Gitahi 

I've always considered myself as an AI optimist, but I'm slowly shifting to calling myself an AI pragmatist. 

00:01:18 Gareth Mitchell 

Ohh, so OK, so not a pessimist, more of a pragmatist. OK, say what? 

00:01:18 Wairimu Gitahi 

Why? 

00:01:21 Wairimu Gitahi 

Yes, exactly. Exactly because. 

00:01:24 Wairimu Gitahi 

The reason why I'm thinking about that is because we can't deny that AI is here. It's ubiquitous. 

00:01:31 Wairimu Gitahi 

It's for everyone. It's not like the way some technology comes in and then it's for the rich. And then the poor can't access or these kind of differences. I mean, generally everybody can access a certain form of AI, ChatGPT and so on and so forth. And so it's here. So we have to accept it’s here. 

00:01:51 Wairimu Gitahi 

The second thing, it's like what I'm thinking about is that it's an extremely capable tool, so it will take you to another level, to a higher level in terms of your profession, I mean you can do, for example, business strategy drafts. 

00:02:07 Wairimu Gitahi 

Of course you have to look at it. We still have need the the human eye. But all the same, it's here. It's extremely capable. No one can deny that. The other third reason and the final one is because it's limited and it's risky. That one we know we we know that it has hallucinations, biases. Sometimes it doesn't know how to predict things well. But all the same it's there. So I feel like these three points have made me like shift. 

00:02:37 Wairimu Gitahi 

So I'm more of a pragmatist. You're just looking at it, accepting it as it is. We're just at the beginning. There are so many things we don't know about it, but we have to accept it’s here. 

00:02:46 Gareth Mitchell 

I like that pragmatism. I think what I'd add is a fairly self-evident point that this is one of the, if not the most profound and important and powerful technologies that humanity has thus invented. 

00:02:58 Gareth Mitchell 

and therefore we need to be careful. We need to go slowly at times and we need regulation. I think is what I'd add to that and I share your pragmatism. Wairimu, thanks very much. Let's jump in. 

00:03:07 Wairimu Gitahi 

Sure. 

00:03:15 Gareth Mitchell 

And coming up today. 

00:03:19 Gareth Mitchell 

To make a big difference, whether you're a pessimist, whether you’re an optimist or a pragmatist, to make a big difference, AI doesn't necessarily have to be big AI. That's one of the messages today as we look at the impact of crop protection technology in sub-Saharan Africa, it's an issue for small scale farmers, but it's a live issue for all of us too,  

00:03:39 Gareth Mitchell 

wherever we are and where, well basically we consume just about everything, don't we these days, cocoa, coffee, rice and bananas. It, much of it comes from small scale farmers. In fact, 1/3 of the food that we eat around the world comes from these small scale farmers. 

00:03:56 Gareth Mitchell 

Meanwhile though, from production to waste, also on the agenda in the podcast, Africa, like many other parts of the world, has a big e-waste problem. So could a recycling plant in Nairobi be helping to alleviate things? That's all right here on the Somewhere on Earth podcast. 

00:04:18 Gareth Mitchell 

But let's start with the tool that helps small scale farmers in sub-Saharan Africa enhance productivity, promote sustainability and boost the economy in their regions. Yes, of course we're talking about AI, but this isn't on the grand scale of the massive machine learning that we often talk about. Well, let's find out a little bit more. We're off to Tufts University in the United States, where they've just embarked on a research programme,  

00:04:41 Gareth Mitchell 

on what they call small AI. In other words, applications with disproportionately large positive impacts on specific lives and livelihoods. These are targeted applications using models trained on specialised data sets. 

00:04:56 Gareth Mitchell 

We're joined by Bhaskar Chakravorti. He's Dean of Global Business at the Fletcher School at Tufts and through its Digital Planet Research Initiative, the school has just released a report called ‘Getting Real about AI for the Bottom-of-the-Pyramid', improving the economic outcomes of smallholder farmers in Africa. Bashkar good to see you again. How's it going? 

00:05:17 Bashkar Chakravorti 

Very good to see you, Gareth. It's going great and I'm really looking forward to our conversation and glad to be back on the programme. 

00:05:26 Gareth Mitchell 

Marvellous. So the report describes itself as a road map for harnessing the potential of AI enabled crop protection technology to enhance yields and deliver dividends at scale to rural communities across the continent. So there we are that sets it up pretty well. Tell us a little bit more about this road map then, Baskar. 

00:05:44 Bashkar Chakravorti 

Yeah, so this road map, Gareth was inspired by, you know what we are seeing as a race and the race is a global one, but much of it is being run by huge, gigantic participants within the United States, trying to get bigger AI models in front of everybody, users and companies, users at home, and these gigantic machine learning models. These AI models gobble up all the data that's available on the Internet and beyond. 

00:06:24 Bashkar Chakravorti 

And so everything is geared towards largeness, and these models are meant to, you know, accomplish all kinds of things, you know, from writing texts to writing code to producing amazing, dazzling graphics and doing all kinds of wonderful things. Now, while this race is happening, and much of our attention is focused on it, much of the hype cycle is also focussed on it. 

00:06:48 Bashkar Chakravorti 

What we would like to put a spotlight on are are much more targeted applications that go towards the small end of the spectrum. So as you mentioned, models that are trained on very specialised data sets and these data sets are, these models are oriented towards solving for a very specific problem, and particularly 

00:07:10 Bashkar Chakravorti 

problems that involve getting critical pieces of knowledge to communities, to sectors that are are starved of knowledge, and if you can inject even a small amount of knowledge into this sector or this community, you can make a huge impact to people's lives and livelihoods.  

00:07:31 Bashkar Chakravorti 

So our first experiment was in small order farming in the developing world and specifically in Sub-Saharan Africa and we wanted to see how much of an impact we could have by injecting a tiny bit of what we call small AI into the farming community and getting a sense of the economic impact on farmers, the impact on the society,  impact on specific parts of the population, women and children and so on.  

00:07:56 Bashkar Chakravorti 

So my colleague Doctor Abidemi Adisa helped to lead this research in trying to evaluate the smaller AI. 

00:08:01 Gareth Mitchell 

That leads in very nicely to where I want to go next, because indeed Doctor Abidemi Adisa is with us with us, as you say, Bhaskar, she's lead author of this report and 

00:08:12 Gareth Mitchell 

you focus, there's loads of different technologies you could focus in on, but specifically you’re looking at AI enabled crop protection technology, aren't you so Abidemi why's that? 

00:08:22 Abidemi Adisa 

Yes. So one of the things that, first of all good to be on the programme, I'm honoured to be here. One of the things that I found in this, during this study and over my years of experience in agriculture is 

00:08:36 Abidemi Adisa 

is one of the challenges that agriculture faces is actually crop loss due to pests. Globally you have about 40% crop loss to pest alone and then in Africa is about actually half of the production goes to waste. So due to pest infect infestation and all that so it was,  

00:08:56 Abidemi Adisa 

for us it was like a good place to start looking at how the farmers, with all their resource constraints, they have limited access to finance and all that, could  

00:09:08 Abidemi Adisa 

easily increase their productivity by not necessarily spending additional money. So what we looked at is an AI application that they can use based on the resource that the resources that is already available with them, which in this case is just a smartphone. So any farmer that has a smartphone or a feature phone that that has a camera  

00:09:31 Abidemi Adisa 

can actually adopt this technology and make use of the technology to improve their productivity. 

00:09:38 Gareth Mitchell 

Would would these be apps effectively like AI driven apps on the phone to tell them for instance, when is a good time to use fertiliser or is it about certain kind of herbicides or  

00:09:51 Gareth Mitchell 

Pesticides. I’m just trying to get an idea practically about the technology that the farmers, the smallholders, have in their hands, driven by AI. What's it actually helping them do? 

00:10:01 Abidemi Adisa 

So the way that the technology that we evaluated works is that the farmer can just take pictures of what their crops look like on the farm and then the the application helps them diagnose  

00:10:14 Abidemi Adisa 

if they have a pest infection or if they have nutritional deficiency on some parts of of their farm and they would be able to to also give them recommendations of what to do. The type of pesticide or fertiliser they need to apply. 

00:10:34 Abidemi Adisa 

And the the good thing about this application is that it's helped them identify the diseases before it is visible to the the physical eye. So it's a way of mitigating the loss before it actually becomes a full blown infestation. 

00:10:54 Gareth Mitchell 

Yeah. OK. So Wairimu is also within this in this conversation with us, of course. So Wairimu, I mean, just getting an idea of why this matters and the extent to which it does. There are some striking statistics in this report. For instance, 32% of global food global, this is food grown globally  

00:11:13 Gareth Mitchell 

is is grown by small holder farmers, so it really matters, doesn't it? Getting some appropriate technology and some help into the hands of these people. 

00:11:23 Wairimu Gitahi 

For sure, it's very, very important because we also know that about 60 to 70% of of farmers, I mean that the agricultural sector employs 60 to 70% of of people in the continent and to the back of that, I like the fact that we're talking here about small holder farmers who are the most vulnerable among farmers and also generally. 

00:11:47 Wairimu Gitahi 

And to that I have a question for you, Doctor Abidemi.  The, one of the entry, for you to be able to use this technology in your phone, you mentioned you need a smartphone or a feature phone. We know very well that this being a vulnerable part of the population, 

00:12:07 Wairimu Gitahi 

getting a smartphone can be very difficult. It's true that now smartphones in a continent are much cheaper, less than $100, but it still remains a challenge for many, many people for sure, including smallholder farmers. 

00:12:20 Wairimu Gitahi 

So doesn't this um, I mean, how are you planning to overcome this entry barrier and be much more inclusive? 

00:12:30 Abidemi Adisa 

So one of the recommendations that we made is how is for the government or to find a way of subsidising or providing like a cooperative fund where farmers can actually have access to a loan, a small scale loan that they can adopt, a  

00:12:51 Abidemi Adisa 

take a loan out and probably purchase a smartphone, but for the analysis of this report, we actually focused on the available resources on ground. So if you look at the first scenario of the report, it was based on the rural adoption of smartphones and the use of digital payments. So what we did was we we combined, we we as assumed that any farmer or rural dweller that is able to make a digital payment has enough digital literacy to be able to adopt the technology and  

00:13:28 Abidemi Adisa 

then if they can make digital payment on their phone, it's almost as good as, they would be able to, they would have a camera which would be useful for this technology. So we are aware of the situation, but that's why we provided some policy recommendations based on those kind of challenges that the farmers will face. 

00:13:49 Gareth Mitchell 

And perhaps we can get an idea of what this means for a farmer's life or a small holders life and a day in their life if you like.  Bashkar, let's come back to you because I see where this report is aiming and we know it's important when we're talking about the crops here. 

00:14:04 Gareth Mitchell 

You know you've got cocoa, cassava, rice, coffee, many staples that we consume all over the world. So I get the bigger picture, but let's just zoom down on that individual small holder. And I'm glad why Wairimu made that point that we can't just assume they magically have a smartphone, but if they have the right kind of handset in their hands, they're setting out to do their day's work. How is this actually helping? What does that look like? 

00:14:31 Bashkar Chakravorti 

Well, what I would imagine, and of course we are sitting, at least I'm sitting quite far away from the ground reality of the farmer. But I can imagine their lives where they have access to such a phone, and it may not be their own phone. It may be part of a community phone. It may be something that they have all manouvre  

00:14:51 Bashkar Chakravorti 

as the epidemic points out, and they realised that their crops are vulnerable and they've they've seen this, you know, many, many seasons in the past. So they are made aware of the possibility of uploading this picture onto a relatively simple website and and then they go about their, you know, their their chores on the farm and everything else that they do. And they very quickly get a recommendation in terms of first, a diagnosis of what is the pest that is affecting their crops. 

00:15:28 Bashkar Chakravorti 

And what kind of actions they may need to take.  Now this recommendation could come in the form of a text message, or it could come in their own language and delivered by audio. Because you know, for many of the farmers this it may be very difficult for them to read. 

00:15:47 Bashkar Chakravorti 

the you know the the the prescriptive aspect of this of this analysis. So it needs to come to them in a language that they can understand and that they can hear. And this too can be enabled by AI. So when we talk about the role of artificial intelligence, it makes its way into a number of different parameters. 

00:16:06 Bashkar Chakravorti 

It matches the picture of the plant or the crop with a a recognised, you know, set of infestations on on a image database. It can then, you know, translate that into a diagnosis and a prescription. And the diagnosis and prescription can be translated into a local language delivered by audio.  

00:16:30 Bashkar Chakravorti 

And then that information the farmer could also share with other farmers who are in, you know, part of the same community. So this is a, you know, a community resource that the farmers can benefit from.  

00:16:46 Bashkar Chakravorti 

And our, you know, our hope is that as this becomes part of their daily routine, part of their, you know, lived reality, they begin to trust this specific application and it becomes, you know, a regular part of farming practise. 

00:17:00 Gareth Mitchell 

Right. And Abidemi, you're you're doing this research. Initially across seven countries, Côte, d'Ivoire, Ethiopia, Kenya, Nigeria, Rwanda, Tanzania and Uganda. 

00:17:11 Gareth Mitchell 

What do we know so far about the the benefits of this technology? Are you able to measure it or predict it so far, maybe in terms of its economic impact, basically, how much richer it might make these farmers and their communities, or in other ways? So what do we know? As far as the you know, the all important evaluation? 

00:17:34 Abidemi Adisa 

Yes. So what we found given the the indicators we used and the analysis, is that an average farmer can actually increase their revenue by an average of 100 U.S. dollars per planting season. Now the reason I included the planting season is some crops can have multiple plant planting seasons for it. For instance, maize can, have planting maturities of about 120 days, so you can have like 2-3 planting seasons of maize in a year.  

00:18:12 Abidemi Adisa 

And to someone in the West, in may look like, oh, that's insignificant, $100 to a farmer might be insignificant, but to a farmer in Africa where his school fees, the school fees of one of the children is probably $50 per per year. That's a lot of money. And that could,  we actually estimated the number of children? 

00:18:32 Abidemi Adisa 

That that could sponsor for one year, which is about 8.9 million children, can have an additional one year education if this technology is adopted and used appropriately. And then we estimated the number of farmers that could benefit across the seven countries to about 14,000,000 farmers in the first scenario, sorry, in the third scenario. So those are the kind of projections.  

00:19:02 Abidemi Adisa 

Overall we expect that the seven countries can have an additional 6.1 billion U.S. dollars in additional revenue just by adopting this technology. So it's it has like, a huge impact both on the individual farmers, thy the economy has large at large, then it flows through the economy. 

00:19:18 Gareth Mitchell 

And their communities. Yeah, yeah. And OK. And helping access to education, which itself has its own kind of flywheel of ongoing economic benefit. Wairimu, just a perhaps a a final or semi final thought from you on all this. 

00:19:33 Wairimu Gitahi 

Yeah, I'm. I'm quite excited about this AI tool. You mentioned some things that I think are very key about, you know, making it much more inclusive. I mean, the issue of language I think is great because it's usually a big barrier. The issue of phones and the issu, but I wanted to ask something about, still  

00:19:53 Wairimu Gitahi 

on that line of inclusivity, there's still a big problem about Internet connectivity in Africa, particularly in the rural areas. 

00:20:02 Wairimu Gitahi 

There is also the issue of gender. Women don't have as much access to phones to the Internet, and I just want to hear what you plan to do about this or what you recommend in the report. 

00:20:15 Gareth Mitchell 

Yeah, and and what? What I'll say, actually, whoever wants to come back on this is I was thinking in our podcast subscription extra we could talk about the gender gap issue. So I think that's a whole conversation itself and it's fascinating. So why don't we hold that for the, 

00:20:29 Gareth Mitchell 

for the extra conversation that we're going on to have, but whoever wants to take this Abidemi and or Bashkar, let's just finish and briefly, if you will on that very important issue of the the digital divide, if you like that we heard there from Wairimu. 

00:20:42 Abidemi Adisa 

Yes, let me go first, then I'll hand over to Bashkar on that. So one of what we found out is all the technology that we actually evaluated can actually work offline, so you only need intermittent Internet access, so that makes it easier for the farmers to probably access their data when they have some data and then get access to the information they need and then still continue to 

00:21:12 Abidemi Adisa 

save images and all and whatever they have until they have Internet access. So it does not require 24/7 access to Internet to be able to use the technology. And as we mentioned, the one of the recommendations that we also included in the report,  

00:21:30 Abidemi Adisa 

that is provision of a more stable access to Internet and all that. But we all know that I'm from Africa, so I know that relying on the government is not necessarily an easy thing to do for Africans. So I'll just leave it there and let Bashkar speak briefly. 

00:21:45 Gareth Mitchell 

Yeah, Bashkar just just finally and briefly. 

00:21:47 Bashkar Chakravorti 

Yeah, I think the the comment that there is a gender divide is a is a very real one. But at the same time, we also know that women farmers, you know, they're quite a they they form quite a substantial segment  

00:22:05 Bashkar Chakravorti 

of the smallholder farming population and challenges of getting access to a smartphone and the challenges to being able to use it, you know, exist across the entire community. So you know this is something that we want to add as part of the policy recommendations, which is, you know, to increase digital literacy and increase digital 

00:22:25 Bashkar Chakravorti 

access, particularly to women farmers, which is an important part of the benefits that accrue not only to the the productivity on the farm but also of productivity beyond the farms, because the added income that comes in, the women then translate it into additional benefits such as sending children to school, such as, you know, saving for the household and these are. 

00:22:52 Bashkar Chakravorti 

You know, add on effects that add to the overall economic impact of what we are talking about. 

00:22:59 Gareth Mitchell 

All right. Well, we will leave it there. Thank you very much. Bhaskar Chakravorti and Dr Abidemi Adisa and we'll see you over on the other side, as it were in our extended subscription version of this podcast. Well, Wairimu, I think we should talk about e-waste now. 

00:23:15 Gareth Mitchell 

I've been reading a few interesting things around this whole subject, like a UN report recently said that electronic waste is rising five times faster than documented e-waste recycling, which was mildly depressing. In 2022 the waste was up over 80% from 2010. 

00:23:32 Gareth Mitchell 

And is on track to rise by 1/3 by the year 2030. And of course we looked at the problem in Brazil a while back. You might remember that dear podcast listener, but Africa has a growing e-waste problem too, and the stuff isn't recycled in the same way as in developed countries.  

00:23:52 Gareth Mitchell 

So Wairimu, I'm looking at you here. Now you're about to tell us about this recycling plant in Nairobi and whether that could be helping. So set this one up for us. 

00:24:00 Wairimu Gitahi 

Sure. Before I even talked talk about the recycling plant, I wanted to talk about. How waste is? I mean, where we take waste. For example, here in Nairobi, we've got a very big dump site, it’s actually one of the biggest dump dump sites in East Africa. It's in a place called Dandora in the Eastlands area of Nairobi. And this dump site was declared full like 20 years ago. 

00:24:25 Wairimu Gitahi 

But every day we have like 800 tonnes of waste being deposited there. 

00:24:31 Wairimu Gitahi 

And then what happens is that you have women, men, children. Sometimes they're digging through the rubbish with, often with their bare hands and trying to look for e-waste because they can be able to sell it. Of course we know that e-waste also has other health hazards. 

00:24:52 Wairimu Gitahi 

Like you know, the mercury, the lead and so on and so forth. 

00:24:55 Wairimu Gitahi 

So it's dangerous, but most of them are like this is the only way we are earning our living and so on and so forth. So the reason why this processing plant called WEEE Centre is important is because they are solving this e-waste problem because as I talked about the 800 tonnes. Increasingly a big percentage of that waste is e-waste, so 

00:25:20 Wairimu Gitahi 

in comes the processing centre WEEE and what it's doing is that it's trying to also collect this waste in a in a way that is not a health hazard. It has different points within the country where people can dump, you know, their computers they're not using, the electronics they're not using.  

00:25:40 Wairimu Gitahi 

It's partnering with companies. It’s partnering with schools and so on and so forth so that that that the waste can be collected in a less hazardous way and either recycled  

00:25:50 Wairimu Gitahi 

or dismantled to be used for other things. Of course, this creates also a problem for the people who I just mentioned, within Dandora because they are poor people. And so I think. 

00:26:00 Gareth Mitchell 

Yeah. So they could lose an income. 

00:26:01 Wairimu Gitahi 

That the way. 

00:26:02 Wairimu Gitahi 

Sorry. 

00:26:04 Gareth Mitchell 

Yeah, sorry, sorry Wairimu. I didn't mean to cut across you. But what? Your point being that they these people who are going through the electronic waste trying to extract some value from it and sell it on, they they could lose an income because of this new centre. 

00:26:15 Wairimu Gitahi 

Sure. So I would like to see a way where they incorporate these people, 

00:26:16 Gareth Mitchell 

Is that it? Yeah. OK. 

00:26:21 Wairimu Gitahi 

in the sense that so that they have an alternative because then it becomes very difficult for them to stop putting their lives in danger. 

00:26:27 Gareth Mitchell 

Yeah, OK. That's. And just to, I'm sure people have worked this out already, but when we talk about WEEE, it stands for Waste from Electrical and Electronic Equipment. So  are you reasonably, are you optimistic, pessimistic or pragmatic then about this centre, do you think it's going on the right lines and especially if it,  

00:26:45 Gareth Mitchell 

and the people running it, think through the implications further downstream, as it were, in terms of communities and individuals who rely on going through all this stuff to just eke out some kind of income. 

00:26:57 Wairimu Gitahi 

Yeah. For this one, I must admit, Gareth, I'm torn in between because I do know that, you know, I recognise that this centre is collecting this waste in a very safe way, in a way that is not a health hazard. But also I have a problem because these other people who live in Dandora are, you know, are poor, are vulnerable people. They will lose their, their source of income. 

00:27:20 Wairimu Gitahi 

So I'll talk like a typical Kenyan or African and say perhaps the government can come in between and help, you know, mitigate the gap. Or maybe this processing centre will also have an idea about mitigating their gap, because for sure they'll lose their their their income. 

00:27:38 Gareth Mitchell 

Yeah, sure. Cause I mean, I was just going to finish off by asking you about what you know, you you say like, a typical Kenyan, how people feel. Are you getting any sense of what feelings are and what sentiment is towards this centre? Or are people even aware of it, for instance? 

00:27:53 Wairimu Gitahi 

Um, from at least the interviews I've heard from from a few journalists who've been there, is that the people are still keen on collecting and they're still keen on, this is the way we are now living. I've been collecting e-waste for from this place for about 20 years. Should I? Why should I stop now? It's true my chest hurts a little, because of, you know, inhaling some of this powder form that, you know that all these electronics are composed of or when they're degrading and so on and so forth. 

00:28:26 Wairimu Gitahi 

But they said, but but they keep on saying, but we have no alternative. So that's why I'm a bit, you know, torn in between it's a good idea, but what about people's livelihoods? 

00:28:38 Gareth Mitchell 

Yeah. OK, very important point. Wairimu, I think we'll leave it there. Thank you very much indeed. If you quick quick credits for you making it all sound lovely today, is Dylan doing the sound engineering for us through the glass? The production manager is Liz, the producer and editor is Ania. You've been hearing Wairimu who’ll be back with us next week and I'm that bloke called Gareth. Thanks for listening, folks. Take care of yourself, won't you? And goodbye.  

ENDS