 Welcome back to the AI for Good Global Summit here in Geneva, where I'm now joined by David Hughes, who's a UN Fellow with the Food and Agriculture Organization, a professor at Penn State. I know you've heard some of these earlier interviews and I think you're quite motivated to counter reply to some suggestions about what's being done in Africa, so the floor is all yours. Tell me about it. Right. So the fundamental constraint we have in Africa is the distribution of knowledge for smallholder farmers, which are the predominant employment on the continent. And unfortunately, this is due to a lack of human capability. So in Kenya, where we work, there is one human expert for 3,000 farmers, and we've tried to change that problem for the last 40 years as an international community and we haven't moved the needle. The approach that we're taking at Penn State through a platform called Plant Village is to leverage AI and cloud solutions to enable decision-making in the field by farmers. And so we have tested our AI solution, which diagnoses crop diseases against human experts on the continent, and it is twice as good as existing human capability. And that's a great prospect because as a piece of software, the marginal costs scale up. So critically with our AI tool, it works offline, so it's Africa-focused. It is developed and co-developed and designed by African women farmers in concert, and it works with African scientists. So all of this is a great leap forward. And now, through the United Nations, we are intending to take our tool, Plant Village, and roll it out across 70 countries in 20 languages to tackle a really big problem that we have called the Fall Army Worm. So there's lots of opportunities here, and indeed demonstrated capability, that when you have the tool and when you have the partnerships with local organizations and you work with a farmer-centric design, you can solve problems. But one thing I think we're saying off camera is that business can't do this alone. I don't think business can do it at all. So we're looking at a group of individuals who have less than $2.50 a day. So they have no ability to pay for this service. So therefore, if you want to give them advanced machine learning capabilities in a phone, what is the economic prospect? Well, the economic prospect is free, but of course it's not free. 15 years of the Facebook experiment shows that the negative externalities of free services are disrupted democracies across three continents. And so I would argue that we've done that experiment that hasn't been good. A successful model is tax and public funds. We've seen how well that's worked both in the EU and in the United Nations. So I believe that somebody should pay, it should be governments, and it should be for the benefit of their citizens. In addition, it should be governments in the north. Because if we don't solve the problem now, we will have to mop up the mess after the drastic effects of climate change hit these populations. And by drastic, I mean climate collapse. We're seeing complete loss of the maize crops in many parts of East Africa at the moment because of the twin cyclones. And we will see hundreds of millions of people become food insecure with a world that is between four and eight degrees Celsius hotter than it is now. Yeah, but David, I thought the whole point about the mobile phone was that it actually gave empowerment to small farmers. For example, they can now call up and see what the prices of rice in the local market at the trading. And they could also create their own kind of like social networks with other farmers so that they can actually almost create a union to negotiate the price in bulk between them. Absolutely. That's why we partner with local charities. We work on an Irish charity called Self Help Africa. They have an existing network of 28,000 farmers in Kenya. And we want to layer our solutions on top of that. We don't want to get in the way of local businesses, local markets and local entrepreneurship. And we want to enable that. Absolutely. That's a critical aspect. And in our tool, once you get a diagnosis, you then have the ability to reach out to seed supply networks and get the correct seeds for your particular question and need. So these farmers who are on 250 a day, like you said, if they all group together and they use mobile phone and AI together, that gives them a powerful tool to work with it, strength. Absolutely. We have a model of 20 farmers, a lead farmer and 19 following farmers. We give the lead farmer the smartphone and then she helps her neighbors. And we've shown in our trials that this lead farmer is going to move 20 to 30 times more when she has a smartphone than when she doesn't. She's empowered and she's very proud to be a community member helping her individuals surrounding her. And we're also trying to figure out, well, how many phones do we need in an area in order to have the coverage? And then if we can use all of these phone networks in an area, can they be canaries in the coal mine for climate change? And we take that learning and we cascade that out through traditional media like radio and TV so we can have a stacked approach. Surely as well, the hundreds and hundreds of really light, small, cheap satellites being created and launched all over Africa will help as well. Absolutely. We have a very large satellite product at UNFAO. It's called WAPR. It's enabling the decision-making on the ground by looking at how your field is performed in the context of water use over the last 10 years and what's the immediate future. We also work with ESA, the Sentinel Systems, and we also work with private enterprises like Planet and Digital Globe. Satellites are critical to help these decision-making. So now we have a phone which is pulling down satellite records on the field in local languages driven by AI enabling decision-making that's beneficial for this farmer in her situation. Just before you, can you remind me again you talked about something called Planned Village at Penn State, if anyone's interested. Yeah, Planned Village is a global public good multi-sided platform that we developed back in 2012. It's a very simple concept, make knowledge that helps smallholder farmers to enable their decision-making process, make that free globally. It has been working on AI solutions since 2015 and it is currently the leader in offline AI diagnostics and crops and now it integrates other things. That's why it's multi-sided. So it's a global public good. The goal of Planned Village is to get to 400 million smallholder farmers in the next 10 years working collectively as a global public good. Well that was David Hughes there from Penn State and also of course working with FAO in Rome. Thanks very much. Great.