 Just to get things going while you're thinking about the question you'd like to ask I'll just start things off but the common question for everyone and that is Following and maybe we can start Mary with you at the end And the question is if you had Just one silver bullet and you could spend it on whatever you wanted in order to take Your dream and what you talked about, you know to the next level or get it to what would that be and Sort of a follow-up given that I suspect most of the people in this room are people who are technologists What would you ask of these, you know, what could these people in this room do for you do to help you achieve what you plan to achieve? Okay, good question. I would say I would start with the platform that I didn't get very much time to explain But we have this idea about this software development platform that would take data Through an API link and then could then plug into many apps or models on the other end So it's just that that operability and the software functions that would allow all that pre and post processing to occur That's where I would invest first because there are there's demand for that information and there are lots of innovative Natural and social and biophysical scientists doing really amazing models out there and there's lots of data So it's to me that the first bottleneck to get rid of is the the platform the interoperability Like actually several people on this panel have talked about too. So that's her and the help From experts in the room would be thinking about what sorts of Pre-processing for either if you're an app developer if you're a modeler and you use that in your work. What sort of Metrics and information do you need and that helps people design the pre-processed pre and post processing algorithms? But also just if you have expertise on how do you put together such a platform? I mean there's lots of prototypes out there. We're prototyping one But it's a basic software engineering help and data pre and post-processing to help inform your Decisions is the kind of advice that I would love No, I think I do what they're doing at grow intelligence actually That was pretty cool No, I think I think that that's actually an example of In that in that part of the world and that in that sector of I think what what we is a real problem more broadly that It's easy for us to say if we had a bigger computer Here's the here's what we do with it if we have more storage we do with it But really you know that in terms of understanding climate risk. There's a real gap Where as you heard for Africa the the level of data availability of what? What is what makes vulnerability? What are what's the fundamental? Condition, you know, we have been really restricted to these pretty gross like annual scale You know World Bank type metrics which are valuable, but I think there's a lot of Potential to really understand a lot more subtlety both In places like Africa, but also here here in the US. So if we look at a at a city for instance, we know You know from from a lot of really detailed social science work It's been done that if you have a heatwave in a city there, you know the the the damages You know the the morbidity and mortality that occur in a heatwave or really Concentrated based on based on socio-economic conditions based on the strength of community networks And that's that's come out of you know really place-based long-term study, I think we're at the point now with again with search and Mobile social data where you know, we can we can really scale that up to you know to to assess vulnerability across Cross-urban urban areas based on based on those kinds of data. I think also we heard this morning about The kind of the the home The home data big data sort of internet of internet of things and and the home the whole home data system And I think in in terms of vulnerability of infrastructure. There's probably a lot of Potential for you know when there is an earthquake those sensors That are built in provide information about structures and and that potentially provides information about vulnerability of infrastructure as well So I think we certainly we certainly can can use a lot of help on You know the ingesting those those data algorithms for analyzing those data I think also the how we how we determine what vulnerability is based on this data Listen up Like I said, we're the youngest so we can use the most hope I think for for me you know the silver bullet now lies in actually Going through that third layer. I talked about which is really building the computational Capabilities to a level where you're doing all sorts of predictive analytics and that's you know people problem When I started the company, I moved to Kenya For two reasons it had open the capital markets and it probably had one of the most advanced technology infrastructures in Africa And so I was hopeful that I would find the talent I needed to build the tools we wanted to build in it as I discovered the type of data sets We are going to use all I had was a defined problem when I went there when I discovered the types of data sets We're going to use it quickly became apparent that the talent actually doesn't exist there unfortunately, so We just opened an office in New York as a way of addressing part of that problem So, you know, I think for us the silver bullet is is is more people It's it's really, you know We've we've done a lot of the hard work of going through the discovery process and figuring out how we want to go About cracking the problem. We just need more people excited about the problem thinking about the problem giving us ideas And I think that will well that will get us there Yeah, I mean so for me it feeds off what Sarah just said, which is Maybe one level up and you know on a more diversified scale its products So the biggest barrier to applying technology to solve, you know problems like the one that Sarah is addressing Is that you don't have teams on the ground or companies in these markets that know how to build these products They don't even know where to begin And so it goes, you know, it goes beyond even just talent and you know, Sarah opening an office in New York is a reflection of that. I think You know, I you're not gonna see in my opinion anyway You're not gonna see the local teams and the local knowledge in the local context emerge And so I think increasingly the way that the world is going you do see these pockets of expertise and because of where communication technology and everything else is going you can Access these different, you know pockets of expertise to address problems that are not local in in nature And you started to see that with you know, India and outsourcing and China and manufacturing and everything else So I think one of the things you know for for you guys and that that We think through I'm thinking through a lot right now with Sarah and everyone else is how could you actually more efficiently build Products so that you're not just sitting there waiting for the talent and everything else on the ground to emerge But you're tapping into you know people here Companies here in the ecosystem that knows how to build products But pairing it with the local expertise and knowledge that enables you to build products that actually address the problems that need to be Solved because you know, there's a disconnect if you just sit here in a vacuum and try to write algorithms Okay So snares the microphone. Yeah, it's let's start right here with Joe So this isn't a big data question. It's little data. Actually It's for Sarah. So you mentioned the silver bullet I was thinking the silver bullet for Africa as far as I can tell was a little data problem in that Women's education is the fundamental problem And if you correct that a lot of great things flow from it like reduced birth rate more Entrepreneurial ship among women who are liberated from their masters and all that all these great things follow. So do you see that as as Not a stumbling block, but it as a precursor to all this other Investing that you'd have to do to set the seed is you know Educate the women because he was happening Nigeria right now. So is is that even a precursor in your mind? so I guess Jeff and his introduction of me briefly mentioned I was a trustee of the Mandela Institute and That's you know on top of my girl hat I wear this activist hat and that activism is mostly driven around, you know, basically good policymaking and education and and really Setting education not and measuring education not by the number of people we educate But by the quality of education we give and in particularly we do with another organization that I sit on the board of called truth We focus a lot on child marriage girl, you know, 13 year old girls getting married off is very common in many African societies today and So yes, I think it's it's fundamental I I almost I just have to wear two hats to to address them simply because one problem versus You know one can still and needs to be cracked because that the first the problem You're addressing is is about behavioral change and that's just gonna take time And I think we all have to speak loud and push, but it won't happen overnight. So It's the it's a two hat problem Yes, sir So many of you spoke about the challenges that you have with manipulating data and managing data but in this In this area today's conference one of the issues that were not addressed Was the fact that there are large stumbling blocks when it comes to instrumenting the environment and the world in general The putting down of sensors by the millions across the globe, right? Have you guys given any thought to what would it take to instrument the world so you get relevant data timely data That you can actually get more interesting information from and the reason why I'm kind of mentioning this is Nobody has come to me And Told me that here are some interesting sensors I would like to have to get my job done and these sensors should cost no more than say 10 cents Nobody is coming up with these kinds of requests and And these are very relevant requests because in the day of 300 millimeter wave for technology is Getting a chip that is 10 cents is a piece of cake But nobody is actually looking at it and asking the right questions about how do we instrument the world? To actually get the data we want and I think this is something that we need to think about and I'm wondering if any of you Have given that some thought And if so, I'd like to kind of hear your comments If not, you don't have the time now I give you my email address I'm the director of a science and technology for IBM research and we're always looking for interesting challenges to help you guys do your job It's a great question and by the way if anyone in the room wants to weigh in on that that would be fine too But oh if I gave you a bag full of you know a big bag full like as big as this this room bag full of 10 cents instruments What will you do with it? Yeah? Yeah, well, so in terms of in terms of Weather and climate prediction certainly instrumenting the ocean is is critical and there is a network now of floats that are distributed globally, but they're very sparse and basically they they They're geolocated. They they submerge they take Take a profile vertically come up to the surface that gets transmitted to satellite they go back down and There's that that has definitely helped our our predictability in terms of What we know about the only know that's developing right now as one example So that's definitely improved short medium term range predictability But really what's what is needed as much much denser network of Sensors in the ocean to know what that what the current state of the system is because that really ends up influence in the atmosphere The second I guess I would just come back to Is not not answering your question, but maybe inverting it is that I think you know in some ways we're with remote sensing and and Instrumentation on land we know a lot about physical system. We know less about people kind of in real time and I guess that's where my my hope is that we can Take advantage of some of some what you're talking about some of the the devices that people are carrying around with them To try to learn more about that part of the of risk Yeah, Mary. Yeah, so we have I agree about the I mean the using cell phones We're involved in a project using cell phones to monitor social impacts of water fund payment schemes throughout Latin America and now Just starting in India and Africa But the physical the biophysical the two biggest ones that we see that would be great to have Some help with our groundwater data worldwide and soil moisture data so those two for predicting all of our crop agriculture production and all the water supply issues that that people have talked about those are all big parts of the projects that we are involved in and those two Parameters are really tough to get in most places So I love to talk with you about a 10 cent sensor that could help us get better data At lunch we were discussing this with bill where he he asked he asked me You know are you guys gonna try and do and build crop insurance models off of this and I said I wish because in the African context and like in the US you actually the Physical distribution of weather stations across the continent is very scarce. It's very sparsely Populated and so if you just take a map of location of physical weather stations around the world And you just look at it you'd be shocked at how sparsely kind of populated it is in Africa But more importantly if you then add to it one more layer, which is when was it installed? It will be like in the 1960s basically and so most of these weather stations aren't Working and we've looked at all sorts of ways around You know bridging that gap and and spent a lot of time with European companies that are currently Installing some of the weather stations that are going up and when you go through the budgets and you look at the line item put in for Vandalism and theft and contingency it's equivalent and sometimes greater than the cost of installing the actual weather station and so I Always say if we could figure out a way There's like a raw area and a raw space in the world today for us to experiment with Sensor networks and installing all sorts of sensor networks just to look at this problem. So I'll take a bag full of tin Craig Lewis with the clean coalition a question for Mary I was really impressed with the solution of coming up with the flicker photos for essentially your crowd sourcing sensor network and I Also really appreciated your emphasis on the importance of a platform solution And I was wondering if you've it seems to me that there's there's platform solutions out there for environmental data and including from Teradex, which is Founded by Bob Wenslow from out of Stanford And wondered if you had thought about how to I guess kind of put your platform on top of some of those environmental platforms and Whether there's some useful pathways there. Yeah, that that's a great point Yeah, we've definitely been in the midst of this kind of a landscape analysis to find out Who's already out there with certain? We know some of them because we use them and trying to figure out how to make them interoperable So not definitely not Reinventing anything but trying to figure out what are the big Functional pieces that we could tie together through these API links. That's really the the design and the all these new data Like planet labs that's spending spinning up from around here, too It's really interesting. There's a lot of good both platform and data And so it's a matter of getting the the whole soup to nuts of the the flows that we see needing to create spatial maps of Ecosystem service value around the world and linking those together. So that's the the real challenge. I would say not starting from scratch Yeah, thank you One last question perhaps Yeah Hi, my name is EJ. We're almost grad student at Cal Poly Pomona So my question is was actually a result of the last presentation on agriculture But maybe someone else on the panel may have info on this A lot of the seems that a lot of the focus on the future of agriculture has been on improving infrastructure and Production and the economics of agriculture as a general whether in Africa or globally Has I've also read that a lot of the produce here in America is Isn't necessarily 100% utilized. So how do we? track the waste from I Don't know from restaurants or maybe perishable foods that are produced them from farms large-scale farms and Has there been has this been accounted for in the algorithms or any research? I can Think you rightly point out that it's two different sets of problems in the US It is a food wastage and it's a food wastage problem It's the fact that it it does get through some level of processing and it gets wasted once it's hit a restaurant at home a supermarket, etc and developing markets the Harvest actually it's a post harvest loss issue. It never makes it to market. So It's and that's why it's it's an infrastructure problem. It's a markets problem When you look at the numbers around them, we actually have done So one way in which we're building out this platform is we're actually doing work Along with partners along the way and so one of our partners is the Rockefeller Foundation And we've spent a lot of time looking at what they're launching as their food and food and spoilage program food spoilage and waste program and We're investigating a lot of these trends and what we found is that in certain core grains in Africa Africa would be food secure If we simply eliminated the post harvest loss issue So if we only knew where crops were grown when they were grown and we could get to them You would be net Basically your supply and demand balances would flatten out. So it's it's a core problem But in the developing market context, it is a wasted. It is a it is an infrastructure Issue and in the US and in Europe. It's a spoilage issue after processing And there have been a lot of initiatives that have popped up But there aren't you know, and there are big numbers that get released But there are an actual Initiatives that I've worked with or know of in this context, but I haven't spent as much time in the US thinking about the problem All right, I think in the interests of giving you a break and I think we'll We'll stop right there. I I just want to add one thing To this discussion, which I think is is getting to a very interesting point this morning one of our speakers Jeff says Said and not surprisingly that that that big data is a sort of a science and engineering challenge And I think after listening to the panelists often, I think, you know, we could we could add a whole A whole bunch more dimensions to that I didn't actually agree with him and I'd be interested actually in your reactants not right now But because I think what we're seeing through this panel that it's a lot more than just science and engineering And I think that's what makes it so fascinating and so difficult So let's thank our speakers once again and we'll take a break