The Future of Food, AI + Food
5:30PM May 7, 2020
Good morning, good afternoon and good evening and welcome to episode seven of the AI for Good Webinar Series. We hope that you, your family, your friends, and your colleagues are all staying healthy and safe. My name is Fred Warner from the ICU, International Telecommunication Union in Geneva. And it's my privilege to introduce today's webinar. Now the EU is the United Nations specialized agency for information communication technologies. And we're also the organizers of the AI for Good Global Summit, hand in hand with XPrize Foundation, and in partnership with 36 UN agencies, ACM and Switzerland. And the goal of the AI Summit is to identify practical applications of AI to advance the Sustainable Development Goals, and scalable solutions for global impact. Now, like the most of the world, but summit has gone digital, and we're moving forward with weekly programming between now and the end of the year, allowing us to eat reach even more people than with a physical event. And today's webinar could be considered as part one of the AI and zero hunger breakthrough session that would have taken place in Geneva, had it not been for the virus. And before I introduce today's moderator, I'd like to go over a few housekeeping issues. First of all, we've disabled your microphone, so please use the chat and the q&a. function if you wish to communicate, and it's the responsibility of a moderator to identify and ask today's questions, to the panels to the panelists rather. And we're counting on your active participation to create a very interactive session today. So without further ado, I'd like to introduce today's moderator, Caroline Kelton. She's the XPrize. She's a project manager at XPrize. And she's leading their research and roadmaps on the future of food systems. So Caroline, without further ado, I'd like to welcome you and the floor is all yours.
Thank you for our urbanite today's title, the future of food, AI and food. And over the next hour, we will be digging into the various issues that currently exist around food systems and food security and discussing how AI is currently and will be shaping the food landscape of tomorrow. I'm joined today by a great panel. And we'd love to hand it over to them to introduce themselves and tell us where they're joining from. I'll start with Lauren.
Good morning. Good afternoon. Good evening, everyone. My name is Lorin Fries. I've focused on the future of food systems for many years and lead a strategic advising firm focused on the future food systems with respect to technology and innovation called future table. I'm joining you from California in the US. hand it over to Marin.
Good morning, afternoon, everyone, wherever you are, my name is Marin goals. I'm a circular economy scholar and activist. And I happen to work at Denon as the global director of open innovation and circular economy for food and I'm joining you today from the Netherlands.
Hi, everybody. My name is Bernhard Kowatch, I'm the head of the World Food programs innovation accelerator. We are based here in Munich in Germany. Hello, everybody.
Thank you so much and welcome everyone in the audience. I'd like to start this webinar with the Question on a topic on everyone's mind today, which is COVID-19 and link that to the future of food and food systems currently. And the question would be global food security projections are very grim, especially considering the impact of COVID-19. And I'd love to know from all of the panelists, what can we expect in the coming years and who will be the most affected by this pandemic around food security specifically? So,
yeah, if I can start, I think, for what food problems? perspective, we just launched a report. Actually, it's about one or two weeks ago, our executive director brief the Security Council on this, where we actually base an analysis on country level economic and food security indicators, which unfortunately forecasts that the number of acutely hungry people will increase from hundred To 35 million people to 265 million people. So there's a real risk of multiple famines and acute hunger that might happen. Now, how is COVID-19 actually influencing food security? Like the the issue here is not only about like trade restrictions, there's also a question about like ability to actually perform your duties. And especially if you think about developing, when people are earning a wage on a daily basis to sustain the lives of them in their own families, if you cut off that employment, that has severe effects on you and your families and the ones that you actually provide for now, this is in particular a challenge that we think that the acutely poor and more food insecure people are like kind of the people who are like daily in formal labor markets, specifically also like urban settings where maybe like they might be more severely cut off from lockdown, but also, it's often times, already affecting communities that already are affected by shocks, by conflicts, climate change, and other aspects that make the lives already right now in the livelihoods put them at risk in terms of like stretched food supply chains. Now, what we think about this is then, really, in the future of food in this country that we're seeing right now, this is shocked that exacerbates current problems. And so we need innovation like AI to actually do something and have impact for the people that we serve.
Let me compliment much of what Bernhard said and actually talk a little bit about the impacts that come from the ecosystems around these food insecure populations. And let me give three examples where I think there's a potential crisis but also some windows of opportunity. It may be helpful to point out that some of the consumers hardest hit will actually be an import dependent nations. If we look at the continent of Africa, you know, their their food import bill was projected to reach 110 billion under $10 billion By 2025 in a small island, developing states or other important dependent regions will be very hard hit. But I think and I imagine we'll get to this later on in the coal, this will interact actually with a almost a reverse shift in globalization towards more decentralized food systems. So there might be a real impact now, but also some opportunity as we look forward to readjust some some of that system. A second impact I wanted to talk about is on workers. And right now we are appropriately celebrating and recognizing the essential role of farm factory and food workers. And also at the same time, we're recognizing that any human interaction with the food value chain, at least in the context of a virus presents some risk. And so you see a lot of activity in the in the investment space, for instance, around robotics and automation. So the question of the future of work in this turbocharged digital age, I think it is a real open one Third and final set of people I want to talk about that really have to do with who is experiencing the impact of this crisis in hunger terms. That group has to do with the entrepreneurs who formed the SME, or the small and medium enterprise sector. So in places like Africa, these folks, these small businesses are actually the backbone of the economy, and those who actually provide the sorts of food and food related services to their communities that we so desperately need. So how to support those SMEs. And that kind of sector, I think, is a really important and imperative question for this moment.
So let me build on what Lorin just said, I totally recognize the point about entrepreneurs and I also, from my perspective, and my job, take the look at startups. And equally there are lots of risks. What will happen to the ecosystem at the end of this crisis, but I also see an opportunity because I actually believe that this crisis is making the weaknesses of our system, including the food system tangible. And in which actually find some hope as, even though, as was just described, the less powerful will be hit hardest, the powerful, those that so far felt untouchable will be impacted. Now, this unprecedented shock to the system is making the downsides of the linear system that we've created, very visible and even undeniable, even for the naysayers and the skeptics. And I actually take hope from that, that this might be the start of an acceleration in a system shift.
Thank you, everyone. Um, I think all of your answers touched on something very important, which is you can't really disconnect food from the system and the whole value chain when you think about the food industry. And Merijn, I'd like to start with you, and a question. something you've advocated for for a long time, which is a system level revolution and a shift towards a circular food economy. I'd appreciate if you can tell us more about that concept and that shift and whether it can help us prevent future crises, health and others. And also if innovation has a role in that shift and what that would be.
All right, let me take a stab at that jam packed question. But I love talking about it. If you look at our current food system, which I would call linear food system, it's built on a mechanistic world view, we see the world as a giant machine, which in my opinion, is the product of the Industrial Revolution. And this system is built on a premise of infinite growth, interestingly, within a finite world, and one of the key drivers for this growth and drivers of that system is efficiency. Now, what efficiency leads to is intensification. Which one could argue increases actually the breeding ground for viruses, like COVID. It's also specialization and the removal of all redundancies from the system. And I think that's key because the removal of this these redundancies leads to the fragile system, a system that is prone to shock, which is what we're seeing right now. Now, in contrast, we look at the circular economy, which I would define as a food system that has a positive economic, social and ecological impact by design. This system, in contrast, is grounded in complex adaptive systems paradigm, just like nature sounds complex, but it's basically just like nature. This system doesn't aim for infinite growth, but it's aimed at a dynamic balance, again, just like nature, and this system isn't driven by efficiency, which is trying to do more of the same with last, but it's actually driven by effectiveness, which is about doing the right Think and such systems. They build actually. And they create diversification, decentralization. And as such, what they create is resilience. Now, if we then look at the role of innovation, which was the second part of your question, I believe that innovation is crucial in system shift, as I just described, complex adaptive systems don't change in a linear planned fashion. They're not built on the very mechanistic notion of interim extrapolation. They change through emergence. And I therefore believe that innovation and experimentation are the very adaptive power of the system, just like procreation and mutation is the adaptive power of mankind. So I think that it definitely can have a positive impact as long as we make sure that we harness this power. Innovation and directed in the right way.
Thanks, man. With that, I'd love to turn to Lorin and explore with you some of the examples of how AI innovation has impacted and will impact the future food systems at different levels of the food value chain.
Now let me zoom people through the the food value chain all the way from farm to fork. So starting on farm, and I think a lot of you will be familiar with some of the applications of digital technology, including artificial intelligence at the farm level. And so often that's used to increase yield but it can also be used to optimize for other interests. So using sensors, drones, robotics, automation, etc, internet of things on farms, I think one of the things that we will see actually coming out of COVID-19 and picking up and accelerating forward in general is producing food elsewhere, including indoors, something called controlled environment, agriculture and using a lot of those same digital tools. Through farm management systems, I think that will be a very, very interesting outgrowth of this of this crisis and maybe something that can help us towards those more decentralized systems that were mentioned. As we move along the value chain of things, supply chain logistics and the efficiencies that can be gained through digitization and some of the the artificial intelligence applications will be absolutely key. What we see breaking now is largely as a result of those supply chain linkage failures or channel switching between you know, food that was going to restaurants or or hospitals and or hotels and is now going into retail. And so actually having very smart systems to be able to direct the food where it's needed to go will be key. And another opportunity along the supply chain is traceability. And I think that's becoming more and more important to consumers. As we get to the retail end of the value chain. We've already seen the e commerce fundamentally reshaping how people buy food from Alibaba and Amazon on down to marketplaces. And I think that there are a set of innovative marketplace spaces such as in Nigeria, for instance, that are doing a better job of linking supply and demand actually allowing to manage to prompt supply, reducing waste in the system, increasing economic opportunity for those involved. Finally, we get to the eaters, the consumers. And I think that one of the applications is where food and science meat such as for the human microbiome and meeting advanced tools, like artificial intelligence, maybe even super compute to support the complex research analytics around that space. Final that I'll say is, is around intelligence. And I think that one of the things that Bernard and his team and others who are looking at the macro picture around this this crisis really need is the intelligence around hunger shocks. And so whether it's satellite imagery to surveil landscape or it's looking at the interactions of markets to spot where there's going to be a food shortage, we tools like artificial intelligence, super compute power, again, to be able to bring us the sort of intelligence that can help us spot a crisis as far as we can in advance and mitigate the effects of it.
Thanks, Lorin. And that's a great segue to a question we have for Bernard about your work at the World Food Program through the innovation accelerator. And I'm very curious about how you are utilizing and promoting applications in your food security and resilience programming internationally.
This to provide some context as one food program where the world's leading agency fighting hunger, so it's both on the side of like, saving lives in emergencies like food or cash emerges, but also changing lives like providing supporting national system for school meals, Mother and Child Nutrition working with small farmers. So I think it's actually within that spectrum that we're working on the multitude of solutions like on the If you weren't on the front end before to actually identify where are hungry people in how hungry are they? One innovation that we've supported that's actually online gnosis called hunger map live, so you can go online hunger map.web.org. And what you can see there is a real time analytics platform. It also visualizes food security, hunger, and most recently also COVID-19 instance and also scrapes, news platforms, so that we have near real time reflection of what's actually happening on the ground. So it's linked up with databases, but also like scraping news platforms. And that's what's actually that we see a lot of innovation that regardless like you, you're quick in responding to emergencies, you're faster in getting the right food to the right people at the right time. And that can have a huge impact and you're not talking about like increasing your program efficiency by a percent by 2%. You might have 20 30% efficiency or effectiveness gains, which is really exciting. In the same vein, there is another innovation we've been supporting. It's called Plus, it's a school menu optimizer. And so if you think about it, like a lot of times like in, you know, the countries where we operating it would be in a school, there's a kitchen where somebody is cooking a meal for the children that Patricia's into doing what they're doing all the time, because that's what they've been cooking for years. Now, what the tool actually does, it optimizes the costs with the nutritional intake. And they've been able to prove that they can actually reach 30% more children with the same amount of money by optimizing the, you know, nutritious recipes with the nutrients that go to choose schoolchildren, which is really, really exciting. And then maybe just to look at it from what Lauren was also mentioning with like, what are the consumer like so we've been, for instance, with smaller departments, we've been supporting a startup called Hello tractor. They are actually from Nigeria, originally no expanding in Africa. It's an Uber for tractor in one they do is they actually renting out those tractors as a for profit model. So instead of you owning it as a small farmer, you can rent it. So you benefit, the tractor owner benefits, and also the startup, which is amazing. So it's an optimization that you actually have to do. And just as a last example, a recent one that has been launched now Somalia country operations, actually launched an innovation on E shop in delivery. So you can imagine, like, your favorite delivery platform that you have, and over $100,000 of food that's being delivered to people, also because of COVID-19, social distancing issues, that it's actually safer in the better customer support for people to actually receive food assistance, so that they can actually receive what they're entitled to.
Things were known for that overview, and with that, I want to turn back to Marin and something that was mentioned earlier in this call is around urbanization and how urban centers are going to be some of the most hit by this pandemic and the crisis that we're in. And the question is trends are showing that we're shifting more towards urbanization in the future with majority of us living in urban centers over the next decades, and I was wondering how AI can accelerate the transition towards a circular food economy, which is a very natural process as Marin described, but in an urban center, where it's more concrete buildings and less so of the natural environments that we can think of.
Thank you. So first of all, I wonder whether this pandemic might actually slow down the urbanization. I mean, by force factor, we're now all confined in our homes. We're all sitting behind our computers on zoom, having this very session and we're figuring out ways to connect remotely. And I would imagine that this might have a lasting effect on the very notion and need for physical proximity, including the proximity to the big urban centers and Metro couples, as the heart of social and economic activity has driven urbanization, driving urbanization, so, just the question you throw out there. What I think about the food and the agricultural system today is is too complex to fully understand with traditional analytical methods. Building a circular economy of food will add additional complexity and this is where I think that AI is needed. If we look at the levers of system level change one of the most powerful one is to change the goal of the system. And in my opinion, so core economy is changing the wide what the why question basic, but another very powerful lever is our information flows and feedback loops. And if we then look at, for instance, the farming site, the lack of data or the lack of access to knowledge and data leads to unpredictability and risk, which is one of the major barriers to a transition from conventional I would say the fleet of agriculture to for instance, regenerative factors, practices, and farming is highly, highly local. So you need lots of data and you need data that's relevant for you on the field where you are at for your crops. AI can help overcome this gap, helping farmers to de risk this transition by making the world's knowledge and database easier to both feed, and to read or access, allowing them to identify the crops and crop systems that would work best in their specific environment. Secondly, if I look at the physical act, Also informational distance between farmers and producers on one side and markets on the other side, I believe that AI has the power of reducing the information lag. and reducing this is distance. I see potential for AI to help connect producers and consumers anywhere in the world to each other in real time. And then thirdly, and finally, if I look at both consumers and corporates, the availability of holistic information on the social and ecological impact of our choices in the food that we source, in the food that we develop, the food that we produce we sell or the food that we eat is essential. Real Time access to realistic and understandable data will allow everyone from a farm to a C suite to a supermarket shelf, to use their dollars, their power to vote for the world we want. And those are all elements where I think that I definitely can have a massive positive impact in enabling different future regenerative future.
So it does seem from from this answer that a connected system is very, very important to achieve the potential of AI. And I want to turn to Lauren with a question about the collaboration across food value chain stakeholders and various innovation approaches. How important is that? And how does that play out to achieve the best potential for AI and innovation and impact those who needed the most?
Thanks, Caroline, and in a word, collaboration is critical. There's a world of opportunity at the intersection of technologies like AI and issues like hunger or social impact more broadly. But we need a multi stakeholder approach to that we need business to come alongside government alongside civil society alongside a range of the disparate players involved. I think that we need to ask three questions that guide our way I think we need to ask AI for what AI for whom? And AI at what cost? So AI for what, what are we solving for? I think a lot of us have seen a technology firm or an app developer sort of produce a solution and go and look for a problem. And if we do that, for food systems that we're going to accelerate the food system we currently have, which I think many would agree, is not optimal. It is not solving for nutrition, it's not solving for Regenerative agriculture, it's not solving for equity. So I think we need to say what do we want to solve for in food systems. So take the reverse of that nutritious, equitable regenerative food systems put that at the center as sold around that so Director innovation and technology including AI towards That AI for whom? You know, I sit here in California where there's a lot of discussion about self driving cars, for instance. And I think the question of the morality of the machines that we're building and the intelligence that sits within them is critical, the ethics, the morality, the biases, we know that AI, for instance, will perpetuate the biases in our society, unless we're clearly redesigning it. So whether it's through insights from MIT's moral machine or elsewhere, the algorithmic Justice League has good insights on this topic and how do we actually put ethics and and equity at the center of our AI systems, including an especially as applied to our food systems, ai at what cost COVID is prompting unprecedented digitization it's really fascinating to watch and in some cases, it's wonderful and and can help us meet this moment in different ways. Also, data is The new gold and much of it is owned or will be owned or owned or controlled by very large players, governments, big businesses into something Marin said before, you know, how, how is that? How are those data accessed by the people who it's about or the people who use it, farmers consumers. So I would posit we need more open source systems, more protections, more savvy citizens and tech technology consumers. So all to say this is complicated. Collaboration is very difficult. I'm someone who, by virtue of having worked at the World Economic Forum with a range of different sectors and now working with clients, including in the tech sector, but being myself very focused on SDG to the Sustainable Development Goal to an other hunger and food security challenges, that interlocutor role is difficult. These actors speak different languages have different cultures, and actually having more people I think at the center who can translate the power of technologies like AI into social good, I think it's critical. And I recognize that there may be many in the crowd who play that role. And I encourage you to do more and more of that we need you.
Thanks, Lauren. And with that, I want to turn again to Bernard to give us some of the context in emerging countries and from your work. What does a scale up of innovation and AI in those developing country context look like? And what needs to happen to ensure that the benefits of AI to Lauren's point about for whom really impact those who need them the most?
I think I structure is a three across three points. I think the first element of this is like how does investment into AI in developing countries look like in in food? And the truth of it is generally speaking investments. If you look at venture capital, money or business Angel like generally is underrepresented. Like Food compared to like other sectors if you want, right, and then and even in there, a lot of the investment goes into like food delivery apps in developed countries where, I mean, there's nothing wrong with like providing funding for these types of apps and startups. But like, I think there's lots of more potential that actually can be done also for developing countries. So this is work in particular thinking about, like, local talents, that could be supported in some of the hotspots in Africa in the Middle East and some other areas where across Asia or Latin America, like where there's talent, there's interesting startups, there's definitely drive in those startups typically have a much better knowledge about the local innovation ecosystem. The second element of this is probably about like how do you bridge the worlds What are currently the AI hotspots in you know, you're probably talking about like Silicon Valley, some parts of Europe, in China in other areas in the world where you have lots of artificial intelligence talent researchers Where they are on the forefront of some of these developments? And how can you provide some of the technology and make it more egalitarian inaccessible to other parts of the world could be by, you know, companies providing, you know, resources for free pro bono or like also focusing on building up, like local innovation hubs in Africa, some of the tech companies have started doing that, to actually also hire local talent, and so make it more accessible for different countries to actually do that. And the third element is about how do we inspire people who are let's say, serial entrepreneurs, you are a founder of startup serial founder, you have done it before you've been successful. And to see that it is a potential not only for charity, but you can actually create a business that has a positive impact on the planet in ending hunger. And I think this is a huge opportunity thinking again about the example that I had about Hello tractor and like there's a couple of others people can check On our website innovation with wp.org. Another one was like sweetness supporting a startup called Cloud to street. They use remote sensing to provide flood monitoring, preparedness to governments, local communities. And it's a it's a for profit startup. And I think, actually, the the knowledge that it is possible that you can actually create a startup you can create jobs for yourself and for others and still have a positive impact is something that we have to broadcast more actually open to that. And you know, maybe for some of these scaling questions that you have, Carolyn, it's maybe some of the solutions should be a nonprofit solution or something that should be offered at no cost is an open source platform. It might also be a possibility escape.
Great, thank you so much. This has been a great discussion so far. And I have one last question before I turn it over to the questions from the audience. And the question is to all the panelists. I'd love to hear from you. What's your views on the future of food As AI becomes fully integrated into the food value chain and the food industry, what does that future look like? And I'd love to start with Lorin, on that question.
Sure. So I think there's some bright spots, some open questions. So let me share a few, I think some bright spots, I think we'll have greater efficiency and intelligence by applying AI more to the future food. I think those supply chain efficiencies that I referenced, I think could get us to a world with very little waste. For instance, I think the intelligence applications are many crisis spotting, optimizing for what we want, including for nutrition, etc. I also think that AI can help us to really accelerate new types of industry within the agri food space. I had mentioned before, but I'll mention again, indoor agriculture or again, controlled environment. Agriculture, I think is a really interesting way for us to reimagine production, and it has to do a bit with what Marin was saying as to, you know, maybe we're not going to be urbanizing quite as quickly, maybe we're actually going to be flowing backwards into more rural spaces. And we're going to becoming more decentralized as a as a global society in terms of, of our food system. So actually bringing food production closer to where people consume, it could be really interesting a range of technologies to do that. Other types of industry, alternative proteins, I think, those of us who look at the blockages for instance, in the meat supply chain here in the US are switching to other types of proteins. And that's everything from cellular agriculture to plant based proteins to things like algae and fun guy, you know, and, and there's a ton of applications for digitization and again, artificial intelligence, their third type of industry, you know, Biological Sciences, where food and nutrition meet those sciences. I do think there are some open questions. You know, how can we enable As a sort of agency and ownership in decentralized food systems with more centralized data systems, right, there's almost an ironic tension here that as we get more decentralized in terms of our food societies, we actually may have more control in a central point in terms of who owns those data or control some of the algorithms behind them. I think another open question, again, as I mentioned before, both COVID and AI accentuate and accelerate in equities. So how can we redesign for greater equity?
I'll maybe end this point by saying I was part of a are viewed a webinar yesterday by SG were one of them talked about what they call the urgency of the long term. And I think that's a really profound question. What's the urgency of the long term for food systems right now and and how can AI play into that? I was reading the report that that SG shared yesterday, and I have a question for all of you. So what do g did? FedEx, Uber and slack have in common? Well, I learned that they were all founded during an economic recession. We're moving into an economic recession right now. And I think the point here is that innovation can and hopefully and probably will be born out of this moment. And so I invite us to ask what is the good that we can design out of this for our food systems, in addition to some of the impacts that we started with at the beginning of this conversation, which will be deeply felt? How can we rebuild food systems for the future in the way that we want, again, around regenerative agriculture around nutrition? We can design that future that we want and we need to do it now.
Thanks Lorin, Merijn, would you like to share next?
Absolutely. And I'm gonna build on what Lauren said. I see AI as an enabler and I agree with Lauren's question that she asked previously on AI for what? And for me that what question is answered with the vision of a circular economy, a food system that has a positive social and ecological impact by design. And I like the fact that Lauren actually just teed this up so perfectly by mentioning the word of designing a food system a number of times in her response. And reacting to that sense of urgency. I was on a call yesterday, with Salzburg, global. And somebody mentioned that actually, the moment is now where we thought that we had six to eight years to react. We actually have six to 18 months, because in the next six to 18 months, trillions of dollars in aid packages will be released to restart economies all across the world and making sure that some of that money or actually a large chunk of that Mining also goes to the redesign of our food system i think is important because it will set the direction for decades to come. And to transform towards this new system, which I believe can literally become fuel and the flywheel for regeneration, beyond just food, but food at the center will only be able if we also get back to nature, get back in touch with nature, through regenerative farming, by applying biomimicry and starting to look at nature as a source of solutions, with 3.8 billion years of life experience, rather than just a source of raw materials. However, in contrast, and Oren already mentioned controlled environment, agriculture. I think that AI also is a platform innovation, and in In my opinion, AI is one of those things That opens the door, as Steven Johnson called it in his book. It opens in adjacent possible. And I think that the circular economy of food is possible and power in part by AI.
You, Bernard, would you like to take us home with this one before you open up?
I guess I'm a technology optimist. Like, oftentimes people would ask me, so is it possible to actually achieve zero hunger by 2030? And the answer is, it is possible. Yes, we do need to end the wars, we need to tackle climate change and there's certain underlying issues that have to be progressing. But technologies like artificial intelligence, offer up opportunities, that could be those breakthroughs, that step change that are required to really get towards the goal of zero hunger. And like, I just want to make this you know, plea of support of like, you know, out there, people Or like, watching this also, like, if you're inspired by that or like, if you're working in AI, like one yes, obviously, there has to be an underlying infrastructure in place in a lot of developing countries, connectivity, access to internet, you know, devices like this is all things that need to be in place. But then again, what we are seeing this point, even right now there is communities like very, like poor communities where it's maybe a new job, but it's the entrepreneur that has the smartphone that provides services to the rest of the common communities all of a sudden, rather than everybody individually, owning a smartphone. And I think this is where like, we'll see a lot of change in that regard in artificial intelligence will help us make that next step change, even from the examples that we see right now is we saw that, okay, you can cut food waste. If 30 to 40% of food globally, right now is being lost because of food waste or post harvest losses. Imagine what we could do with all that food that's currently going waste. Similarly, the example I gave you the school meals, if we optimize diet and they said, That's nutritious meals that taste good, we can optimize nutritious intake of things that reduce the cost and you actually have more for everybody in and then thinking about like, how can you be just a lot more accessible for people in developing countries like take for example, a smallholder farmer right now, maybe they have had disadvantages like because they were not able to go to school and they didn't learn to read and write. Imagine a world where there's low cost devices, where there's a voice assistant and all of a sudden that voice assistant provides the smallholder farmers with precision agriculture advice. Imagine what like how that could change. And like this is one example of like, imagine a future in which it doesn't mean in 10 years from now, this the phones exist, artificial intelligence exists, like voice interfaces, exist. So like, there's lots of things that can happen right now even.
And like, this is where I'm really passionate about like, okay, let's think about like, how can we create the social impact unicorn? So like, let's define how these kind of, instead of a billion dollar valuation, we're focusing on creating startups that impact maybe a billion people. It's really something that is, it's, it's possible, we just need to use the energy and the creative, the bright people out there with who are the technology, artificial intelligence with to actually do that.
Awesome. And that's exactly what experts tries to do through its competition. So it's great to see that everything is coming together. Um, it's now time to open up for questions from the audience. And we've been receiving almost 40 questions over the course of the webinar and some emerging themes happen in these questions. One, one major one was the question around data and One question or a couple of questions that came in was around the governance, the privacy, the ownership of that data along the value chain, and how AI can. What's AI's role in that game of ownership and privacy? If any of the panelists would like to take that question or tag on please, please feel free. Lorin, maybe?
I'm happy to do so let me tee up what I think is the need and leave it to some of the smart people who are attending this webinar to figure out how to solve it. You know, I think we need to look at where data is being collected, such as about farmers and about farms, and understand who will have most access to the insights and the control of the recommend nations that are made from those data. So it sort of comes back to what Marin was saying around information and feedback loops, right? If you're pulling information about a farm, but it's going into a broad system that then creates some decision making and recommends to you a certain product. And that's what comes back through the voice that Bernard was mentioning, that you actually take a little bit of agency away from that farmer to make other different choices. And that's a way that's that's actually blind, you know, that some of those systems of power actually made a little bit more, you know, are a little bit shadid. And so I don't have the solution here. I know that there are many who are working on against open source platforms where we can use artificial intelligence to be a bit more democratized, both in the governance and the access to those data. And I would invite if my co panelists know of good examples to offer them, I think there's some of the words But none that I would point to as being sort of centrally successful at that moment to rebalance.
Maybe just add a quick comment on this also, like, I think there's two aspects that are relevant in the, in the ethical considerations. One is, I think we need to make sure like we I mean, like everybody, like so like, if you're working on artificial intelligence is to like, bring in other voices, like make sure that like the knowledge that you're creating is like, some of it may be proprietary, but like, just to increase the dialogue, also, that it's not the one Artificial Intelligence engineer who decides on ethical considerations, but you actually need to bring people up to speed communities that are affected by local governments, people across the world to actually understand some of these implications that they might not be actually be aware of. So like, there's definitely a question in that regard. The other one is how can you leverage trusted in entities as such like in I mean, we have been successful for piloting a blockchain system where we digitize shipping documents bill of lading, from Port of Djibouti into Ethiopia, in one of reasons why we have been able to do that with the governments with low communities with the port shipping agents, because it's clear that we have the benefit of the people in our hearts. And like this is something that is just also something to remember when we're talking about artificial intelligence. If you design a system, and you make it clear from the outset that something is governed by the local government, or the community or its self govern, I think that increases the likelihood of acceptance of substitutions.
Great, thank you. Thank you both so much. Another theme also that came up was around the challenges that need to be overcome for for the potential of AI to be achieved. And specifically, some folks mentioned the political contexts that sometimes create like a pull factor when you're trying to push Something forward. And another was around incentive models that currently exist that are damaging. So economic like subsidies or things like that that could hinder the potential of innovation given the rigid existing incentive models. Merijn or Bernard or Lorin, please feel free to jump in on that one.
But maybe just to start, like, I think wherever there is an inefficiency in the system, somebody is benefiting of it. And you have to be aware of this. Like it's not limited to artificial intelligence that's limited relevant for any type of innovation or startup work. And you do need to depending on what you want, I do think there's a there's an element of sometimes in startup culture, people talk about, like we need to disrupt everything. Maybe sometimes it's more a question of partnerships, rather than just disrupting, where like you need to work with people you need. To work with the communities like in I think that's, that's an element that you have to be aware, I would say, and, I would just encourage everybody to also think about like, how do you actually do that on a on a day to day basis, if you take the first step. It's always about like you're trying to solve a problem. And in a lot of times, these solutions are very basic, like, oftentimes, it's just like solutions that are looking for a problem and it doesn't actually solve anything.
Let me come in on that. I love the question about incentives, because I think at the end of the day, people don't act because they should do something they act because they're incentivized to do that in whatever ways including through the market. And I think this comes back to how we design our AI What are we optimizing for? One of the myths I think that we've somewhat moved beyond but not fully is the idea that we just need to produce more to feed the More More and more yield. And as we've grown, I think grown into a next phase to say we need to focus on what we're producing it the nutritional value of that and how we're producing it. So if, for instance, the sensors and the entire Internet of Things system at a farm level, we're not only optimizing for yield, but we're also optimizing for carbon capture, for instance, and was giving us some information about how much carbon was being captured in the soil or other types of biomass and was actually feeding into a marketplace for effect carbon capture, I think we really start to Marin's point to redesign our systems through those incentives. Yeah.
It's a very broad question. So we I think we could fill a whole whole seminar on this one. But I'll jump on what Lauren just said. Indeed, policy is is often counter acting innovation. And hindering innovation and slowing innovation down. But changing policy is tough, because policies often actually the product of politics, which is the product of popular pressure. And so far, we've been really good at showing what's wrong at bashing what's not good. But we haven't really created a utopian view of the future yet. There is very little out there actually, that displays a positive future for the food system. And I think that creating and generating that actually, is very important. And for instance, we've worked recently with thought for food, Swiss innovation platform and launch a challenge on the circular economy food and we had with without any press almost, we had 3000 startups from 175 countries, mainly Gen Z's that submitted their problem. Now, I'm not saying that any of them are necessarily perfectly circle, circular, but I'm saying that amongst them, all the key elements and pieces are there. And we need to stitch that together and tell that story. So that we can also mobilize power, popular pressure, I would say, to to overcome some of these challenges. And on the point of yield, I totally agree. We consider today's food to be cheap. But today's food isn't cheap at all. For every dollar of food that we produce or revenue that we generate in the food system, actually, we pay as a society $2 $1 in social cost, and $1 in health related costs. So the food that we're buying that we think is cheap isn't isn't cheap at all. So these externalities as we call them that are not in in the price today. They are problematic. We need to make sure that the food system compensates farmers for Everything all the value they create beyond just the commodity that comes off their field, but also the social value, also the ecosystems value, etc.
Thank you so much for these answers. Um, one other theme that came up in the questions was around duplication of efforts, especially when we think about innovation and all the little startups that try to solve the same problems. Sometimes a duplication happens and the question to panelists is, how do we deal with that? Is it positive or negative?
Yeah, this is a this is something that I truthfully can respond to that is like, obviously, wasteful duplication would be not smart. And I think we should try to avoid that. So like, part of what we're trying to do like everything is on our website, innovation wi fi.org. So we're trying to also broadcast more information about what works, what doesn't online. And I think we all should try to like every accelerator, everybody who's engaged in the system like as much as we can. It's my advice to you please share it, make it publicly available, what works, what doesn't, especially if it's for the benefit of people there. But there is an element of competition isn't necessarily bad, especially if it's startups, or if it's early stage high risk innovations. Like a lot of times what we're seeing is that with the startups that we're working with, it doesn't matter how well the first pitches or how first how well the first round of is, it's, it's the quality of the team that determines whether the startup will be successful, which means and sometimes you're lucky and sometimes you are not. So you might have two startups that on the surface look very similar, but one of them might be successful, the other one might not be. So I do encourage us to also look at that from the element of if you follow startup logic, like nine out of 10 startups fail It's not necessarily bad if we have multiple attempts with different approaches, it can actually help us find the one startup that one approach that will help us really make the difference.
I couldn't agree more of an art, I think it, I think we should celebrate innovation in all its forms and encourage it everywhere that it's cropping up, and especially in decentralized ways. I mean, I think, where we see these solutions need to be appropriate to context we need the innovators in those contexts to be creating the solutions. And, you know, I personally work on a range of activities in Africa and supporting African entrepreneurs in the agri food sector. It's awesome to see what springs out of an innovation hub in in Legos or an address above or whatever it may be. So I think that we may have duplication of a kind but especially where it's responding to a local challenge, that context specific solution is going to be really impactful for their community for their market. You know, they're they're all I think some places where we should avoid duplication, we need interoperability of data systems, for instance. So there are some places we should find some broad scale agreements and protocols. But generally speaking, I think let's let 1000 flowers bloom, let's see what turns into that social impact unicorn. And let's really support those information flows. Because, as you mentioned, where we can find other pre competitive spaces or other spaces where people are ready to share, we can leapfrog we can offer tech transfers from certain portions of the world to others. I think it benefits our entire society to do so.
Yeah. I actually think it's interesting. I mentioned it in one of my earlier comments. duplication isn't isn't the problem. The fact that actually that we we dislike duplication. Again, it's this very mechanistic worldview duplication also means redundancies. And if we think about the way systems change and the system should change It's gonna be through emergence. It's not, it's not an unplanned fashion that the system will change. And if you think about emergence, it's based on experimentation. So we need as many experiments as we can get, because nobody has a crystal ball and those which experiment which startup, which in which innovation is going to set off that next loop, which one is going to hit the positive feedback loop and basically start to tip the system. So I really think that we need to think about our language. That's one of the things I'm trying to say here, because as much as we're aiming for a new paradigm, the old paradigm is so deeply ingrained in us that we typically go back to it and then to the point of failure. I mean, I just loved when I when I watched it, I hope everybody you've seen it, but the movie about general magic, which basically one of the things I took away from it is is failure, actually failure. And I think when we think about startups, and I agree, I mean, the stats are Not rosy. Luckily, that doesn't withhold many entrepreneurs from starting it. But you know, what's failure? How do we define, say failure? Is that purely in business success and business outcome? Is it an impact? Is it in opening that adjacent door for the next innovation? We might have to ask ourselves those questions.
In kill, this has been extremely insightful. And I think we have questions to fill another hour, but we were approaching on the on the mark to end and I want to one final question that came in was around gender. And the question was, is AI being used to empower rural women from developing countries in terms of facilitating their access to markets or even land rights? Given that women are the main food providers for the entire world?
I'll jump in and say I think we need to ask the question, How can it be used? I mentioned before that artificial intelligence can accentuate The biases in our society and that those are biases both in our cognition and also in the access that we have to assets. So there have been some examples, for instance, where a platform that was using social media activity to approximate bankability was then offering financial services to the unbanked. Now, wonderful design wonderful ambition. You know, very creative. This is in East Africa, by extension, because men were more likely to own the phones, and therefore more likely to be interacting with people and women were less likely to be it learned that men were more reliable, and therefore it should lend more to men. So we see this sort of learned behavior of the of the intelligence really creating a disproportionate advantage to men who have access to the majority of assets in the system and disproportionately so especially in agriculture, I think the question is, is really Important welcome, if my co panelists know, of ways that we're already correcting for that, but I think that's absolutely imperative that we designed to do so.
So what one of the factors that we assessed startups and innovation one is actually whether they are supportive of gender equality, like in their approaches in the goals that they actually pursuing. Now, having said that, the I think it's not always just AI. I mean, there's lots of other it's like, the question is way more complex and localized in terms of like, if you look at the specific outcomes, I think this is where I would encouraged us human centered design approaches that about like, seeing what people's needs are and trying to see like how you best fulfill them. that's relevant with AI and sometimes in an AI innovation. It might be gender equality, or inequality might not be actually the AI component of the start of a project but might be something completely else where it's like maybe disposable income or the phone example. Lauren just gave, like, I think you just have to be aware of those factors that whatever is being designed actually caters for that. Yeah.
And what Lauren said, I don't have the exact the good examples either. But what Lauren said resonated because one thing that really opened my eyes was when I read this reason, I don't have all the details, but bear with me was when I read this piece about the development of the microphone, that actually at the time was tuned to the male pitch. And then as more women started to use them, they were often told that they were screeching or they were screaming. But the simple thing is that the technology wasn't developed with the full pitch and frequency basically in mind. So I would say that one of the things is we need to make sure that technology by design is I'd say, gentlemen, but in general is unbiased. Now how to make it unbiased. I mean, I'm probably the least knowledge On a on this entire webinar, including all the participants, but here's what I think that we need to do. I'll trust a specialist that they will help us figure out a way to get there.
Thank you. With this, I want to give a big thank you to all our panelists for joining us today. And thank you to the audience in the attendees for tuning in and asking questions and engaging in the chat. If our conversation today has inspired you to get more involved in how AI is positively impacting food systems, I want to point you towards some very practical next steps. The first is that this webinar is the second in a series of three webinars focused on each of the series AI for Good breakthrough tracks, the purpose of which are to identify practical applications of AI to advance the UN's SDGs. Tune in at the same time tomorrow to hear experts discuss how AI is impacting the domain of the environment. The second thing I want to point you to is x price has been actively working through the future of food in the future forests impact roadmaps to identify a set of core food systems and forests problems, as well as highlight several breakthrough areas that can establish a more food secure, environmentally sustainable and health hazards the word world. So to get access to those, please visit the XPrize website or click on the link to the chat in the chat box to learn more, and to learn how you can partner with what we're doing at XPrize. Thank you again for joining us and we hope to see you in the future Earth webinar tomorrow, may 8 at 9am pacific time and goodbye