Can AI solve the global drought crisis?
52:38 Share

Welcome to today's episode of Lexicon. I'm Christopher McFadden, contributing writer for Interesting Engineering. Today we sit down with Joseph Ayoade, an award-winning IT expert and AI specialist who shares how cloud computing, big data and machine learning could revolutionize drought management. From predicted weather models to smart irrigation, we explore the future of AI-driven climate solutions. Let's get into it.

Gift yourself knowledge. RU+ is a premium subscription that unlocks exclusive access to cutting edge stories, expert insights and breakthroughs in science, technology and innovation. Stay ahead with the knowledge that shapes the future. Joseph, thanks for joining us. How are you today? I'm fine. Thank you. How are you Chris? I'm very well. Thank you. Thanks for asking. For our audience's benefit, can you tell us a little bit about yourself, please? Okay.

I'm actively evolving. That's the first thing I'm going to say about myself. I started my journey as a physics and electronics engineer, moved into broadcast engineering. And from broadcast engineering, I decided I want to know more about programming because I needed to solve problems for myself.

my friends who came to me, you know, I have a lot of friends who have one technological issue or the other, and they would ask me, is there a solution you can create for this? And I would tell them, sure, there is. Let's think about it. So until then, everything was going fine. Then a question came and a solution was asked from me if I could make a generator go off at a particular time. And I said, yes, it's possible, but

That didn't work out fine. So I learned that certain solutions are better if your machines are intelligent. And that got me into actively seeking how to understand artificial intelligence. That was all the way since 2015. So I realized that the best way is to go through software development,

then data analytics, and now the property in artificial intelligence because artificial intelligence is based on all these things I have mentioned from software development to data analytics and all that. So that's my journey. That's about me. I believe with this story, I've told you a lot about who I am now. All right? Okay. Great. You're like that ever-evolving. Rolling stone gathers no moss, as they say. Yeah.

Let me see. Right. First question then. So you kind of touched on it, but just expand a bit more. Can you tell us about your technology journey, particularly what drew you to physics, electronics, data analytics and artificial intelligence? Yeah. So from the story I've explained before, you would see that curiosity, curiosity, a desire to keep understanding how things work, a strong push.

which you can't explain, right? So like most scientists would, and I believe you too would, just this really, really desire to solve problems, to understand how things are working. And that's what keeps drawing me down to it. So if you check, you'll find out that after graduating with my bachelor's in physics and electronics, I moved on with a master's in data analytics.

And now also a master's in artificial intelligence. I have developed many projects. I can't even keep track anymore. So worked on several. That's just so much I could talk about myself. So that's what drew me, trying to solve people's problems. And also, look, I would like to mention that when you look at other problems that have been solved by people, it's quite inspiring, right?

You look at the work that has been done by just even the AWS guys, right? You look at how the development of the cloud, you'd see that it's ingenious, right? You remove a lot of worry and headache from people who don't have to go through the pain of infrastructure anymore. I remember that was a living hell before in the past. Leveraging on those tools

infrastructures were like being saved from a nightmare. So you look at how innovations keep helping us as humans and you don't want to stop too. You just want to ensure that you do your bit. Yeah. So that's it.

Absolutely. You mentioned AWS. I think I'm right in saying this. Amazon, the actual store, either doesn't make money or loses money, but AWS is so profitable, it doesn't matter. Yeah, 100%. Incredible. Amazing how that's changed things. How long has AWS been around? When did it first deploy?

I've lost count. I know it's over at least. It's over a decade, going to over 20 years, about 20 years or thereabouts. Yeah. Yeah. I think I remember hearing about it 15 years ago. I could be wrong though. Anyway. Yeah. So changing tracks, but how can artificial intelligence be leveraged to address droughts? And what are some of the most promising AI applications in this area, if any?

Well, now you've touched where I like, you know, this is AI. This looks like where I'm stopping. But I think maybe I'm smart, so they say. I'm not sure about that yet. The AI is a game changer. And if I think I'm smart, AI has been built to be probably 50 times or 100 times faster, smarter, cleaner,

and ingenious. I think we are actively looking at a point where humans, we only have one more thing remaining for them and that's intuition and vision. AI can solve basically most problems now and if you can integrate AI into a machine, that's it. The machine simply becomes alive.

Yes, we know it does not have emotion, but it comes alive. So AI is the game changer. Now for droughts, I've worked on several problems, but I've never seen a problem as, what I call it, pernicious and difficult as drought. Now I come from a part of the world where drought is a pain and you would not be so inspired until you are first-hand

a victim or you experience the impact of drought. Now, yes, you would wonder how has the society or the community been tackling drought all along in the past. There are basic ways that has been used. If they were successful, we wouldn't be here talking about this. So they were not. Is AI solving the problem completely? No, not currently.

there has been no single solution that addresses it completely. And the reason for this is because tackling the drought problem is multifaceted, right? It's more than just deploying algorithms. It has to do with humans. It has to do with governments. It has to do with weather. No one can control weather. Yeah? It's very fickle. But...

Recently, we found out that AI's ability to predict weather is touching over 90%. It's not very high. You wonder how that's possible. There are predictive models that have become very, very promising. You look at LSTM. If you have heard about LSTM, well, if...

anyone who is in the AI field will know about LSTM. Now, when you use LSTMs, you find out that they have the ability to manage sequential data, time series, right? Time sequence, yeah? They can make predictions based on time. But LSTMs are not just the only one. There are models like the Informas, which are very, very...

Well, so far we've tested a bit. They have an edge over LSTMs. So when you talk about solutions, artificial intelligence solutions that are promising, I won't tell you about maybe an organization or a research. I'll tell you the fundamentals, the fundamentals that will make it possible. So fundamentals that will make that possible is the modeling.

The AI models that make this possible, I've mentioned two of them, but we have to talk about hyperparameters. How those models can be tweaked to address these conditions is what matters. Now, you talk about whether we can predict weather, right? What is AI? AI is about prediction. If prediction can go as high as 90%, then you're almost good to go.

So you can have a plus or minus, most three days, one week, your prediction. So it's really promising, I must tell you. It's encouraging to, when you think about it, it's very encouraging. It's something that it's what's delving into. You talk about also when AI is implemented in irrigation strategies. I need to mention that.

Now, the mistake most communities make during drought is they think water is all it's all about, right? Water is not all about that. No, drought is not all about water. In fact, if it's all about water and you give the water, you will lose nutrients because of leaching. So if you just drop it, throwing all the water in, there will be runoff, a lot of runoff because when the soil is baked,

it becomes dusty on top. And when you put in water there, it takes away your soil erosion and leaching. The nutrients go way down. AI comes in there too. Now, AI is not, for example, AI is not going to tell you what amount of water would be necessary, right? And that really gives you the ability to tackle droughts. Then you go one more step, you talk about satellite data, right?

You integrate satellite data, although satellite data is not so useful in this aspect because it's about, I think, a satellite vision. Data can only be about 36 kilometers. So while it's useful on a continental level or on a regional level, we are talking about communities. It's too close.

So, all right. So that's what I can see about how artificial intelligence can be leveraged to address drought. Yeah. Superb. Again, I mean, yeah, I can't actually physically do anything.

It's done its modeling. So it relies on humans to enact any recommendations it makes. You touched on the government or governmental policy being a potential problem with drought. So where do you see governments, especially if you want to speak about from in Africa, where they tend to fail and where they tend to be quite successful with regards to the drought management? It could be as politically honest as you like.

Yeah. You see, Africa is not like every other continent. And every government has different means. Before I go to Africa, I will talk about, let's use Israel. Israel tackled the drought problem before AI. And it's complete. Like, the solution was complete. Yeah.

100%. And they add even 20% more water to give, to sell. Right? All they had to do was, so a lot of desalination, deep, if there's a name it's called, I think it's something deep irrigation where you use tubes underneath the soil so that you don't spray the water on the soil due to surface runoff and evaporation. You sink it right into the soil.

The level of technology deployed by Israel is applaudable. And this were in the 1970s, 1980s, up to the 1990s. So how did Israel make that possible? It's because the government had the willpower to do that. The willpower. And two, they had the financial resources. Willpower is good.

But the resources, Israel is very wealthy. I think the whole world will agree to that, that Israel, excuse me. Now, the whole world will agree to that, that Israel is quite a very wealthy nation. I know about that. Now, when you talk about Africa, Africa as a whole, excuse me, is not on its own. The African regions, the governments are kind of fragmented, right? There's corruption, right?

there is willpower. The willpower might be there, but it lacks the cohesiveness to be able to deal decisively with that problem. And the funding. It costs a lot to implement the type of technology that Israel implemented. So...

In terms of the government, it's very difficult. Currently, there are government organizations, there are governments that are working in tackling drought in Africa. I think there are some in Burkina Faso who are doing really well. And in Mali, Nigeria is also trying a lot. They are doing all they can.

in managing drought and you would ask how are they doing that they're not deploying technology like israel is doing they are approaching it using more traditional means yeah so the people are not largely waiting for the government although there are government programs and nga programs that include planting trees digging ditches and filling it with humus

these are more people-propelled, less costly approach, which is working out fairly, but not as invasively or not as aggressively as Israel's own. So that's what I can say about that. Yeah, that makes sense. I mean, the people on the ground would be more responsive, quicker to act, but with something like drought, you need more of a bigger...

bigger uh what we're trying to say big picture thinking don't you really especially africa's obviously a huge continent so these governments are gonna have to try to find a way to coordinate together right to try and solve this problem i would have thought yeah yeah and then some technology isn't that necessarily that expensive really is it i mean drip irrigation as far as i know is relatively cheap and a very incredible solution

Yes, it is. But you will talk about the skill sets and management and maintenance. Yeah. The skill sets, not so many in Africa. Not so many. It was fair enough. Moving on then. What role does data analytics play in predicting and mitigating the impact of droughts? And could you share some examples of successful implementations, please? Yes. Look, without data analytics, there can't be AI.

Because under data analytics, you have big data, you have visualization, data visualization, and it's interpretation, which AI leverages heavily on. We have data cleaning. You have all that packed into data analytics. And let's say, I would say probably half of what AI does is based on big data.

An AI model cannot make a proper prediction if it does not have data to train on. The more the data, the better your predictions. The more the training, the better your predictions. So data analysis is the backbone of AI. And so

There is no success story with AI without data analytics coming into it. But on a more basic level, if we were not to talk about AI, would data analytics play a role? Has it been useful in the past? Absolutely. Yes, it has. Before AI came in,

There are simple models which are basically under statistics and data analytics, which has been used to predict the weather, used to predict drought, and able to mitigate the impact before it comes. Because it's very difficult to stop drought. You could say, okay, let's do desalination, let's do salting. The salting of the sky is to make rain calm.

Those have downsides too, yeah? But data analytics is what helps people forecast. Without AI, you can at least to a larger extent rely on historical data to form opinion and be able to back it with proof. So if you say, all right, this year from past data that we have, February would have rain.

From past data, yes. When you take your gradient and all that, okay, you could have rain in February. It's becoming difficult to predict with just data due to global warming. Global warming is not a constant. The reason is because the rate of warming up keeps changing every day, the warming up of the earth. And when you throw this into the equation as a function,

Everything about your data changes completely. And that is where AI has to come in. AI can see the pattern, the change. It can follow the trend of the warming up and put in that factor when it wants to make the prediction. Data analytics can't do that. It can't. So data analytics on a lower scale, it's very useful because when you have sensor data, when you have satellite data, when you have weather data,

you pull all that together, that's a massive data, a lot of features and instances all packed together. With that data, you can do some really good maths with basic, maybe you do some, what's it called? You do some ARIMA, right? You could, to an extent, figure out what will likely happen. And also,

Data Analytics is very, very useful in terms of your deployments in the farm.

For the farmers, like, okay, on average, if we have this, we know what amount of water we need. We know what time of the day, on average, that we'll have to carry out the irrigation. We know what amount of fertilizers to use on average. When we used this amount of fertilizer two years ago, the yield was this. When we used this amount of fertilizer five years ago, this was the yield. Okay, let's check.

What's the yield today? That's data analytics. It makes that possible for you to do all that. So there has been successful implementations, of course. And before AI, it was all about data analytics and making these decisions. So that's all I can say about that. That's fair enough. I mean, with an artificial intelligence, it would include an element of data analytics anyway, right? They're not...

separate entities. There's a lot of overlap, isn't there? It's pretty important to understand that. Okay. Fair enough. So given your experience with cloud infrastructure like AWS, we mentioned, and Firebase, I think, how critical are these platforms in deploying scalable drought mitigation solutions? Okay. Um...

I don't want to say this because some people will be like, oh, are you a marketer for AWS? But no, I'm not. I'm just trying to look at how visible it is for you to create a platform where you can manage to contain massive millions of instances of data

and features, probably features running to hundreds or probably over 50 features. And you have that working on your hardware. And I'm also wondering if you have the compute power and how many of that you have to get. I'm also wondering how you are going to be able to get the technical skill to build a network that will be powerful enough

to manage that. Yes, there are a lot of network engineers. Yes, there are a lot of sysadmins who are really good. How much would that cost you? And then you talk about management. Look at all that work. So when you talk about cloud infrastructures, I have leveraged on it. I would say maybe I can't do without it because...

it takes away a lot of headache and work where I need to spend one year building an infrastructure, a minimum of one year building an infrastructure that I can deploy my AI on and which might not be scalable enough. Now, let's have an example. For one of the organizations I work, I'm not going to mention the name, there are certain servers, really good servers. As of then in 2010, between 2005 and 2010, really good servers which were acquired

for some tasks, right? They were really good. They were the latest to them. They were the best. Working powerful, you know, when it was deployed, the whole network, everything was going fast. Business was moving. More customers came in. The company grew. 10 years later, fast forward. The company was struggling. There were so many downtimes. What's the reason for this? Scalability.

The scale of users, the success resulted in increase of users, right? Increase of users automatically increase in traffic. While you can have easily, probably increase your network, your bandwidth, your throughput to yourself, do you have equipment that can manage that traffic coming in? Are your switches working?

Are your layers powerful enough to withstand that? So while we could improve the pipelines, the equipment could not bear that load. So we keep having the bottoming, right? The equipment's peak, they fail. The peak, they fail. So what do we do? Typically, you keep increasing your processing because you don't want to change the whole hardware. If you want to change the hardware, you have to start configuring afresh again.

You have to talk about how you are going to also make it possible for different hardware to work together. New ones, you get new ones, you deploy it. How does it work with the old ones? So the upgrade is very difficult. So that can be very, very cumbersome. So when you talk about AWS...

or you talk about Firebase, you don't have to worry about all those things. You want, it's elastic. That's what they call it, EC2, right? For instance, EC2 Elastic, right? So with Elastic computing, you can keep upgrading, right?

non-stop you get more customers you're great you get more data you're great you get more sensor readings you're great you keep and you don't have to actively do the upgrade the mansions the cloud servers you are using already has algorithms in them to know that oh your traffic is going on and then on its own it scales to meet the demand right so um

These cloud platforms are critical and they make scalable AI possible. I know that you would wonder, is it possible to have AI that is not scalable, that remains where it is? Yes, it is. Yes, it's just that you will be stuck in the past. But your AI has to continue to evolve. And if it continues to evolve, you have to keep getting...

new processing power. There was a time when you have AI5, you could do AI, you could run some models, traditional models, you could run probably, you could go as much as XGBoost sometimes with I7, you could do extreme gradient boosting, you could do some assemble methods, traditional AI assemble methods with those types. If you are stock day fan. But when you need neural networks,

where you have to move up, you have to do CNN, you have to do RNA, then you can no longer use I7. You can no longer use I5. So you now have to proceed, move forward and be scalable. To do that, you need compute engines to work with. And the way things are going, you also have to talk about quantum computing now.

You have to think about that now because, yes, we have already touched the peak of processing requirements. Our silicon chips, so far we know, can no longer cope with the demand, with the compute demand that we need. So now we have to move forward. We have to move and get that. Microsoft talked about the test about three weeks ago.

They released their first quantum chip. There are still a lot of bottlenecks with that. So cloud, even with all these things, cloud still shines, right? It still shines. You know SageMaker, right? AWS SageMaker. It does well in crunching ML models, right? It does well. And it can also handle terabytes of satellite data. Firebase, right?

On the other hand, you know, the day I discovered Firebase, that day I felt like I've been given a... When you're so tired, let's say you're tired and you're just weak and somebody just gives you caffeine. That was what happened. Firebase is fantastic, especially when you have to deliver mobile apps, mobile applications that work with AI.

You don't have to worry about listeners. You don't have to talk about events. It's inbuilt into Firebase and Firestore, right? They take care of all that for you. And there is no limit. There's no limit. It's nonstop. It can keep the storage capability of Firebase. It's like we said just like before, elastic. It keeps...

expanding it keeps expanding but then I know you talk about the cost too right it costs a lot sometimes because the higher your traffic the more your payload then you have to but if you have higher traffic then you should be making more money so that should be that should be it should be proportional right yeah so that's it

I'm just wondering with basically everybody jumping on the AI bandwagon, so to speak, it's going to put a lot of pressure on these servers to, what can I say, the demand for them to expand. I wonder if the pace of AI will ever outstrip

how these servers could adapt. Because, I mean, somebody has to upgrade the servers, you know, even with Amazon. Right? So, whatever region they're at. Yeah, it's going to outpace them completely. We are on a countdown. But, you know, they're not sleeping, too. They're actively working to do that, yeah. Absolutely, yeah. Anyway, okay. Right, so...

From your international experience with regards to droughts, how do different regions approach drought challenges and what lessons can be shared globally, if any? Okay, so this time I'll start from Africa. I'm starting from Africa because it's the most basic approach. Yeah, so the most basic approach is drought.

Usually the best, to be honest. Yeah, which is the best because it's usually the side effects or the side, the problems that might result are very minimal because it's always mostly natural. Let's take a very simple one that was invented in, is that Burkina Faso long ago by a man

I think the man's name should be Yakub or something. So what happened is the drought is making things. I think that's Yakuba. Yeah, Yakuba, yeah. Yakuba. Yeah, he does not have AI. But what he had to do, he had a shovel, right? So since he has a shovel and he has a bit of tools, he knew that there was one way out. And that's by adding humus to the soil.

So what he went is, what he did is, like, if you remember what I said at the beginning, I told you, it's not about just water when it comes to drought. No. It's about water management, soil and water management, a combination of both. So you take up soil management, half of the problem is solved. And that's what Yakuba did. So what he did was that he went around digging ditches, just digging.

holes around and he fills it with humus, with manure, organic matter, and he keeps pouring it in it. When the rain falls, water collects in these shallow pits, right? Water collects in them and they kind of make the humus active organically. And that resulted in trees falling

And certain types of plants like sorghum, like millet that do really well in a rich environment, that made them bloom. In essence, the hard-baked cracked soil soon became a plantation of sorghum and millet. So traditional way, perfect, doesn't affect nature, no salting, no side effects, just straight up, boom, working, hectares of land, deserts, droughts.

was resurrected, so to say. What I would call this is that it's natural, it's very, very helpful, but there is a downside, time. Time, you have to wait. Because your demand...

Can it cope with the wait? So that's the question. Can your demand cope with the wait? So if you have no choice, your demand has to cope with the wait time for the implementation of these things, which is what Africa unfortunately has. Africa has to wait. So that solution worked well.

In most parts of the sub-Saharan Africa, typically before, you know, Africa is kind of divided into two. We have the part above the Sahara and the part below, beneath the Sahara. Now, the part beneath the Sahara is where these problems happen. That fringe, the fringe of the Sahara, where people actively live, civilization is there.

That is where we have these challenges. Now, as you come southwards, this is less of a problem because now you have the Atlantic Ocean. Moisture coming in, the wind of the Atlantic Ocean coming in between around February to September. So drought is not really a major issue at that period. But as you progress northwards towards the Sahara, drought is a problem. How is it being mitigated?

Ways I've told you, dick beats, we're putting some humors there. Plant trees, trees are really helping. There has been a lot of aggressive push

NGOs planting thousands and thousands of trees in these regions. It's working beautifully well. It's working and it's even becoming a problem. It's becoming like a concern because when you look at the Sahara Desert, it's gradually becoming green. If you look, there's a report by, there's a report around last year or there about early last year where the Sahara Desert was noticed to be

some parts of Sahara Desert is blooming green. If the Sahara Desert, if we lose the Sahara Desert to become a farmland and it's actively green, you would say, lovely, now we have more money for Africa, right? But there's a downside because the Sahara Desert is what keeps the balance of the ecosystem, is a major player in the ecosystem of the world. There are some winds, some air movements,

Aerosols that move from the Sahara Desert every year in massive tons that move into the Atlantic Ocean and to the Amazon forest in South America. They travel that far. Now, those aerosols contain phosphorus. They contain very low nutrients that are crucial for planktons. Okay.

nutrients that are crucial for the forest, the Amazon forest. You lose the Sahara Desert, those nutrients become unavailable. What you get is a decline in the number of planktons in the sea. And also it impacts the Amazon forest. If the number of planktons in the sea drops, what do the whales eat? What happens to the food chain?

So while the Sahara Desert might be like a curse and the drought that comes with it, we still have to... I'm sorry about that, Liz. No problem. While the Sahara Desert might look like a curse with the drought, it also plays a pivotal role in the balance of the ecosystem of the world. So that's for Africa. I've given you quite much more than you require in terms of tackling droughts in Africa.

You talk about India. India is a very massive country. In fact, sometimes I say India is a continent. So there is...

It's common in India. But India has done a lot of work in terms of drought. They've deployed drip irrigation, right? It's interesting that India actually deployed drip irrigation in tackling drought. And I've mentioned Israel's method by using desalination methods

and taking water from the sea. When I was in Dubai, I noticed that, I wondered, how is it that a desert is blooming? So they tackled it with both desalination and salting, right? They make the rain calm and...

although it seems artificial, but it's working. It's working. So you have more water than you even need. There was a flood even last year in Dubai, which was rare. So that's about that. The Israel's national water carrier in the 60s still remains for me the most ingenious, the most powerful, because that's...

For the fact that you have a problem before and now you don't have 20% more water, don't even give up. It's still the best approach for that region. Now, it might not be applicable. You might not be able to deploy something like the National Water Carrier concept in a place like Mali. You might find it difficult. Why? Because while Israel is not landlocked, Mali is landlocked.

Is it Mali or Burkina Faso? One of the two of them. Either Niger, Mali or Burkina Faso. They are landlocked countries. There is no sea. So desalination is absolutely impossible. If you say it's possible, you can say, okay, you take permission from other countries like Nigeria and then you sign an agreement. You get your pipes to the sea. It's still a lot of work and dependence is there. And it's very expensive. I don't know if...

if a country like Nigeria can, for example, provide, fund such a massive project. So we wouldn't say somewhere like Nigeria would have to deploy a national water carrier. They might not be able to do that. They have to look towards other ways of solving the problem.

So that's all I can say about that. Yeah. That's fair enough. Yeah. As a nation, like if it is Burkina Faso, you mentioned, I don't think I'd feel comfortable trusting my neighbors to not destroy that pipe if it's so critical, you know, to my future. Yeah.

Not just picking on Africa, anywhere in the world. Would you give your potential enemy that much power, really? Much power. If you think about that, recently, just last year too, there has been a lot of political disturbances. You are aware? Where Burkina Faso were having challenges with their neighbours. Let's not go into that. Let's not talk politics. So let's move forward. Yeah, absolutely. What else was I going to say?

With anything that humans do to solve a problem, there's always unintended consequences we haven't thought about or simply completely ignorant of. So with, like you say, the greening of the Sahara Desert and its consequences on other parts of the world, other ecosystems, this is another area AI could be very, very powerful, couldn't it? It's like, right, we want to do this to this area. You'll say, yeah, you could do that, but...

there's a 50% chance, I don't know, of you impacting this part of the world. It could lead to an increase in flooding. Yeah, be careful what you want, what you wish for. That talks about the considerations of AI.

That's right, yeah, absolutely. We did have a question there but we're running out of time so I'll just jump to the end basically and then we'll close up. I thought it's been fascinating though. Let me see, so what advancements in AI and data science do you foresee having the most significant impact on drought management over the next decade? If you can answer that. Okay, so what advancements in AI and

which will be pivoted in the future? Yes, new models are coming up. New transformers are doing really well. The informer I mentioned before is a transformer. And you heard about Grok. That's an LL. It's a large language model, but they also use transformers, right? You think about DeepSeek too.

You combine DeepSeq, you combine Grok, you combine Lama, you combine Tagipiti, you look at the progress they've made within three months. That progress is exponential. If AI continues to grow at this rate, before the end of the year, there's a likelihood we will run out of compute power. Great. Yeah. So, yeah.

We can't talk about AI moving forward if we don't have the compute power. So we have to first all talk about, are there infrastructures that can cope with the growth of AI? Are they readily available? Are they affordable? The answer is no, not yet. So we are waiting probably for quantum computing to move forward. Without that quantum computing, without that speed, without that processing power, AI may develop, AI may get better,

but it will work like a malnourished child or a malnourished person. It would not really grow that fast enough. It would grow, don't get me wrong, but the growth would not be as projected. It would not be as projected as we are thinking. So we would say that

The advances in AI currently at the pace it's going would successfully, yes. Already it's doing that. It's solving a lot of problems for people all over the world. But there are bigger problems that are yet to be able, the AI is yet to be able to solve completely. One of them is also the drought. The drought, yeah?

So, yeah. So can AI solve the problem of drought completely? When you talk about solving the problem of drought, the problem of drought is, like I said, it's multifaceted. So even if AI gives you the model that can accurately 99% predict weather and the future, AI can't control the locals. It can't. And AI can't control the government.

So while AI progressively is getting better and improving, there are still other requirements that AI can't meet. So this, as it applies to the drought, so it applies to every other part of human aspirations and endeavors and problems. So people say AI is going to take jobs, AI is going to take jobs. I tell them, no, AI will increase the jobs,

Because when AI gives a problem, we need infrastructures to implement it. We need government to fund it. So manpower comes in. New jobs, new type of jobs, new skill sets is being required. So AI improving is 99.9% absolute. Yes, it's going to be better. The borders are going to get better. But it fulcrums on its dear compute power

Are all the factors in consideration available to meet the growth? If those things are missing, the growth would peak and stay where it is. It's not going to progress more than that. So that's the point where we are now. Interesting. I haven't considered that. So it will peak and plateau unless something happens. And then it can free it up to keep growing. Fascinating.

Excellent. Okay. I hadn't thought of that. I hadn't considered that. Interesting. Right, well, that's all my questions, Joseph. Is there anything else you'd like to add that you feel is important we haven't mentioned? Well, um...

Maybe there's really not much I have to talk about anymore. I've said practically everything that's on my mind. I was excited about this. I'm still happy I'm having this right now. Thank you very much for the opportunity. Our pleasure. Christopher, thank you so much. Thanks to Interest in Engineering for this opportunity to talk about AI, which I love doing a lot. So I also have young people who...

look up to me and would have liked to know more about AI and how it works. People have very weird ideas about AI. If you listen to what people are saying about AI, you could like, no, no, no, no. You don't need to get freaked out.

It's awesome. You're not going to lose your job. You're going to be fine. It will be Skynet either, yeah. Yeah, yeah. I mean, you did a great job making people so scared about that. So it's not that. It's not going to happen. Be it Matrix, be it Terminator, no. So the young chaps, you can relax, you can throw in your energy and help to improve this technology, which is one way

of rescuing us later as we find out from the arts, helping us from probably other disasters that might happen, predicting disasters and all that. This is our game changer. Brilliant. With that then, thank you for your time, Joseph. That was very interesting. Also, don't forget to subscribe to IE Plus for premium insights and exclusive content.