[00:00:05] David Bergvinson: Well, welcome back to Grounded Intelligence, we look at how to responsibly use artificial intelligence to empower farmers around the world. And today’s topic, we’re gonna be talking about farmer data. recall maybe our last episode, we talked about localization of data to develop artificial intelligence tools. Well, this topic is really important. This is about farmers’ data and how we ensure that they realize the value of contributing their data into these systems. I’m really excited with this episode to have two experts joining me today. So, uh, joining me from IITA, the International Institute of Tropical Agriculture, as a scaling specialist, is Dr.
[00:00:56] David Bergvinson: Thompson and, uh, [00:01:00] he’s been working on, digital platforms to support farmers, so we’re really excited to learn about his experience, especially as it relates to scaling agronomy efforts. And also, uh, by Ify Muna, who’s the director for agriculture with DevAfric and has a rich history of farmer empowerment and looking at food systems.
[00:01:22] David Bergvinson: And so I welcome both of you to the show, and we just invite you to amplify your bios, especially as it relates to, uh, farmer data. So Thompson, would you start off, please?
[00:01:32] Thompson Ogunsanmi: Yeah. Thank you very much. Uh, just as mentioned, I’m Thompson Ogunsanmi. I work with, uh, International Institute of Tropical Agriculture. I’m based in, uh, Accra in Ghana, and I’m working with a lot of smallholder farmers across three countries. Uh, basically coordinating the use of an innovation we call AkiliMo, and we have been able to work with a lot of farmers, uh, across the three countries.
[00:01:57] Thompson Ogunsanmi: Currently, we have, uh, over seven [00:02:00] hundred thousand, uh, database of farmers across the three countries. And what I do basically is I, I… We do, we do– we, we take some of this innovation to the smallholder farmers. We work with, uh, different value chain actors across the value chain who are connected to these farmers.
[00:02:18] Thompson Ogunsanmi: So our working is not directly with the farmers. We are working with those partners who are engaging, uh, with, uh, with those partner organization directly. And we have been able to scale the use of several innovations with them. And one of the things we have also done is we usually ensure that we get insight from this data that we have collected from these farmers, so that we can improve their systems using the insight that we are generating from that.
[00:02:45] Thompson Ogunsanmi: And, uh, for… With IITA for the past seven years, that’s what I’ve been doing. But aside that, I’ve also worked with GIZ for some years, and I’ve also worked with the International Fertilizer Development Center as well. Thank you very much.
[00:02:57] David Bergvinson: Wonderful. Wonderful. Great. [00:03:00] Well, look forward to our conversation. Uh, Ify, over to you
[00:03:03] Ify Umunna: Thanks for having me. Uh, my name is Ifeoma Nna, Director of Agriculture at Debafric. prior to that, I spent, uh, over a decade working with farmers, uh, across different organizations. At Sahal Consulting, I was the gender specialist, so I worked directly with female farmers to try and uplift them, uh, through access to finance, access to market, access to information, uh, such as, uh
[00:03:28] Ify Umunna: Which became Nourishing Africa, which is an entity aimed at helping, uh, MSMEs in the agricultural sector across Africa. I also worked at AFEX, where I led the strategy that works closely with, uh, over a million farmers across Nigeria to really help them scale as well. and now at Debafric as well, trying to put that all together and, and, and make things move for farmers.
[00:03:52] David Bergvinson: Wonderful. Well, you both bring, like I say, diverse complementary perspectives to this really important topic around farmer data. So thank you [00:04:00] both. Um, you know, just to set the stage, you know, one of the challenges we see with, uh, you know, the, the earlier digital divide was farmers not being able to f- fully participate in, in that digital transformation more than two decades ago. And unfortunately, you know, we’re at risk of seeing the same thing happen with artificial intelligence as it relates to especially advisory services, but also at other services like insurance, market access, et cetera. so, you know, our topic really is going to be how do we ensure that farmers not only contribute, uh, towards these models and data systems, but also benefit from them? And so, uh, you know, with that, I really want to get at some of the, uh, design features and even policy-related features, ’cause our next episode’s gonna be talking about policy. So this conversation will really be helpful for informing our next conversation around policy. So, um, Thompson, I’ll start with you. You know, just around, uh, the architecture of [00:05:00] in farmer data, and it sounds like you’ve got over 700,000 farmers on, on the platform that your team has been developing. How, you know, how do you design that to ensure farmers actually really benefit from the provision of their data? What does that look like?
[00:05:16] Thompson Ogunsanmi: Yeah. I-in terms of it, the, the starting point has always been the, the farmers themselves. And, uh, you know, it has been difficult to change because farmer are often treated maybe as a source of, uh, data rather than participant or stakeholder in the ecosystem. So we, we have been able to recognize that, that farmer provide information, and they do that repeatedly, maybe through government organization, to development project, to advisory services, to platforms, and that generally.
[00:05:46] Thompson Ogunsanmi: But the information most of the time remain in s- in, in, in the system. And the farmer may have to maybe provide some information again and again without building a portable, uh, lasting [00:06:00] record. So how we have been able to work with them is, is not directly with them. We are connected to agricultural extension agent who we have trained on how to engage with farmers.
[00:06:11] Thompson Ogunsanmi: And whenever you take them through the use of any innovation, you get their details. And that information, uh, gets to us via a, a, a sort of platform that we have established, and the partners working with the farmers are also connected to that. So they have access to that, but we have partitioned it in such a way that you can only have access to the database of, uh, farmer that are connected to you so that you don’t have access to the other one, uh, generally, except you have an agreement.
[00:06:39] Thompson Ogunsanmi: But the, the platform is open to everybody. If I communicate with another person, I could have access to that. But basically, uh, I see generally beyond what I’m doing, that it has been more or less like a, a sort of a maybe flow upward. That is, from farmers to project to platform, or maybe an [00:07:00] institution with that equally strong flow of value back to the farmers.
[00:07:03] Thompson Ogunsanmi: And you know, from our side, one of the thing we do is that we improve the innovative process in terms of scaling, and it is the data we are getting from farmers that we are using. We analyze them, we see the loopholes we improve, and we take it back to them to affirm if these are okay and these are their intention.
[00:07:20] Thompson Ogunsanmi: Maybe a sort of a feedback learning process that we do from time to time. But I think it has been very difficult in changing this. That’s what I’ve observed. Not simply, uh, about the technology problem. It is also about institutional incentment. There’s need for trust, there’s need for governance, there’s need for ownership, and also that power should be with them, for them to have the sense of ownership that this is coming from us, and it belongs to us, and we should be part of that decision-making in the whole process.
[00:07:49] Thompson Ogunsanmi: But maybe if I want to conclude on this, I would just say that there is a need for more farmer data, which is good, which is enough, but we need to redesign the [00:08:00] system so that farmer themselves and participating organization have visibility and also the voice, and they also have value in the data ecosystem.
[00:08:10] Thompson Ogunsanmi: Yeah. Thank you very much.
[00:08:12] David Bergvinson: Yeah, no, you– And actually you mentioned the word I’m always looking for is trust. You mentioned it in our first question, so I think you get a gold star for that, um, because it’s so important that we not betray the trust of farmers, especially for this topic of farmer data. Um, but you also painted a picture of a fragmented, uh, system.
[00:08:29] David Bergvinson: And so, um, you know, Ify, I’d like to talk to you a little bit about the infrastructure that, uh, really brings together, uh, these different data assets that live in isolated environments and, and how can we unlock that? Because in the case of India, where they have Aadhaar, for example, a biometric-based system to a unique ID to farmers, and then that resource is shared with different providers so that collectively they’re all or able to share data around [00:09:00] that client or that farmer. Um, what are your thoughts on how do we address this in Africa in particular?
[00:09:05] Thompson Ogunsanmi: Yeah, I, I think– okay, sorry
[00:09:09] David Bergvinson: So I’m gonna, I’m gonna go to Ify first, Thompson, but you’re welcome to add on after
[00:09:13] Thompson Ogunsanmi: Yeah, please
[00:09:14] Ify Umunna: Uh, I think it’s important that we also try and instead of just saying fragmented, we put it into context of the life of a farmer. So for instance, if I’m a farmer and I have given my information to the Ministry of Agriculture, and then a donor-funded project comes along and asks for data as well, and then I put it onto an agritech platform, my data is spread across three different, uh, to which my data’s been extracted, but nothing is coming back to me. Um, and I think fixing this requires some basic infrastructure. Uh, we talk about a persistent farmer or, or common identifier which a farmer can have, um, which would have standards and clear definitions as to what that, that means. But it’s not just a, a technical or structural issue. There is an [00:10:00] institutional part that needs to change because, um, oftentimes we think about who owns the data.
[00:10:06] Ify Umunna: When we talk about interoperability, the question of who owns the data, where does it sit, who has access to it, um, and what are the incentives for it to actually be interoperable? For many, you know, private sector, uh, uh, agencies and the likes, they don’t… There’s no incentive either financially or therefore or thereafter to really try and see how best to, to connect and to ensure that farmers are centered in this. So I think in, in trying to address it, we’ve seen the likes of, you know, AgriStack in India, um, trying to address this and, and they have had some, um, good wins in that, and then lessons that we can learn. But even if we were to put that in within an African context, there would still need to be some tweaks made from, uh, what they’ve been able to achieve in order for that to be applicable within our context.
[00:10:57] David Bergvinson: Yeah. A- and, and Thompson, please add on to this, [00:11:00] especially from the perspective of the infrastructure, uh, component of this a- and the projects that you’re currently, uh, operating in. How, how, how do we get the infrastructure in place, um,
[00:11:10] Thompson Ogunsanmi: Uh,
[00:11:10] David Bergvinson: to, to realize this?
[00:11:11] Thompson Ogunsanmi: yeah, i-i-in terms of that, i-in practice really, there, there is need towards, uh, treating farmer data as infrastructure. And i-in, in this case, it means that is it possible to continuously maintain a foundation that makes other agricultural services to loop into that and for them to build on, on, on that so that farmer can benefit via other services that have been put in place.
[00:11:36] Thompson Ogunsanmi: But what we’re having today is that the data is gotten and it ends there. And many organization, programs, they all collect data for their own immediate purposes. When the program ends, the information may not be useful to the farmers any longer or to be reused in such a manner that it will bring benefit even to the farmer, maybe as another set of services.
[00:11:58] Thompson Ogunsanmi: Because most of the [00:12:00] people or organizations who are collecting the data, they collect it for a particular purpose during the lifespan, maybe purposely of a project. A project has end five years, then this data is locked up and somebody want to render a service with the farmers, we have to reinitiate the process again, asking farmer the same question that somebody has asked before.
[00:12:20] Thompson Ogunsanmi: So that shows this requires several things working together. A sort of trusted way of identifying farmers to, to ensure that they are part of the whole process and they can reuse the data as an infrastructure. Then also to have a sort of common data standard so that system can communicate to each other.
[00:12:40] Thompson Ogunsanmi: This is very, very important. And also, will there be need for a sort of clear responsibility for data quality, security, and access? Because most of the people or organizations who do not want to give out data, they think more in terms of, uh, maybe this is an investment they have made themselves, and they want to keep such information to [00:13:00] themselves and all that.
[00:13:01] Thompson Ogunsanmi: But we need to look to the side of the farmers that there is need for transparency so that farmer know what information is held and how it is being used to their favor. And it could be in term of maybe sustainable, uh, use of data related to how do they improve access to market, how do they improve access to inputs that they are going to be using.
[00:13:22] Thompson Ogunsanmi: And most of the time that has not been visible. Uh, uh, one of the typical example I can give in one of the innovation that we are promoting is, uh, that I referred to as AkiliMo, is an example that provide a useful illustration because the scaling involves sector actor, several institution like extension organization, farmer groups, private sector that are all involved across the value chain.
[00:13:45] Thompson Ogunsanmi: And the information infrastructure has been used to help all those actor to coordinate better And what it means is that if I come to the platform and I’m working with a, uh, maybe farmer organization and there is a fertilizer company, fertilizer company can share certain information so that[00:14:00]
[00:14:02] Thompson Ogunsanmi: they’ll be able to sell their fertilizer, and this will reduce the– This has reduced duplication, and it has also helped across the ecosystem to be able to improve services for farmer. But if I, if I want to conclude on this, I will say that if farmer data is infrastructure-based, the farmer should be one of the primary beneficiaries of that infrastructure, not simply the person supplying the raw material.
[00:14:26] Thompson Ogunsanmi: They are playing a critical role beyond data, uh, collection we are having from them.
[00:14:32] David Bergvinson: Right. And, and, and so to achieve that, uh, Thompson, do you have unique identifiers for farmers that is then shared with other projects? Or is that just used within your Colima project?
[00:14:43] Thompson Ogunsanmi: No, c-c-currently what, what, what we have, uh, we– The, the, there are several organizations we are working with, they have access to the data, fertilizer company, agro dealers, and all that. And for us in the course of collection, we have unique identifier, and the reason is this: we don’t [00:15:00] want to count one farmer twice or three times.
[00:15:02] Thompson Ogunsanmi: So because most of the time when we engage farmers, they come back. The same farmer can come like three, four, five times. But we are able to identify that this farmer has received this innovation five times. We also try to also check the aspect of how do we ensure a, a, a sort of effectiveness and efficiency of the use of that innovation.
[00:15:22] Thompson Ogunsanmi: If we have en- counted you five times and you have only been able to– not adopt it, then we know that you are not really valid in terms of what we are doing. Yeah.
[00:15:32] David Bergvinson: I, I’m just trying to get at, you know, are you sharing that unique identifier with other organizations outside of your projects under IITA with the government, for example, or, you know, with a value chain actor that’s not tied to IITA? Is that– Is there some coordination of unique identifiers
[00:15:50] Thompson Ogunsanmi: Yes.
[00:15:51] David Bergvinson: that’s just sitting within
[00:15:52] David Bergvinson: IITA?
[00:15:53] Thompson Ogunsanmi: there is. Uh, currently for one of the innovation, we, we have what we call, uh, Akilimo Nigeria Association. [00:16:00] We also have Akilimo
[00:16:01] David Bergvinson: Okay
[00:16:01] Thompson Ogunsanmi: Association. We have Akilimo Tanzania Association. The partners themselves, they have come together to form an association, and they have been given access to that data, and they use that to work with their farmers generally.
[00:16:12] Thompson Ogunsanmi: But the aspect we also need to look into is that, do that farmers have certain benefit with those partners in terms of that data coming to them? So for us, uh, we, we, we are holding that data in trust for the partners because we are working with them. Then it is now for the partners to work with the farmers using that data to make informed decision that is going to affect them positively.
[00:16:37] Thompson Ogunsanmi: But for IITA, data that we have gotten over time, particularly for Akilimo, I’m speaking specifically now, the, the partner organization, even currently the Akilimo Nigeria Association, the data passes through their platform before IITA assesses. So which means
[00:16:54] David Bergvinson: I
[00:16:54] Thompson Ogunsanmi: already we are even getting– we are the one even getting data from them, not the other way around.
[00:16:59] Thompson Ogunsanmi: Yeah[00:17:00]
[00:17:00] David Bergvinson: Great. Okay, thank you for that clarification. You know, if you’ve already mentioned AgriStack in India as an example, um, you know, what are the aspects of that that merit replication, uh, for different African nations, and what are the things that need to be managed better, um, such as inclusion, for example? Uh, any thoughts on that?
[00:17:20] Ify Umunna: Yeah. Um, I think in terms of what AgriStack is doing as it relates to farmer ID, where you can, you know, access information as it relates to land or crop records or credit insurance, advisory services and the likes, um, that is something that we could easily replicate in saying that having a centralized database for such is, is, is great. Um, however, one flaw, for instance, that would work within an Asian context but might not work within the African context is, is linking farmer to land ownership. So for instance, we know that across Africa, uh, women rarely own land that they farm on. Additionally, [00:18:00] you have nomadic farmers who move and do not have land ownership, so linking a farmer ID to a location might not be helpful in the long run and might, we might end up in the same situation where we are now. Um, but I do think that when it comes to the architecture and sequencing that, uh, AgriStack has been able to do, uh, that’s something that we should be able to replicate, uh, on continent. Uh, but also looking at nuances such as land ownership and, um, distribution of, of capacity and time as well is something that we would differ from that of our Asia counterparts as well.
[00:18:35] David Bergvinson: Yeah. Good, good points. Thank you. Um, you know, that sort of leads to a, a point around the data itself and its control. Uh, Thomson, you mentioned that, uh, you know, you’re hosting this data on behalf of an ecosystem of partners. how is that data actually controlled? Like, if a farmer wanted to, for example, re- request their data to be taken off of your platform, is there an instrument or way for that to be done?
[00:18:59] Thompson Ogunsanmi: Uh, [00:19:00] yeah. The, the, in terms of the control, the, the, the starting point is that, uh, a farmer– we, we have the understanding that a farmer should not lose meaningful control over the infor- their information simply because another organization collected it. So there is understanding with the partner organization because we are not working directly with the farmers.
[00:19:21] Thompson Ogunsanmi: We have organizations that are working with them. So we work with those organization. They are able to assess and work with the innovation we have trained them on. Then the data comes to them and they share with us. And for us, we use that to improve the innovation we are taking back. We look at all the gaps we have observed and all that.
[00:19:39] Thompson Ogunsanmi: And in my own view, this does not mean that every farmer needs to manage technical side of their data personally. That might be too technical for them. But it means that farmers should have meaningful rights maybe over what is collected, why is it collected, who can access it, how is it reused, and, uh, who benefit [00:20:00] from its use.
[00:20:01] Thompson Ogunsanmi: So th-those are some of the… The farmer needs on their own, they needs to ha- to– they need consent process in, in terms of understanding that this has been generated from us, how is it being used, what are the benefits that are involved in that generally. And there should also be a sort of clear right to access and correct information, transparent data sharing rule, limit or in terms of secondary use.
[00:20:27] Thompson Ogunsanmi: Also, the need to also protect sensitive information of farmers because that’s also another thing as well. If those farmers are being gotten, do you also push them out and you expose this farmer in terms of certain things generally? And there will be also– there should also be a sort of a, a mechanism for complaint and redress.
[00:20:47] Thompson Ogunsanmi: So w- if farmers are having concern on those, uh, uh, data, we should be able to have that. Maybe that would take me to aspect you raised on policy the other time. Policy-related discussion should also be around data ownership [00:21:00] to make partners more willing to share information responsibly. And this should rather be about stewardsh- stewardship rather than ownership alone.
[00:21:10] Thompson Ogunsanmi: The, the, the, the data is more or less like in trust, although you have committed fund in getting it, but at the same time, the primary thing should be what benefit accrue to the farmers themselves. So to– So,
[00:21:22] David Bergvinson: Right.
[00:21:22] Thompson Ogunsanmi: yeah. So thank you.
[00:21:25] David Bergvinson: Yeah, then that takes us to, you know, the challenges of sharing data, the, these data cooperatives that you’ve just described, Thomas. And then if you, maybe you could embellish a little bit on that, some of the, the promises, you know, the ideology of data sharing, uh, but the realities of data sharing and data agreements that, uh, the data ecosystem.
[00:21:45] David Bergvinson: So thoughts, especially from the African context
[00:21:47] Ify Umunna: Yeah.
[00:21:48] Thompson Ogunsanmi: you, yeah
[00:21:49] Ify Umunna: I, I think, uh, for us, when we look at
[00:21:52] David Bergvinson: Afi
[00:21:53] Ify Umunna: a cooperative model, um, when it comes to farmers, they’re, they’re good because they address the fundamental issue of a [00:22:00] power imbalance. Uh, one farmer alone does not have the, you know, the bargaining power that he w- he or she would need against a tech firm or a large, uh, program donor of the likes.
[00:22:12] Ify Umunna: But if they come together, they have that power. Um, and I think it, it, it serves as a, a good to ensure that farmers are able to express, uh, their sentiments as it relates to data, where it’s going to be used, how it will be used, and how they feel about that. However, um, I think one… There are a few bits where it’s really important that these cooperatives, um, are, one, in line with what they say they’re going to do.
[00:22:41] Ify Umunna: So if they are speaking on behalf of farmers, uh, that they have the legal say-so to do so, that they are able to get the consent of farmers, and that they are actually delivering the information that farmers are actually, uh, asking for and not, you know, there’s no other incentive other than being, [00:23:00] uh, a spokesperson for this- these groups of farmers. Um, I also think that doesn’t automa- uh, a farmer cooperative or data cooperative does not automatically become farmer-centered unless they’re able to do so. Um, and oftentimes I think the, the two get blurred in saying that, yes, you’re d- you’re a f- a data cooperative for farmers, but you have not been endorsed by the farmers.
[00:23:24] Ify Umunna: They, they still are where they were before, uh, such cooperative came on board. Um, so and then that naturally, uh, dissolves any trust that a farmer would have had in the system. Uh, so I think we need to look more deeply into what that means, how we can ensure that, you know, these, uh, these cooperatives don’t become middlemen in the end, where, you know, nothing moves and they become bottlenecks in the process.
[00:23:51] Ify Umunna: But, um, ideally, they should be able to push this forward as, uh, spokespeople of farmers [00:24:00] collectively.
[00:24:02] David Bergvinson: Okay, great. Um, yeah, and that sort of takes us to the topic of, you know, adequate representation. Um, platforms that you’re managing, Thompson, currently, what do you see as the gaps in representation? Is, you know, uh, can you give us some examples of communities or segments of society that are underrepresented in your, uh, platforms?
[00:24:25] Thompson Ogunsanmi: Uh, yeah, I think this is one of the most important risk I’ve observed in, uh, uh, artificial intelligence for agriculture. Uh, if certain farmers are missing from the data, the technology can reproduce that exclusion. And i-if you look into it, we should not assume that the farmers with smartphones, reliable connectivity, or frequent interaction with farmer institution represent the whole farming population.
[00:24:50] Thompson Ogunsanmi: No And what I’ve observed over time, particularly with what we have developed, we, we observed that, uh, currently, uh, it’s about [00:25:00] 70% of the data that we are having is actually not coming from farmers that are having smartphone. It’s coming from farmer with feature phones. And as a result of that, what we have done, we have been able to develop other formats for them to be able to have access like printable guide.
[00:25:15] Thompson Ogunsanmi: We’ve developed videos that can draw their attention. We’ve looked into the aspect of, uh, SMS, USSD, IVR. So we have other channels we have developed which help them that you do not need to have smartphone before you’ll be able to access that. And that means that inclusive data collection also need multiple trusted channels.
[00:25:35] Thompson Ogunsanmi: That, that’s what I’ve observed. Rather than maybe depending entirely on, uh, direct digital engagement. So we need to ask not only how many farmer did we reach, also who did we reach? Which, which, which segment of farmers are we talking about? And at the same time, who did we miss that we have not been able to reach as well?
[00:25:54] Thompson Ogunsanmi: Maybe whose experience are represented and who is not part of the whole process. And when it comes to [00:26:00] the aspect of feedback, that is where we miss it. If we don’t look at it holistically, then the feedback will be influencing a lot of advisory services in future. Particularly example I can give to be specific is the aspect of gender.
[00:26:12] Thompson Ogunsanmi: Gender is particularly important maybe to point towards a gender responsive approach in terms of including disaggregation to make sure that information cut across both male and female, and to also look into stronger participation of vulnerable groups that we, we have in a, a lot of, uh, farming community.
[00:26:33] Thompson Ogunsanmi: You have a lot of people who are not educated. They are totally– they cannot read or write. So what kind of information can we package for them and they are not totally excluded, and we can still get their information for them to be able to have access to extension and advisory services. And the same principle also can be extended to youth as well.
[00:26:54] Thompson Ogunsanmi: A lot of youth, they want information to be packaged in a way that is attractive to them. [00:27:00] And, uh, many of them want to look into the fact that they are– they, they, they, they, they should have the information to be able to do business more smartly. But generally, on this aspect, representation must be designed into data system before artificial intelligence is trained,
[00:27:15] David Bergvinson: Mm-hmm.
[00:27:16] Thompson Ogunsanmi: not treated as correction after technology starts maybe, uh, failing that certain group.
[00:27:22] Thompson Ogunsanmi: But it should be there from the beginning. Yeah.
[00:27:24] David Bergvinson: Correct. Yes. Being very intentional around, inclusion as a design
[00:27:29] Thompson Ogunsanmi: Yeah
[00:27:30] David Bergvinson: for sure. Thank you for that. Um, you know, that sort of leads us to, um, you know, how do we return value back to farmers, you know, after their data’s been collected? You’ve both have talked about, you know, giving value back to the farmers.
[00:27:45] David Bergvinson: I- if you, maybe you can give some specific examples of where farmers have contributed data and they have benefited, uh, from that. You know, maybe access to financial services or markets. So, um, if you could amplify, you [00:28:00] know, what incentivizes a farmer to contribute?
[00:28:04] Ify Umunna: At the end of the day, I think what we know that the strongest model is one where maintaining your data as a farmer remains useful because it’s connected to a service that improves your livelihood. so looking at that, we’d say that of- in historically we’ve done things such as, you know, pay for airtime or given them a stipend for the duration of a project, but then after that nothing happens, and it’s as if, you know, it was all a waste of time. Um, but we’ve also seen new innovations or, or, you know, methodology where if provide a farmer with– a, a, a farmer pri- provides you with data, you’re able to give, you know, accuracy as it relates to weather or, you know, you have access to credit as a result of that, and that has a direct linkage to the betterment of, or the improvement of their livelihoods. So there needs to be… At the end of the day, hu- farmers are human beings. They need to get something in return that is tangible, that is not a [00:29:00] long-term, multiple decades, you know, re- reward for their data. Uh, we’ve seen, uh, cases where farmers will provide information, uh, or data as it relates to, you know, their production, and they get information as it relates to climate or location or, you know, access to markets or access to, uh, inputs, and that has direct consequences on their livelihoods.
[00:29:24] Ify Umunna: So we need to instill that model in what we do and treat farmers as businesses in that a business would not work with you if they don’t get anything in return. Farmers are a business, I think that’s the mindset that we need to change because they already have that mindset. They might just not be able to articulate it as well.
[00:29:42] David Bergvinson: Yeah, very good point. so our, our– we’re gonna wrap up our conversation here, and maybe just, uh, for both of you to provide what was the biggest takeaway f-from this conversation and from the paper on farmer data. You know, if you could leave the audience with one call to [00:30:00] action or insight, um, that would be really helpful.
[00:30:03] David Bergvinson: So Thompson, um, what’s, when your, your big takeaway for the audience?
[00:30:07] Thompson Ogunsanmi: Uh, yeah, I think, uh, my big takeaway is, uh, we need to know that the future of agricultural, uh, artificial intelligence depend not simply on collecting more farmer data, but on building a farmer data system that farmer can trust and benefit from them. And, uh, in my own view, that will require some shift. Uh, maybe for an example, uh, when farmer becomes participant, not just data sources.
[00:30:37] Thompson Ogunsanmi: So we need to get to that level. We need to get to a level where data collected from farmers should be reusable infrastructure. And the last aspect should also look into the aspect of how do we ensure that governance evolves alongside technology. Technology can take, uh, can make data collection and artificial intelligence can be more powerful, but the question of trust, [00:31:00] ownership, consent, asset, and accountability must be considered alongside with technology, uh, development.
[00:31:07] Thompson Ogunsanmi: Thank you very much.
[00:31:10] David Bergvinson: Great. Uh, we need to invite you back for our policy discussion.
[00:31:13] Ify Umunna: Mm-hmm.
[00:31:13] David Bergvinson: Um, if your big takeaway, please
[00:31:16] Ify Umunna: Uh, I echo everything Thompson said. Um, and I think for me, uh, we have to change the way we think about farmers, not just as raw materials as he had said, but as actual partners. Uh, when we work within a corporate or whatever the case may be, we are very clear-cut as to how we engage one another, need to be mutually beneficial, but we just don’t have the same mindset with farmers and I don’t understand why. I think that needs to change
[00:31:41] David Bergvinson: Well, that’s a great statement to end on. Um, well, thank you both. This has been a, a very interesting conversation on a very important topic. And so if you also found this interesting and wanna learn more, you can go to the agx.community to, uh, learn more about the farmer data [00:32:00] discussion paper offers a lot more detail than what we’ve talked about, uh, on this panel. And also just sign up, uh, for this podcast, uh, Grounded Intelligence, to join us on this journey as we explore how to use, uh, artificial intelligence responsibly to empower farmers and ultimately to deliver food security. So, um, thank you both, and I look forward to following up on your, your work, especially as it relates to farmer data and how do we unlock it to deliver value back to farmers.
[00:32:30] David Bergvinson: So thank you
[00:32:31] Thompson Ogunsanmi: Thank you.
[00:32:32] Ify Umunna: having us
[00:32:33] Thompson Ogunsanmi: Yeah
