Investing in Regenerative Agriculture and Food
Investing in Regenerative Agriculture and Food podcast features the pioneers in the regenerative food and agriculture space to learn more on how to put our money to work to regenerate soil, people, local communities and ecosystems while making an appropriate and fair return. Hosted by Koen van Seijen.
Investing in Regenerative Agriculture and Food
351 Paul Clarke - Smart Machines, AI and Modeling: engineering our way out
A conversation with Paul Clarke, technologist, innovator, inventor about technology and innovative tools from various domains, including modelling, digital twins, digital shadows, robots, and other smart hardware solutions that are crucial for the regenerative transition—tools we’ve barely begun to consider, let alone adopt.
We often hear about AI, machine learning, and large language models, but these represent only a fraction of what is currently available. Paul argues that the challenges we face are so immense that we cannot afford to ignore the potential of these technologies. They are essential for building better farms, advancing farm technologies, creating smarter robots and hardware, developing improved food systems, optimising food warehouses, and so much more.
More about this episode on https://investinginregenerativeagriculture.com/paul-clarke.
This podcast is part of the AI 4 Soil Health project which aims to help farmers and policy makers by providing new tools powered by AI to monitor and predict soil health across Europe. For more information visit ai4soilhealth.eu.
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.
This work has received funding from UK Research and Innovation (UKRI) under the UK government’s Horizon Europe funding guarantee [grant numbers 10053484, 1005216, 1006329].
This work has received funding from the Swiss State Secretariat for Education, Research and Innovation (SERI).
==========================
In Investing in Regenerative Agriculture and Food podcast show we talk to the pioneers in the regenerative food and agriculture space to learn more on how to put our money to work to regenerate soil, people, local communities and ecosystems while making an appropriate and fair return. Hosted by Koen van Seijen.
==========================
👩🏻💻 VISIT OUR WEBSITE https://investinginregenerativeagriculture.com/
📚 JOIN OUR VIDEO COURSE: https://investinginregenerativeagriculture.com/course/
💪🏻 SUPPORT OUR WORK https://investinginregenag.gumroad.com
- Share it
- Give a 5-star rating
- Buy us a coffee… or a meal! www.Ko-fi.com/regenerativeagriculture
==========================
🎙 LISTEN TO OUR PODCAST AND SUBSCRIBE TO OUR CHANNEL ON
🎧 Spotify https://open.spotify.com/show/4b7mzk8c9VNM7HX5P3pM4u
🎧 Apple Podcast https://podcasts.apple.com/us/
Thoughts? Ideas? Questions? Send us a message!
Find out more about our Generation-Re investment syndicate:
https://gen-re.land/
https://regenerativo.org/en/laris/
Feedback, ideas, suggestions?
- Twitter @KoenvanSeijen
- Get in touch www.investinginregenerativeagriculture.com
Join our newsletter on www.eepurl.com/cxU33P!
Support the show
Thanks for listening and sharing!
This is a very different conversation compared to what we usually have no deep dive into soil science, regenerative agriculture practices or how to sell nutrient-dense food. Instead, we have an absolute technology expert joining us to talk about a whole suite of technologies in different families, ranging from modeling to digital twins, digital shadows, robots and other smart hardware solutions that are so needed in our regenerative transition, but we barely started even thinking about them, let alone using them. Other smart hardware solutions that are so needed in our regenerative transition, but we barely started even thinking about them, let alone using them. You hear all the time AI, machine learning, large language models, but that's only a small subset of what's available to us now. And our guest of today argues we are so deep into trouble that we can't afford not to model. We can't afford not to use these technologies to build better, better farms, better farm technologies, better robots, better hardware, better food systems, better food warehouses and so much more. This is the Investing in Regenerative Agriculture and Food podcast Investing as if the planet mattered, mattered, where we talk to the pioneers in the regenerative food and agriculture space to learn more on how to put our money to work to regenerate soil, people, local communities and ecosystems while making an appropriate and fair return. Why my focus on soil and regeneration? Because so many of the pressing issues we face today have their roots in how we treat our land and our sea, grow our food, what we eat, wear and consume, and it's time that we, as investors big and small and consumers, start paying much more attention to the dirt slash, soil underneath our feet. To make it easy for fans to support our work, we launched our membership community and so many of you have joined us as a member. Thank you. If our work created value for you and if you have the means and only if you have the means consider joining us. Find out more on gumroadcom slash investing in RegenAg that is, gumroadcom slash investing in RegenAg or find the link below.
Speaker 1:Welcome to another episode today with technologist, innovator, inventor and a leader. Welcome, paul. Hi, good to be here and first of all, a shout out to Henry Dimbleby, who put us in touch and who already mentioned you in our two hour long deep dive. Don't go and listen to that now, people. I'll put a link below if you haven't listened to that yet.
Speaker 1:But mentioned digital twins and said you really, really have to talk to Paul. And here we are and I agree, we had a pre-call and I had the pleasure of doing a bit of background research into your fascinating journey into technology and, of course, the world of food and agriculture. You're definitely not deep in the world of regenerative food and agriculture, but you've played a large role in moving a lot of food, like more food than most people will ever touch in their lives, and so I would love to unpack that what you see, what you've noticed, what are the opportunities of digital twins and other deep technologies that we need, that are already there, that we might not even know how to use yet. So I'm really looking forward to that. This will be a wide-ranging interview and conversation, but first of all, starting with a personal question in this case, how did you roll into the world of technology?
Speaker 2:Well, I suppose as a child, you know, when many of my contemporaries were reading kind of I don't know Innie Blyton or books like that, you know, I was normally reading instruction manuals for cars and things like that and taking stuff apart and putting it together again and maybe making new things out of those parts.
Speaker 2:So I was always fascinated by how things work and how things are built and by machines of all sorts, and so that love of how the world works kind of drove me to studying physics at university. But even then I was fascinated by the kind of the early days of personal computers and I nearly dropped out of my physics degree to do a first startup. But I always kind of knew that I wanted to work with computers and technology, and probably within startups, and so that led me into the computer industry and sure enough, about two years in the opportunity came to do a first startup and I grabbed it. So it's really it's been that way from the start and I suppose over that journey the technologies may have changed and evolved, but many of the themes have continued, particularly the excitement of working at the interface of software and hardware systems. You know, things that move under software control are much more exciting than just on a screen.
Speaker 1:And was it always the case for you like that interface or that connection between the hardware, physical world and the software world, or has that been gradually over time, like your first startup or the first companies you you worked at and with? Was that predominantly the, the software side, or also the, the stuff you could touch the hardware side?
Speaker 2:no, I think earlier on the startups that I worked in were more software focused, but I suppose then the the physical world started to come in in terms of sort of the more real-time applications, so software that interacts with the real world in terms of control, and then that led on to the whole area of what I would now call smart machines. Others might call them robots, but I think you know increasingly that term robots is too narrow to describe the variety and area of smart machines and related technologies to that, including what the modeling technologies you know, like digital twins that are hugely important in the gestation cycle, if you like, of those smart machines and I don't know if you are, if you were or are a foodie at all.
Speaker 1:but let let's say, when did the food world enter your professional life, not just your kitchen? In your journey and there was quite a profound one, I think, in terms of scale, in terms of technology, in terms of moving parts, in terms of especially, let's say, moving things that can go bad quite quickly if they are on the wrong temperature or are put on top of each other in the wrong order. It's very different from moving. I mean, probably also clothes are tricky and a lot of other materials are tricky, but food, specifically fresh produce, even more makes it a very interesting challenge. How did that happen? How did you roll into the world of food? Or maybe it was always a dream to be working on food and agriculture?
Speaker 2:So I suppose the place at which I or the point at which I got involved in moving food was when I joined CARDO and that kind of happened by accident. Somebody gave my name to a recruitment consultant and they persuaded me to have a chat with somebody and they then in turn persuaded me to go and have a look at their first generation warehouse in Hatfield. And when I stepped into that kind of Aladdin's cave of technology I was completely blown away. It was their first generation warehouse in Hatfield, so it was conveyor based and a long way from where the technology got to during my time there.
Speaker 1:What was so impressive when you stepped in? Just for people to understand was it the noise, was it the smell, was it the size? Just for you, coming from what you were doing before.
Speaker 2:So I'd had some involvement, certainly with robots, before that, but nothing like this. This was like a roller coaster or a theme park of food.
Speaker 2:It was, you know boxes of groceries flying around over many kilometers of the conveyor and being, if you like, controlled, like you know, trains on a track taking the right kind of turning left and right at junctions and eventually ending up in the back of vans. And so it was all about machines for orchestrating, if you like, that picking and packing of grocery, controlled by very sophisticated software. And I was just blown away, you know, I just thought this is like the ultimate kind of train set that I've stepped into and it was mesmerizing, you know. And I think a few months later I joined to do a one year consultancy project. I ended up then joining permanently and I ended up staying 15 years, which, given the journey I had been on with startups, I was very, you know, surprised by.
Speaker 2:In a sense, it was never my game plan to do that, but it was a kind of a, you know, it was a love affair with, with the technology, but also with the culture, you know, of the company, which was incredibly inventive and creative.
Speaker 2:And, you know, very like what you experience in startups and I had experienced in other startups, even though by then, by 2006, when I joined, it was still quite a big company and by 2006, when I joined it was still quite a big company, and so I think it was this blend of technology and invention that attracted me, and during that period I became CTO, I grew a huge technology division and then handed that over, and then for the last three years I ran the kind of advanced research side which is, you know, for those people listening who know organizations like ARPA or, in the UK, ARIA, you know it's the sort of the sharp end of the innovation spike or the innovation factory at Ocado Atacado, and we were looking at all sorts of future potential technologies in that division that might be relevant and applications that might be relevant to the future of the company. And then, in 2020, I left and decided that what I wanted to do was Right in COVID.
Speaker 1:Well, yes, it was in the midst of COVID, which wasn't easy, but it was Interesting times for online retail. Yeah, well, yes, I mean, it was an the midst of COVID, which was an easy, but it was.
Speaker 2:Interesting times for online retail, yeah, well, yes, I mean it was an accelerant, certainly, you know.
Speaker 2:Arguably, online grocery delivery became sort of the fifth emergency service. You know, it was something that a lot of people relied on, but also all sorts of kind of automated processes became important because of trying to keep people out of areas that were dangerous and indeed there was a desire for all sorts of other smart machines to help deal with the pandemic. But you know, one of the challenges with smart machines is that they have a different gestation cycle. To what does that mean? Well, as in if you think about how a new piece of technology is ideated and then designed and created and tested and prototyped and eventually put into production, you know that's what I would refer to its kind of gestation cycle. As in, you know, a gestation cycle of an animal or a human, but software and hardware has a gestation or a life cycle and a gestation cycle. So it's and digital technologies, software and AI and so forth. It's very different, and the digital and the physical worlds are different and the really exciting thing, things happen when the two come together.
Speaker 1:So you need which is such an interesting point, that because I think many people underestimate which is such an interesting point. Because I think many people underestimate and I will put a video below of what you also did at Ocado because retail online is incredibly difficult, incredibly small margins and cutting edge and, at scale, you need to move a lot of things very, very, very smartly and and that led to an incredible cycle of innovation and from you and and others as well, to to what? How do you push the boundaries, how do you do this smarter, faster? Which 10 years ago, when we thought, or 20 years maybe, when okado started online supermarket or online groceries or retail in general, was more almost like picking from the supermarket itself, like the same layout, the same, like it was basically just moving it online, which, of course, is not ideal, just as we're still stuck with the same typewriter letters on our laptop, which is not ideal. Um, so, just to the immense innovation that has happened there by merging or by using both the hardware, the robotics, which has changed dramatically, I think, over the last years, and the incredible sort of director software. We now have to do the dances, and COVID has pushed that and pushed it to the limits, probably in certain extents, but now you Like, what do you see now, with the possibilities we have on all the suite of technologies?
Speaker 1:Like it's not one, it's not just digital twins, it's not just living labs, it's not just the swarms of robotics you've built? Like if we, I don't know, lay of the land is maybe the wrong intro into we're now talking in November 2024. It seems like everything is accelerating, but what do you see where those physical and non-physical will touch? Like, what are exciting pieces for you? Of course, we're interested in food and ag, but I'm also just having you here to see, ok, what should we know in food and agriculture? What is what is exciting to you? What is interesting, what is meaningful?
Speaker 2:Gosh that's a good question or big question question. I think the first thing to say is there is an incredible frenzy and focus, you know, on AI at the moment, and arguably not AI in its inclusive, its wider sense, but in large language models and generative AI, and they're incredibly important and powerful technologies. I was on the AI Council in the UK and helping to, you know, be a part of a group advising government and you know I'm a huge fan of AI in all of its flavors and it is a. It's a much bigger toolkit, if you like, than just large language models and generative AI. But those are what is in ascendance at the moment and that's what's kind of captivated. But those are what is in a sentence at the moment and that's what's kind of captivated, if you like. People's attention, including, just, you know, normal people for whom AI wasn't a big part of their world, and for some it still isn't, but for many, many more now it is because it's become much more accessible. So AI has had its kind of chat GPT moment. Some other technologies haven't yet, but the point is it is one of the other flavors of AI, is embodied AI, which is when you put that kind of intelligence, if we want to call it that, inside a machine, inside a smart machine, and it makes the machine. That's the smarts, if you like, in the smart machine. But it's a symbiotic relationship because the smart machine also can be the arms and legs of AI. It can allow AI to escape, if you like, from inside your desktop or whatever, or a server, and get out into the physical world and do that with scalability and agency. So smart machines as they go around, often they're controlled or orchestrated by sort of centralized kind of AI systems, and sometimes they're autonomous. So sometimes, like autonomous vehicles or drones, they're able to perform without minute-to-minute control. They probably still need some degree of orchestration because they need to know what you want them to do. But it's a blend of the two, and so now we've got AI and now we've got smart machines. And then the third member of the holy trinity, if you like, is what I would call synthetic environments. So synthetic environments is a family, another family like AI and, in fact, another family like smart machines, because there are many different kinds of smart machines.
Speaker 2:It's probably worth going on Synthetic Farmers, just to point out that you often say to people you know, so what robots do you have in your home. Well, I've done that. And they go I don't have robots. I said, well, I think you do. And they say what I said well, do you have a washing machine? They say, oh yes, it is a robot. And the point is, the most successful robots are the ones that we don't think of as robots anymore. So a drone is a robot, a 3D printer is a robot or a smart machine, an autonomous vehicle. In fact, a normal connected car is a very sophisticated smart machine with hundreds of CPUs in it controlling everything that's happening. So there are lots of things out there, lots of smart machines out there, hiding in plain sight.
Speaker 2:You don't recognize, you just don't necessarily think of them as that. So there is a whole family of smart machines, there's a whole family of AI technologies and then there's a family of modeling technologies called synthetic environments. In that family it's got five members. It's simulations, emulations, visualizations, digital shadows and digital twins, and once again they are very different, they're complementary, but they form a family and once again there is an evolutionary cycle in there in the sense that a know everyone, a lot of people are talking about digital twins now and it's in the zeitgeist and you know it's got people's attention and that's great.
Speaker 2:I'm a huge fan of digital twins, but unfortunately the term has now been so hijacked. Poor henry had to, has had to have uh, listen to me bend his ear about this, because unfortunately a lot of people talk about digital twins. You know they don't mean digital twins. Sometimes what they're talking about is just a static sort of 3D diagram or render of a physical thing in the real world. That's a powerful thing. It may be very useful. It's not a digital twin. And the key thing is a digital twin has to have that modeling element, you know. It has to be able to predict the future. It's not just a data store, it's not just a sort of a collection of data about some physical asset, the physical twin, but then the other key thing is it needs to be connected to its physical twin in both directions. So if we think of an example of you know a weather forecasting system, that's a digital shadow. It takes a lot of data in from you know weather stations around the world, it models it, it predicts the weather in the future, and then you know, we get it on the news, so to speak. That's a digital shadow because the connection is in one direction. What makes a true digital twin, in my terms and in most people's terms, is that then, when you take those insights and you use it to control the real world, then you've created a digital twin.
Speaker 2:So, anyway, the point is there is a family there of technologies. What they do is they allow you to understand how complex systems work, they allow you to optimize those systems, they allow you to sort of delve back into the past. They let you predict the future. They let you understand what the implications are of the interventions you want to make on the physical world, whether they're the intended or the unintended consequences, so incredibly important in many areas, including when you want to innovate and model new technologies before they exist. So if you're inventing a new kind of robot or a new kind of smart machine, if you can do the modeling in silico, as it's called, ie in a digital model, it's much cheaper and faster and less risky than building prototypes in the physical world. But eventually, once you've sort of hammered out the design and you're confident that it's going to work in this digital model, then you start building the physical prototypes and putting them through their paces. And that's when you start moving on to things like living labs as being environments in which you can do that testing. So there is a whole ecosystem, if you like, that supports this blend of technologies.
Speaker 2:But the really exciting alchemy happens when you start cooking with all of them together. And in fact, coming on to the subject of food, you know my view on transformative innovation and invention is that it's very much like great cooking. You know you have to think about the sort of the utensils. So the technologies, the building blocks, the talent and so forth. You have to think about the utensils, so the tools, the frameworks that allow you to assemble and manipulate those raw ingredients. But the recipes are incredibly important. So the culture, the ways of working, the creativity, the ways of solving problems, so that it's that blend of the ingredients, the utensils and the recipes, just as it is in cooking, that makes it transformative and you need if you really want to get the most out of it, you need to work on all three of those together. So it's a sort of a it's an amalgam.
Speaker 1:And when you sit with somebody like Henry Dimbleby or with, maybe looking at a farm or at a food operator, of course coming from the extreme high tech side of the food, like moving the food as most efficiently as possible to our plates and the end consumer what do you see when you imagine I don't know how much time you spend on farms, but is it I'm not saying laughable or you would imagine what we could do here, or does it get very interesting for you? Or are you thinking, oh my, is also this really seems like we're in 1950 in some cases. Or like what do you feel or sense when you're interacting with, with the food and agriculture system? Uh, if that happens a lot. Or uh, if you imagine like an average farm in the uk or an average farm on in your area so my, my first journey into agriculture and farming was long before any of what I've just talked about happened.
Speaker 2:My father, who had been a corporate lawyer all his life, retired and suddenly announced because he'd been very much involved in the kind of the birth of the Green Party in the UK and was hugely interested in the concept of organic food and self-sufficiency, which was a kind of a concept that had come out then, and a famous book called Of the Same Name by John Seymour, which he and my mother read and they were captivated by that.
Speaker 2:And I was just about to go off to university and they said, oh, we've decided we're going to go and buy, you know, a farm in Wales and that's what they did. So just when most people, you know, retiring at the age of 70, you know, might have put their feet up and maybe done a bit of gardening, you know, on the allotment, my father and my mother and indeed all of us in different ways, headed off, you know, to this hill farm in wales and it was an extraordinary eye-opener about just how challenging farming is, especially with animals and the you know it's a 24 by 7, you know job and in all weathers and with all the unpredictability that weather and conditions bring, and I think the Even though we have been getting better at that, but still We've been getting better but I can tell you, on a hill farm in Wales it's pretty rugged.
Speaker 2:You do get some sunshine, but it's snow, wind and hail is all too common at that altitude and you end up. It's very challenging, but it was also an amazing sandpit for creativity. So I used to spend my time building machines, you know, with my eldest brother, you know welding stuff together, making crop sprayers and making new machines for doing things in the farmer. But we were also doing things like building a biogas generator to turn, you know, the pig shit into gas and anyway. So it was a sandpit, you know, for starting to explore different ways to, in a kind of a very kind of simplistic way, automate aspects of that but also and unoptimize it, although we weren't building, you know, digital twins of it, but the. So that was my first taste, if you like, and I came away from that and it's still. I look back on that as some of the sort of the four or five most precious years of my life, you know, in terms of what I was able to experience there, but also with a huge degree of admiration, if you like, for people who do that, you know, day in, day out, you know, for the whole of their lives. So then obviously got into food in a different way. But then when Henry, during COVID, asked me to join the advisory board of the National Food Strategy, particularly in terms of offering kind of technology advice, you know, I jumped at it because I just thought this is it's such an important. It's important at the best of times, but it was incredibly important, you know, during the pandemic, so but also not just the food system as it is, but the fact that we need to be, you know, finding ways for it to be more sustainable, both for us as humans in terms of healthy eating, but also in terms of the impact on the planet. And obviously, if anybody listening hasn't read the National Food Strategy, I would commend it to them. It's amazing. But one of the things in there, which I think was chapter two, was about systems thinking and was about and one of the recommendations was around, if you like the need to create a data ecosystem for the food system. And perhaps come back to the second one. But the first one is important because systems thinking was incredibly important in terms of Ocado, what we did there, because you need to optimize it in an end-to-end system, but it's also incredibly important in all sorts of systems.
Speaker 2:The food system is a very complex system of systems, but it sits within a wider set of systems, whether that's energy and transport, and logistics and healthcare and so on, so an international trade. So our everyday lives are like an amalgam of many different kinds of systems and if we want to understand them better, data and modelling are absolutely key. You know how we collect the data that we need to build the models, but then you know building the models and then using those to answer questions, but also to do things like optimising the behaviour of those smart machines we talked about earlier and then coming back to your question. So when I look now at agriculture, or indeed many other systems from you know, transport to energy, to defense, to health care, you know there are extraordinary opportunities, not just for AI, but for smart machines and for modeling, and for remote sensing and the internet of things, and for living labs, you know, and so that's part of one of the reasons why one of the first things, one of the organizations that I, a government organization, I'm still involved in it's called the Robotics Growth Partnership, which advises government on smart machine and robotic technologies, and back in 2022, we authored something called the Cyber Physical Infrastructure Vision, which is really about how we can use that blend of technologies that we've been talking about, but at a national scale.
Speaker 2:Or, as I like to put it, how do we build a better Lego set for the UK or, ultimately, for the planet? Because we need to change the way in which we build things, whether that be infrastructure or public services or products all kinds of products and services because we need to do it in ways that are more sustainable, more efficient and more affordable. So that's one of the things that's really excited me, you know, in recent years, is that idea of how we use all of these different kind of technologies, but at a national scale and, ultimately, a planetary scale. And that includes so, if we want to zoom back to agriculture, the opportunity to use automation and optimization to improve yields, to improve sustainability and, if you like, find new ways in which to turn photons from the sun into calories which is the game that we're really playing, you know, when we grow food but to find new ways of doing that that are not only healthier in terms of the products but also kinder to the environment, and one of the challenges we face with that is when you want to compare conventional agriculture with alternative forms of agriculture like vertical farming and algae and lots of other things like that, you know the problem is it's not a like for like, and conventional agriculture, as many of your listeners will know, you know has a lot of hidden externalities.
Speaker 2:So you know whether it's pollution or soil erosion or effects on flooding, or chemicals, pesticides you know all of those things have an impact, if you like, on the environment.
Speaker 2:But when you are working out, if you like, the net cost, if you like, not just to the consumer but just society, they often don't get taken into account and therefore we need to find ways to understand those externalities and once again, the National Food Strategy talked about that in detail. And one of the ways we can do that is through modeling. You know we can produce, you know, potentially digital models of a theoretical, if you like, farm or field or whatever it is part of the agriculture, and model those externalities and understand, or start making a step towards understanding, what their true cost is, if you like, in a whole of life sense, so that when you buy an apple yes, you're buying an apple, for you know whatever it costs, but what does it cost the planet to grow that apple, and what would that be like if it was grown in a different way? And obviously the whole push towards sustainable agriculture is trying to get to somewhere in between whereby we can have the benefits of conventional agriculture but without some of the current negative unintended consequences.
Speaker 1:And because it feels like there's such a powerful tool set like these three families and within them, um like an immensely powerful tool set, um readily available or coming available or in many cases um. But it also feels it's mostly ignored by food and ag. I mean, we see a lot of investments going into ag tech or we saw and that's cooling down a bit, um, robotics have a name but it always feels very superficial in general, like it doesn't feel part of a much bigger vision and and it sort of seems I don't know why. That is why, why food and agriculture um hasn't embraced or hasn't at least interacted with like the simplest forms of just modeling one or two farms and seeing what are the externalities, what are the net negative and the net positives that it's producing, et cetera. Like, how do we, if we would change this in management, what would happen without changing it?
Speaker 1:One of the many issues with farming is you have a harvest every year, or even less if you do a perennial, and just the iteration cycle is so long and so slow that it's very, very difficult to imagine or to have a culture of innovation. A lot of this suite of technologies could fix part of that, because you could iterate very quickly or you could model and understand what potentially will happen, and yet I don't see people talking about, okay, what would, uh, an agriculture system in the uk. I mean, you've done some of that, I think, with the food, but like, what could it look like? Or what somebody on the podcast just a few hours ago said? We lack the imagination, like we cannot even imagine what a food system looks like, and we just stuck in.
Speaker 1:Okay, this just gets worse over time. Let's just put a few band-aids left and right. What do you feel? What's the blockage there? Is it the mindset? Is it age of farmers? Is it we just don't care too much about the countryside Because the tools are there and seem to be there and only getting better every minute?
Speaker 2:So it's a lot of different things. So the first thing I would say is look, the technology is there but in many cases needs to mature further. You know, smart machines are getting smarter, they're getting more capable, you know, particularly on the back of AI, not just in terms of the intelligence within them, but the power of AI in the design process, if you like, of the smart machine in the design process, if you like of the smart machine, Really, even before they're built, even before they're built.
Speaker 2:So in the kind of stage of what you would call computational design, the ability for AIs to come up with new solutions, if you like, for making smart machines smarter and better, often by connecting AIs to simulation models, because when you put those two together, all sorts of things happen, because the simulation allows the AI to explore the real world, the physical. It understands things like the laws of physics and how things move, and gravity and the constraints that come with that. So it's really important that you do that.
Speaker 2:It's good to know gravity, I'd always say yeah gravity is good, you know it's kind of difficult, it wastes a lot of energy when you fight it, so, as poor old drones know, but you know it's the point is you need that blend of not just AI going off, you know, generative AI coming up with all sorts of weird and wonderful solutions which actually wouldn't work in physical, but when you put it connected and this is what has been happening in, for instance, in architecture for some time I mean, you've seen extraordinary buildings that have been built where you know the design often has come from an AI, but connected to an engineering model that actually puts constraints, if you like, on the creative juices of the AI to make sure that it is structurally, you know, has structural integrity and could be built and could be maintained and all the other things, so that it's that combination.
Speaker 2:Well, so there is work to be done in making smart machines smarter and more capable, and that's something that I and others are very much involved in, and we're about to publish the latest version of a strategy for smart machines that sort of takes us out to 2035.
Speaker 2:So it's kind of a work in progress.
Speaker 2:But that's part of it is the evolution of smart machines, but the other part is, as other podcasts were saying, it is the sense of the possible. It is our ability to imagine futures, if you like, where these technologies have been deployed and what would it look like, what would we want from it, what problems could it solve, what problems might it create? And for those of us who live and breathe these technologies, it's easier to do some of that, but none of us can do it perfectly. But for people who don't use these technologies, it's much more difficult, and that's why we need to find ways to share use cases, to share videos, to share models of it, so that people can immerse themselves in those kind of futures and get an idea of, oh look what they did. You know over here on this farm or in this, you know industry or whatever, and how might my business different or similar to that? And once again so modelling has a role to play in helping explore, helping people understand and have that sense of the possible.
Speaker 1:And I often feel we miss that, like if we look at a landscape we look at, most landscapes are very degraded and are not full of life, definitely, but somehow, because maybe we've lived there for a long time or we're used to the landscape without a lot of trees or without a lot of insects, or somehow over time it slowly went down and, like our baseline just has been slipping away and we're like this is normal and somehow that it's not normal, or at least not how it could be in 10 years or 15 years or whatever cycles we are going through.
Speaker 1:And so I think there's, apart from managing what we currently have or managing it way smarter, there's a huge potential of okay, what could it look like? What are the different scenarios, what are the different options with climate modeling, with what, what species actually could live here in 10 years, 15 years, and what management? We need to manage complex systems because a farm we've made them simpler over time, but we probably need to complexify them over time, which brings a lot of challenges and we have a smaller workforce and we need more machinery and smart machinery to do that. Like I don't think many people can even imagine what a farm looks like in 15 years. We sort of think, okay, maybe a smarter tractor and that's sort of it, but of course it's not going to look like that well, things like autonomous tractors and and you know, but it doesn't have to be microdosing.
Speaker 1:It could be a much smaller one.
Speaker 2:Well, yes, but it's. I was going to say that, yes, it's about you know the machinery to work on conventional farms and fields and so forth. The sad reality is that climate change means that areas used to be farmed in that way, you know, many of them are becoming infertile. And it's not about you know, putting an autonomous tractor on there. Actually, just the soil quality and the amount of water that you can bring, the kind of crops, if you like, you'd need that would withstand the increased heat. All of that means that we're going to have to find new ways also to grow things in areas you know would be called extreme environments, if you like, or else they're just going to become deserts. So I think there's a whole area of exploration, you know, for this kind of technology. That is about how do you create ways of growing food in controlled conditions but in a landscape that is perhaps now no longer amenable to conventional agriculture, you know. So you know, and people have been starting to do that. You know using solar power, and you know, effectively, huge greenhouses in, you know, deserts and places like that, and it's you know. And how do you recycle the water, and so you know. That is a whole branch of exploration, but that's going to be very important. So it's not just about, you know, bringing technology to bear. You know, on farming as we knew it, it's also going to be about that.
Speaker 2:But then there's another strand, which is back to the whole thing about how do we turn photons into calories in other ways, which somebody called Ian Boyd, who was also mentioned by Henry, I think, on that podcast that he did a good friend of ours has done a huge amount of thinking in this area and talks very eloquently about it, and we are going to have to find alternative ways to do that. You know, which obviously some people have been exploring with lab based meat and other people have been exploring with. You know, can we turn things like algae, you know, into palatable food? You know there are many different ways, and other people are trying to grow stuff in bioreactors, particularly because some of those technologies are going to be important if we want to to grow stuff in bioreactors, particularly because some of those technologies are going to be important if we want to grow food, you know, on other planets, whether that's the moon or Mars or whatever.
Speaker 2:So it isn't just about the Earth. Now, you know, people are thinking very much ahead to ways to do that kind of food creation, if you like, in very hostile conditions. So it's a multi. What I'm trying to say is it's a multi-pronged or multi-strategy approach to how we might solve this very complex problem of sustainable, healthy food production, and to do that for the whole world, you know. So we've got to make sure that the ways in which we would do it don't are sustainable and don't just increase the problems that we have that have led to climate change and to ask a few questions.
Speaker 1:We we always like to ask what would be your main message to the investment world, the financial world either people investing their own money or other people's money about this, um, these technologies, or this family of of different technologies, um, but also the work you've been doing. What would be? Um, of course they, they, let's say we do this on stage and they listen to this, but they also forget, and people don't don't remember a lot of things. If there's one seed, one thing you want them to remember and to really guide their work through, let's say what would be the one thing you want them to take away?
Speaker 2:Well, I'm not sure I'm going to stick to one, but I'll stick to a few. But I think the first thing I would say is that, in my experience, everybody focuses well, not everybody, but many people focus on what I would call the outcomes, the verticals, you know, whether it's technologies or solutions or sectors or whatever, and of course, those are incredibly important because that's where the results come. But I'm much more interested in what I would call the horizontals, the enabling technologies, the enabling solutions, the things that let you scale, that let you be more efficient, that let you optimize, and we've been talking about some of those. So the modeling, you know, the AI, the smart machines they're all about things like scalability and efficiency and reproducibility and resilience. And efficiency and reproducibility and resilience, and that blend of horizontal technologies that cuts across and supports the verticals often doesn't get enough investment and it doesn't get enough interest, I would say, you know, from governments. And there's an S at the end of that, you know, because it's a recurrent problem, because it takes it's like slow cooking, it takes longer often to build those horizontals and everybody wants the fast food, the kind of the outcomes you know, and I think that's got to change because we do need to build things better.
Speaker 2:Wherever you look, you know whether it's services, whether it's products, we need to build better. We need to build for maintainability. We need to build for recycling and reuse. We need to change the materials that we use so that they are more sustainable, and so, once again, it's back to that systems approach. We need to look at it from a systems approach, but we need to change the Lego. We need to change the recipes that we employ in how we go about building what we call the built environment, which is basically everything that is man-made person-made, if you like, so it's everything that isn't the natural environment.
Speaker 1:And even a big chunk of that is man-made or managed.
Speaker 2:Well, exactly, or man-affected, you know, as in you know. So the natural environment and the built environment have, you know, crossed over in, often in not such positive ways. So the first thing I would say is focus on the horizontals. But also one of the things about the horizontals is they lead to intersectionality. They often cut across. They're the things that connect one industry maybe with another. So solutions in one industry or one sector that you can then steal and adapt to solve problems in another and I'm an intersectionality nut I think it's incredibly important.
Speaker 2:I think, you know, if we stay, you know I used to say to people when I was at Ocado you know, working for me, don't go out and hang out with people and build networks with people who work in online grocery. Go off and talk to people who build lawnmowers or nuclear reactors or whatever, because you know what you may learn in those other areas. You know you may then be able to bring back and it may be incredibly powerful, but it's also likely to be a source of novel thought and creation. Because, you know, if you're in an echo chamber of people doing the same things as you, it's probably just of people doing the same things as you, it's probably just the same. Old things will go round and round. So intersectionality is very important and it's very important in innovation and invention. So it's very important to help tackle some of this growing list of exponential challenges that we've made for ourselves as a species.
Speaker 2:So we need to get as many novel ideas into the solutions as possible, and those horizontal technologies are part of the kind of the web, if you like, that connects all those different verticals in the different areas.
Speaker 2:So that's the next thing I would say is invest in the kind of the tools that let us do things faster and more scalable, but also let us join up the solutions from different so that we're not building everything from scratch. Because obviously, the more that we can build shared solutions across different systems, you know that's good for everybody. It takes costs out, it means we don't have all sorts of different spares, it means we don't, it means we have greater interoperability. But the answer is you know you need to design for that, you need to think about that beforehand and therefore, the more in which there are standards and the kinds of standards that actually we take for granted in the software world. So we take for granted in the software world. So open source and cloud has transformed how we do software. Well, we need the equivalent for hardware. We need that same sort of those same kind of shared solutions for hardware.
Speaker 1:And what would you do if you would be on the other side of the table as an investor? Let's say you wake up tomorrow morning and you are in charge of a significant amount of money. We usually say, let's say, a billion, in this case, pound sterling. I'm not looking for dollar amounts or exact amounts. I'm looking for what are big buckets, what are things you would focus on for a very long term potential investment strategy. But if you had to put that kind of money to work which is a bit of a weird question, I understand, but it's just. I like to ask it because when resources are almost unconstrained, like what do people prioritize and what do they focus on?
Speaker 2:Well, I would definitely focus on exactly what I've just been saying, which is those enabling technologies and solutions to accelerate many different kinds of outcomes. I like to think of them as wormholes. They allow you to perhaps out-compete other players who are doing things in a traditional way, If you can find a wormhole that acts as a shortcut to that. The most powerful one I've ever come across is modeling, but you know and in conjunction with AI now, as we've been talking about so I think that would definitely affect my investment. Things that make you go faster and accelerate are incredibly powerful. Secondly and this has very much affected, if you like, the portfolio of startups I work with yes, climate technology in all its shapes and forms is important.
Speaker 2:One of the challenges is that a lot of typical investment sources don't really like to build or invest in hardware.
Speaker 2:They prefer digital because it's cheap to build and to distribute and it's quicker to build and less risky.
Speaker 2:But, unfortunately, most of the things that we need to deal with the challenges in the physical world that we've created, which is not just climate change, it's how do we have affordable and sustainable health care around the world, how do we feed everybody around the world and how do we have better resilience to future pandemics? Because we're definitely going to have them. So, therefore, that's all going to involve physical interventions in our physical world, and so we're going to need hardware, and so we've got to get better at building it, like we've been talking about, but we've also got to encourage new forms of investment in it, Because if we want these technologies and there are some very exciting climate technologies emerging we're going to have to find a way to fund not only their initial research but their scale up and their adoption you know, and that is often where they hit trouble, because and that is where nation states, I think, need to step in, you know and sovereign wealth funds and so forth because it's no longer a challenge, I would argue for traditional investment.
Speaker 2:This is the future of our species and this planet that we're playing with here and therefore, you know, there is really nothing more important, you know, that we could invest in, than the solutions to these kinds of exponential challenges, and we need to get really serious about that and I'm not sure.
Speaker 2:Yeah, if a billion would be enough to unlock, a billion is definitely not enough there enough, but maybe what one can start doing with that is is trying to nurture, you know, some of those really key emergent climate technologies and but also maybe do more of the modeling that allows you to demonstrate, if you like, what those different solutions be. That's kind of what you know. An area that's always interested me is the whole idea of building Earth twins. How do we build digital twins at a planetary scale, of both our man-made systems and our natural environment? How do we collect the data scalably from, you know, on land, in and under the sea, in the air? How do we use space observation and satellites to do that? But building, we need all that to feed the earth twins. But that's something that a growing number of people are starting to talk about and there are some interesting projects there. But it's critically important because we've left it so late to intervene.
Speaker 2:We can't afford to get it wrong, and I think the last thing I would say is perhaps that, in my experience, most innovation, as it's called and I don't really like the word innovation, because I'm much more interested in invention which is more radical but because innovation tends to be incremental most innovation is focused on playing what I would call the current games a little bit better, and you know, an example of that might be the fact that the fascination at the moment with humanoid robots which you know are incredibly complex, costly machines, basically to try and mimic what humans do using human tools and human processes, with all the limitations that come with those, and that's about finding a better way to play the current game of human-created products and services.
Speaker 2:I think the future lies in finding new, completely new. We need to change the game. We need to, as I've been talking about in this podcast, radically change the way that we build solutions, and because we need to do it faster, but also because we need to do it more sustainably, but to do it with a systems approach. So don't just find ways to automate, if you like, what we've done before, but think about it, you know, from scratch and think about it at a systems level. So, once again, I would want to bring that to everything in that innovation portfolio thinking, you know, in terms of systems rather than just a set of individual investments.
Speaker 1:It's a great point to wrap up, and I want to be conscious of your time, but still ask one final question on the magic wand, which we always ask. But if you can change one thing overnight, what would that be?
Speaker 2:I love that film, the Martian with Matt Damon, you know as the astronaut stranded on Mars.
Speaker 2:The inventor and trying to invent new ways to grow food and survive until he can be rescued. And I think at one point in that film he says an iconic phrase. He says I'm going to have to science the shit out of this, you know. And that's always stuck with me and I feel, for the reasons I've just been talking about, you know we've left things so late, as I said, that we're going to have to model the shit out of everything you know. So I've made that point already.
Speaker 2:But it doesn't stop there, because one way of looking at how we run this planet is that it's, you know, it's like a bunch of squabbling tenants, you know, who haven't been looking after the building and who are now really rather surprised and alarmed that they're about to be evicted.
Speaker 2:The point is we've got to find ways to manage this extraordinary planet as a sustainable, connected ecosystem, you know, and the first step there is modeling.
Speaker 2:But it doesn't stop there and this is going to bring us full circle to digital twins, because, as I explained at the start, a true digital twin doesn't just model the physical twin, it helps to optimize and control it and the complexity of the world that we live in, you know, and we've seen this with. You know biodiversity and the unintended consequences of the changes we make, and often we intervene with one part of a food chain and it has, you know, disastrous effects elsewhere. Well, we need to turn that building going back to that analogy of the squabbling tenants into a sustainable, managed ecosystem, and I think technology can really help us there. So the magic wand, you know, ultimately is how do we take that step from modelling through to actually managing our planet as this connected ecosystem, sustainably and for the long term, and with a much greater level of understanding and, ultimately, compassion and humility than we have brought to bear in the past.
Speaker 1:I think that's a perfect way to wrap up this conversation. I want to thank you so much, paul, for coming on here, um, and for sharing your, your journey, your, your lessons learned, what you're working on now, the potential, the opportunities and, and definitely um, bringing a whole suite of tools into our mindset that often are. We might mention something, but we don't really mean digital twins. We might mention something on AI, and some things currently get the spotlight, let's say, but it's definitely not the full suite of things that we have to work with. I would even say I don't think we have the luxury to, as you said, to not model and to not understand, because it is 15 seconds to 12 or even worse than that, and so we've been walking in the dark for long enough to, and we have these powerful tools, and I don't think we have the luxury to not put them to work, put them to good work, with all the complexities that come with it. Like, modeling Earth is not going to be easy for any stretch of the imagination, but we definitely need to give it a try. So thank you for coming here and share about your work, your journey and what it means for the food and agriculture space. Thank you very much. Thank you so much for listening all the way to the end.
Speaker 1:For the show notes and links we discussed in this episode, check out our website investinginregenerativeagriculturecom. Forward slash posts. If you liked this episode, why not share it with a friend or give us a rating on Apple Podcasts? That really helps. Thanks again and see you next time.
