Episode transcript: Is the global food crisis a problem that only tech can solve

DAVID ELLIMAN

Hello and welcome to Tech Tomorrow. I’m David Elliman, Chief of Software Engineering at Zühlke. Each episode we tackle a big question to help you make sense of the fast-changing world of emerging tech.

Today, I’m joined by Illtud Dunsford, CEO and Co-founder of Cellular Agriculture Ltd. Founded in 2016, the company is developing hollow-fibre membrane technology to produce cells for cultivated, or lab-grown, meat. Illtud isn’t only a business owner; he’s also a farmer from a family with generations of experience in raising cattle for meat.

So who better to help us consider today’s question: Is the global food crisis a problem that only tech can solve?

So if the subject of today is ‘can technology solve the global food crisis?’... I mean, you have such a unique position coming from the traditional side of farming and then taking the decision to apply new technology, which is obviously potentially an extension to that industry, but also a challenge to that industry.

So, I mean that leads us really nicely into telling us in simple terms, what are cultivated foods? What is cultivated meat?  

ILLTUD DUNSFORD

It's the process of taking a cell that's come from an animal, so from traditional lifecycle like a cow. And rather than feeding the animal, traditional forage like you would in the UK, you feed a series of nutrients that are very similar to that cell.

And that cell then multiplies in no different that it would within the animal. And that collection of cells, once it's grown to a certain amount, can then be processed in the same way that meat is traditionally processed to make a food product that is as nutritionally equal as what traditional meat would be.

And similarly, you know, a cultivated food product could be multiple things. It could be meat, it could be fat, it could be seafood. It could also be plant cells.

With the effects of climate change, and there are going to be real pressures on certain areas that produce things like coffee or chocolate, where we won't be able to meet a growing demand for traditional products in the way that we currently grow them.

DAVID ELLIMAN

So in a way it feels like what you are suggesting is that there is a potential process that's available that I'm guessing in time could be optimised so that you can do things, I'm guessing on a smaller footprint, more cheaply, etc., to take away maybe some of the risks of doing it with some of the side effects of doing that kind of farming in nature.

ILLTUD DUNSFORD

Absolutely. If we think of, I don't think anybody believes cultivated foods will replace animal farming. It's something that will be complementary. But in growing a whole animal, you are growing a number of components that we as humans don't consume. So there's a limit to the use of that carcass.

And things like blood and fat, skin and bones don't generally enter into our food chain. People generally kind of understand that there is an impact from cattle, from the methane that they produce through the burps that they have. And that contributes then itself to climate change. And so it's finding a system that, or finding a process that will get you exactly the same output but in a more efficient way and with less impact on the environment.

But we have to also remember that although the product is sort of the same as what it comes from traditional agriculture, you're also reliant on traditional agriculture because the inputs to feed the cells generally come from the agricultural industry. So it's not, as far as I'm concerned, to eradicate farming as we know it.

It's to see how can we produce the same output, but in a more efficient way.

DAVID ELLIMAN

I think sometimes we live in a world where the edges, the extremes are taken too quickly and too literally, and often the truth is within the nuance that lives between all those things.  

ILLTUD DUNSFORD

Absolutely. But I also think that sometimes, you know, the extremities are the things that make you think differently.

The extremities are the things that make you push forward, but more often than not, where you end up being is something that sits quite a bit further back from that, what becomes commonplace, essentially But it's really important to have those extremities and for people to be thinking in that way, to push a narrative in a certain direction for us to think differently and to do differently

DAVID ELLIMAN

And to give you the bounds for your problem.

ILLTUD DUNSFORD

Yeah.

DAVID ELLIMAN

So in software engineering we see similar problems, where we're having to push the boundaries of what's possible. DeepMind aimed to solve intelligence, but funded it by selling to Google and building practical AI applications along the way. The key is what venture capitalists call ‘climbing the ladder’: ambitious vision, broken into achievable stages, and we don't always appreciate how long those stages take to actually achieve.

Waymo, Google self-driving project, started in 2009. They didn't try to solve everything at once. First highway driving city streets and then specific geographic areas, and then 15 years later, here we are, they're finally achieving some sort of commercial deployment, but it's still very limited. As you would expect in something like that.

Ambitions, visions attract exceptional talent. They also attract investment and a lot of hype, but balance matters. For example, Magic Leap raised three and a half billion dollars on hype, but they didn't deliver anything that matched that. The sweet spot is ambitious, but achievable goals that others haven't recognised yet. Think, evolution and not revolution.

So, let's turn now to the technology that Illtud’s company is developing. Where did it all begin and how does it work?

ILLTUD DUNSFORD

I was very fortunate at a conference in 2015 in Maastricht in the Netherlands to meet a group of academics and one of which, Professor Marianne Ellis, who was at that time the head of chemical engineering at the University of Bath.

All her work had been, whether in the biomedical field or in food, specifically on bioreactors, you know, the technology that grows these cells. The research that she'd done during her PhD was on technology called hollow fibre membrane. It essentially replicates human or livestock vascular systems.

So they are small membranes. Little... Imagine sort of middle straws. They transfer the nutrients directly to cells in the same way that we grow tissue within our own bodies or the way that cattle grow tissue. So it limits the stresses that the cells experience when they grow, the nutrients are targeted directly towards them so that they can consume them healthily, and any waste that those cells have are able to be taken away quite simply and naturally in the same way that we would grow muscle within the human body. And so to me, as a non-scientist and somebody who'd come from agriculture, that made perfect sense that this technology sounded far more appropriate than essentially large stainless-steel vats where cells would grow.

The same kind of membrane technology has been used for years to filter water. So there are already factories that have these kind of membrane systems at scale to produce an at-scale process. So although there's major challenges to reappropriate it for another purpose, it did seem logical to me as a business person that you could actually take this kind of technology and build it at scale and cheaply.

DAVID ELLIMAN

That's fascinating. I mean, obviously you are pulling together some other industry experiences in terms of how this particular technology's been employed in those contexts, and for something like this, I should imagine that you've got to have a number of units that are scalable in terms of growing the cultured meat.

ILLTUD DUNSFORD

Yeah, absolutely. So, if we go back to 2015, when we met and discussed about starting the business, if we looked at hollow fibre membranes that had been used for forms of cell culture, there were reactors out there that were being used for that purpose. But generally, it wasn't the cell mass itself that was being grown as a product.

It was the excretion, you know, whatever that cell produced. So, whether it was a protein or whether it was something that would end up in a pharmaceutical product. The output wasn't the cell; it was the product that came from that. So, in the early days, we even had to demonstrate and create a proof of concept at the very smallest of scales, I think our first reactor was about 60 millimeters long. It was a tiny little thing, but we needed to prove that we were able to grow cells.

The challenge over the last 10 years really has been how to scale that to that industrial scale. We had to design reactors that were going through those iterations or scaling points, so that we'd have a large enough population to be able to inoculate a large bioreactor at an industrial scale.

And that's what we've done over that period, you know, we've gone up in stages from one to the next. And I mean, like you mentioned, it's a scale out. It's a modular approach. So rather than trying to make huge, massive stainless-steel vats, there's a terminal size that actually is the one that would work industrially. There's no need to go over a certain point.

To our knowledge, the best of our knowledge, we built the largest or the longest ever hollow fibre membrane ever for any form of cell cultivation application. But essentially we've also realised we don't need to build that at that scale industrially because it works just as well as smaller scales.

DAVID ELLIMAN

That's fascinating. So, I mean, where are we now in terms of the industry? You started this in 2016 and that's coming up to, well, nine years. It sounds like you've made amazing progress within that period. Where does this sit in terms of the future for you, but also the sort of footprint within the industry as a wider point?

ILLTUD DUNSFORD

At the moment, we've always focused on fundamental problems that the industry need to get to industrial scale, and hence why you have nine and a bit years down the line and we're still pre-commercial. So we have another nine months or so before we'll get to the point of continuing that, coming to an end of the first round of R&D that we have in terms of that industrial scale.

But once we've been working on technology that will enable scale in the future, others in the industry have been focused on getting to markets and that's what's really driven the appetite, especially from investors, and probably even the public in terms of considering cultivated foods as a new food product, and you know, that we've seen, especially in the last 18 months, a raft of companies gaining regulatory approval in varying territories globally and seeing products reach market.

And that's both in human food and in pet food. And the things that we really should be celebrating more so than we do, because if I'm perfectly honest, you know, we saw the first product on the market in 2020. Had I ever imagined that would've been possible back in 2015, I don't think I would've ever imagined that.

So there's been immense groundwork done by a number of pioneering companies to be able to get that initial market access. We're still many years away from getting to scale, but that's the challenge that we always focus on because I came to this thinking this is a marvellous idea, but it'll remain novel until it can attain scale.

DAVID ELLIMAN

So when we look at the time it takes to perform R&D and get to market, there's a lot of parallels that we see in software engineering, for example. Google acquired DeepMind for half a billion dollars in 2014 when most saw AI then as science fiction.

Those patient bets during the winter positioned them to dominate when transformer models triggered today's AI boom. True breakthroughs can't be rushed. They need the right confluence of technology, infrastructure, and market readiness.

We've spent some time exploring what the cultivated meat industry is all about, but let's turn to the why. Why does Illtud see this work as so important?

ILLTUD DUNSFORD

If we look at how we produce food currently? We exceed a number of the existing planetary boundaries that we have. Some of that is in greenhouse gas emissions, some of it is in biodiversity. There's a number of areas where we're already exceeding. And we actually produce enough, it's just that we are wasteful in the process, and that's all the way from the field to the plate. The obvious way to do it is to make some changes in the way in which we do it, to consume less, to consume what we need rather than to consume in excess or to waste.

But that's difficult because that needs human change. And human change is hard, especially in an era where we have plenty, you know, especially in westernised civilisation. There's never been so much at our fingertips, whether it's technology or information or food or culture. We can access all these things, either at the click of a button or by going to our local shop and experiencing it.

So limiting choice is something that's very hard for us as humans when we're so used to it. That’s the reason why, for me personally, there need to be other ways that you can offer people the same choice they’ve historically had, but that that choice is better for the planet.

And I've always thought of cultivated foods where it's not one or the other. I would want people to be able to walk into a supermarket and they could buy a product that's made from traditional livestock, or they could buy a product that's made from cultivated foods and they could make their choice. Do they want something that has the same and having parity, you know, that maybe the price could be the same, maybe the flavour profile is exactly the same, the emotional experience of consuming both would be the same. You have the choice of if I'm buying one, I'm limiting my impact on the planet. And that's the only way that I can imagine that we can see that level of change in us as humans.

DAVID ELLIMAN

I guess we've also got the advancing climate change question that I guess forces a greater degree of focus on the availability of food.

I just wondered how your set of solutions can help with that, you know, when you're at scale, then not only can you supply the choice, but I'm presuming therefore you can start to say, well, yes, the choice could also be not just about flavour and availability, can be about the carbon footprint of the end-to-end process of this product.

ILLTUD DUNSFORD

So the starting point really for investment into the cultivated food industry was largely based on prospective lifecycle analysis for the technology and the output in its comparison to traditionally produced proteins. We'll take sort of cattle. It's usually used as the worst offender when it comes to considering these things.

There's a significant reduction in the resources that are needed. So it's just an excess of 50% the land use, 50% the water use. So you can produce the same quality product with the same level of nutrition with nearly half of the environmental impact. That's kind of like a no-brainer really in terms of being able to opt for alternative technology, but you still are reliant on agriculture because you still need those inputs.

So glucose, for instance, is the largest energy source. So you're not having to produce huge amounts of forage or huge amounts of arable crop for livestock. You're producing a significantly smaller amount that you can then utilise for feeding the cells. There's also the issue of, with climate change, there's a real effect in terms of with things like I mentioned, coffee and chocolate, with climate change, the banding around the world where we currently grow those products, that's really seeing the pressure. With climate change that changes the climate and the weather within those bandings and that just means there isn't enough land to produce at the levels that we currently produce.

So if we just keep on going, you know, there's nothing that we can do.

DAVID ELLIMAN

The formula breaks, doesn't it?  

ILLTUD DUNSFORD

Yeah, that's it.

DAVID ELLIMAN

There's sort of two things that occur to me now. One is that what's holding it back at the moment in terms of it being available at scale to the public, and if there's a lead time still, does your timetable stack up against that ever changing and ever shortening timescale that climate change brings us.

ILLTUD DUNSFORD

We could definitely match that timetable. The unfortunate thing is, is that new technologies, especially those that are kind of from venture businesses, that are reliant on investment in the first place in government support. And on hype cycles really for anything.

You know what, and we've seen in other technologies, if you look back at the renewables industry, you know, solar is now quite established. Electric cars are now established. But it takes a very long time to go from the beginning of an idea of an industry through to a place where these things are accessible to all.

And with that, that climate of how you produce those things, whether it's manufacturing, whether it's the way in which government supports it in terms of early stage R&D, or whether the way the government engages in terms of industry centres in manufacturing, all the way to regulation, all those things are important.

And actually the cultivated foods industry is a real oddity because more often than not, things start off in academic research. They gain a critical mass. Government then steps in to fund early-stage innovation, and it's at that point that venture capital then funds the next step when something can be commercialised. In this industry, it was a lot of money from Silicon Valley that came into the industry because an idea had been proved. Very little work had been done in academia. So we saw a boom in early companies that were probably too early-stage at that time to form an industry. And we're now seeing governments and especially UK government are doing particularly well in supporting early-stage research now and going back to academia to support research.

The issue now is, is how do we scale that quickly so that more consumers can actually access foods.

DAVID ELLIMAN

You've talked a lot about, you know, within the agricultural industry that solutions often rely on a number of different technologies being available, working together. Do you see any other technologies being brought to bear on the problem space that you are in to get it to market more quickly?

Do you have a feeling for like when we're gonna hit that inflection point?  

ILLTUD DUNSFORD

I'd be comfortable in saying that within the next five years, I would expect companies to be aggressively scaling their processes. Within the cultivated meat industry, we, for instance, are a company that's focused on bioreactors and the bioprocess, so the technology to grow the cells.

But there's a raft of enabling companies. There are some that are purely focused on the cells and the cell lines that you need as a starting population. There are some that are focused on the feedstocks that you need to feed the cells. There's others that are looking at scaffold materials that you can grow cells on to create some form of texture.

And then there are others that are on the downstream processing element that are looking at the final formulation of food products. So there are a raft of technologies within our industry where companies are focused on, but equally we're looking as an industry outward. We're looking at the traditional agricultural industry.

So for like the feedstocks, you know, is there an alternative that we can look for as an input that's cheaper from the agricultural industry? And it might not be from a primary agricultural product, it might come from a waste stream. Is there something we can utilise that could reduce that cost? So to make the process far more efficient.

So there's work being done at the moment to join the dots. How we find all those components that will essentially get us to that point of scale.

DAVID ELLIMAN

So when we consider the cultivated meat industry and how it's advanced quickly through deep collaboration across disciplines, we've seen this at the fundamental heart of agile software delivery. DevOps, for example, revolutionised how we build software. Before, developers threw code over the wall to operations.

Deployments were quarterly wars. DevOps just didn't merge roles. It created new disciplines. So, what does siloing look like in software engineering? It creates risks, and it can be a barrier to innovation. Modern siloing manifests as this throw over the wall anti-pattern. Data scientists build models without understanding the production constraints.

Security audits after development, instead of participating in design. Infrastructure teams provision resources without understanding the application needs. The risks are pretty severe, so how do we break these silos down and actually get the dots connected? The new opportunities and breakthroughs that are becoming possible in software engineering is the coming together of some of the ideas that have been promoted by the Agile community over the last few years.

We started by breaking things down into chunks, and then we looked at the flow and the rate of change, then building with multidisciplinary teams. But then it was more about trying to build platforms that become self-service for developers, ways to integrate templates that enable developers safely and securely to deliver into the target system, and platform engineering is a perfect example. Instead of developers waiting weeks for infrastructure through tickets, platform teams provide these self-service capabilities. Spotify's Golden Path, for example, offers pre-configured templates in encoding best practices.

Developers deploy services in minutes and not weeks. You get massive productivity increases and less errors because you are using these pre-configured templates. The pattern's clear, it's connection that creates that capability.

All the progress in the cultivated meat industry leads us to the question, what role could it play in the future of food?

ILLTUD DUNSFORD

So I think firstly, I'd have to consider what traditional agriculture might look like in 10 years, because the drive and the need filter in your products will come from the fact that traditional agriculture might not be able to meet the existing demand.

And if we think of the extreme weather events that we've had only in the last few years, projections are that those will be far more progressed within the next decade. And I think we've been slowly edging towards a cliff where we've thought that we've needed a load of new technologies, but actually, until we’re there, it's a nice to have rather than an essential.

I think that the thing that will trigger the cultivated food industry is that we'll become, not a nice to have, but an essential. And on that conversation of consumer acceptance, that's one that I've always considered is, is there a yuck factor currently for this idea of something that is produced differently or potentially yes.

But that's today. That's when we consider that, you know, we have plenty for those who have the privilege to have plenty. And so we don't need to make a choice.

DAVID ELLIMAN

The constraints are different.  

ILLTUD DUNSFORD

Yeah. It's a different, it'll be a different world in 10 years.

We will be limited in our choice to a certain degree because some things won't be able to be produced anymore.

And at that point, I think that really completely shifts the narrative in terms of how we think of food. And I think for me, in terms of considering what the next 10 years, depending on where population shifts, are we gonna be even looking at the same types of retailers in the likes of the UK and the large centralised models in which we produce food?

Or do we need to think differently? Do we need to look at decentralised models? Do we need to produce foods in a different way? Because technology allows us to do that in a way that agriculture often doesn't. There's a huge amount of opportunity there of how we could do something different rather than sticking to a way in which we currently do things.

DAVID ELLIMAN

If we only concentrated on technology, as you've said, there are a whole host of things that need to be done societally, culturally, and so forth. But what are the risks if we just... A bit like climate change. There seems to be a body of thought that says technology will solve the climate change problem, and you realize the risk in that assumption because it takes the emphasis away from trying to prevent things as opposed to hoping that something will come along and save the day.

And I guess there might be this same sort of thought process here, both in terms of sustainability and in terms of optimised sort of meat production. What could we miss? What would be the blind spots that we'd miss?  

ILLTUD DUNSFORD

The key factor in venture businesses essentially is the return on investment for early-stage investors, and that capitalistic model is sometimes very difficult when it comes to just pure sustainability.

Sustainability and the need for more food is the opportunity. But delivering at a price parity or even cheaper, that's a challenge in itself. So if we consider meat, for instance, that climate change has an effect in which we produce traditional meat, the way in which that will likely level playing fields is it'll make meat more expensive. And therefore the technology then is probably gonna be able to compete sooner than it would've if we were looking at technologies now.

That then has economic pressures on us as people generally, because the way that the capitalistic model has pushed us now is that actually it's access to all things is what we want. You know, food as a component of our weekly or annual spend is significantly less now as it was a hundred years ago.

The world is a very different place to then. It's whether we see a society that regresses a bit to that or not, because it's not just a devaluation monetarily of food. It's value to us as humans. That's happened over that last hundred years. It's very hard for us as humans to change.

DAVID ELLIMAN

Yeah, and I guess the drivers, I think as you've been saying that, if the problem was that we wanted to solve world hunger, as an example, then the political will for that would be very different and the political support structure, and therefore investment in research and deployment would be very different than VC funded return on investment, which, as you say, is probably more about choice to the already people who have access to the alternatives.

ILLTUD DUNSFORD

Yeah.  

DAVID ELLIMAN

The technology remains intriguing and capable of maybe more things than it is being currently geared towards.

ILLTUD DUNSFORD

I fully expect that we will see a huge increase of investment, whether it's from venture capital or whether it's from government in the future, because the need will become more apparent.

But the problem is, because climate change is such a slow thing, that need takes a very long time for it to be obvious.

DAVID ELLIMAN

And then it's almost too late.  

ILLTUD DUNSFORD

Yeah.  

DAVID ELLIMAN

So I just, I want to return you somewhat playfully back to the key question that we've set and just, you know, even though you've spent the entire interview talking about this, just any kind of summary thoughts, how you feel about, is the global food crisis a problem that only tech can solve?

ILLTUD DUNSFORD

Definitely not. I think humans can have a huge impact in solving this problem. It's whether we choose to or not. The fact that we have a growing world population means that we have many more minds to find the appropriate solutions. Humanity can do atrocious things, but we can also have some really amazing ideas and they don't have to be just to do a technology that can be very simple.

So, I still have faith in the human race to be able to solve these challenges for the future. But whether technology is part of that or not, that remains to be seen.

DAVID ELLIMAN

So, is the global food crisis a problem that only tech can solve? And Illtud’s answer was definitely not, and it resonates deeply in my world in software engineering too. Technology is a powerful amplifier, but it amplifies human decisions, both good and bad. GPS guides emergency responders and enables surveillance states.

Social media connects families and spreads misinformation. The same technology has opposite impacts depending on the human implementation. Today we produce enough food for 10 billion people, yet 8 billion people go hungry. Technology solved the production problem, but not the human problem. We've seen this pattern repeatedly in software, social media, algorithms optimised for engagement, created echo chambers.

Automation eliminated routine jobs faster than workers could retrain. The technology worked perfectly, but the human implementation was flawed. The global food crisis, like every crisis, will be solved by humans making choices, and they're often political choices. With technology, our increasingly powerful assistant is implementing those choices at scale.

Thanks for listening to Tech Tomorrow, brought to you by Zühlke. If you want to know more about what we do, you can find links to our website and more resources in this episode show notes. Until next time.