 I think Arkin added grain quality in post-harvest because he saw my hand go up. I got scared, yeah. But I just wanted to respond a bit to Ed and just put another thing up. Okay, when we get to yield, yield is one thing, but that's not what rice is. Rice is what's on the plate when you eat. So when you take the hull off, you can lose a huge proportion of that yield due to pre-harvest management and post-harvest, or pre-harvest management, pre-harvest conditions, environmental conditions, and post-harvest losses. And it's really, really difficult to map that into this process. And this is why you haven't seen the quality component, the quality that has been included in this, but it's really difficult to map that into this process and how we quantify it. And how we quantify it per ecosystem. So we quantify it on market prices, market price gains, but then when you put more, if you get higher yielding, higher quality rice, you have more rice, and then that's going to have an effect on market price again. And so it's all, it's a really very difficult process to capture all the components into what will end up on people's plates. Yeah. Okay, Serge. Thank you. Serge Savare from ERI. I'd like to come back to the point which was gross about the uncertainty of estimates, because I think this is a very important element. The first presentation of this session by Dr. Jagme was based on farmers' interviews. The exercise conducted at ERI, on the other hand, is based on scientists' evaluations. And I think that the both approaches have their own values in the sense that they perhaps may help us address the issue of adoption. For instance, if there is discrepancy between what the scientist sees and what the farmer perceives, what the farmer perceives is a fact and it's an indication of potential adoption of whatever technology might be available to address that particular problem. And so that may be important in addressing this kind of uncertainty, which is, the adoption aspect is the most, we felt that the uncertainty is primarily attached with the adoption element. So I think that has a lot of value in itself. Then again, the scientists' point of view is not to be dismissed either, because there are elements that farmers simply cannot see, cannot perceive, and I can hear the argument of saying, oh well, if you bring a soil scientist there, then it will be a soil scientist problem. If you bring a proper soil scientist there, then it will be... Yes, but these things are, I think we have been carefully taking care of them, because they're part of the views in the process. Maybe one little addition to that, because what we will try to do is to very carefully document all of the assumptions being made and when it's being made public for external comments and reviews and input, we want to be able to document why we come up with that number. And in some cases, we will be able to link it to a clear, quantitative or unbiased set of data, or data of publications. In other cases, we will not be able to do this. We'll just say, okay, this is what the five guys came up with, or goats, if they're involved. And I think what is really important is to make these things as transparent as possible, so that we all know, okay, this is what we know, and this is what we don't know. And then, I'm coming back to Donald Rumsfeld, it seems, but it's this process. So we would certainly then like to open this up so that it isn't just the view of a small group of e-residists, and that we can also benefit from other information and data that other people may have. So it's a continuous process. But we have some more time, we can still discuss a little bit more. No more questions? Aliyu. Also, while he's getting the microphone, I see a few people from the private sector in the room here. I wouldn't mind getting a comment from her, from Sengenta, for example, how you do these kind of exercises. Yes, I was just curious about the wee competitiveness there. I'm second. What, and coming from the override, can you say there is another set of estimators which will go against the server? Yeah. Can you elaborate on that? Yeah, so in terms of the wee competitive varieties, there's, so we have one set of estimates. I mean, they've been produced by different scientists, basically. And I think we still haven't had a chance yet to sit down and really resolve the different estimates. So I pulled in the one that was comprehensive. One set of estimates was targeted to direct seated systems only, and the other was comprehensive to all systems. So for the purpose of this, I just pulled in the estimate that's comprehensive to all systems. But this is, again, another area where we have to sit down and have an internal discussion and see where we really stand. This is, since it is a process that's ongoing, we have a gradient of things that are resolved and some things that are still a little bit less resolved. Okay. I see. Do you have a microphone there? Yes, thanks. I came, maybe just a couple of comments. I think we are struggling as well to do this type of prospective analysis. Whatever you do, the only thing you know is that it's going to be wrong in 20, 30 years, that's for sure. But at least the value of and the miracle of doing such an exercise is really forced you to look at the trends and the uncertainties as it has been mentioned several times. So what we try to do internally without disclosing any strength in terms of information here is just working with kind of scenario planning, doing different scenarios based upon the trends and the uncertainties that we have in this overall market. Looking at all the parameters that you mentioned. If, for instance, we believe, and I've seen a lot of papers and discussion around direct seeding over the last few days, that direct seeding is going to be a trend, then we can understand what does it mean and what will be the consequence of that because that's going to drive the different systems. That's going to change the crop management in many areas for farmers in many places in Asia. So once we have identified these trends and certainties, we can also elaborate some scenarios and based upon these scenarios what we work now more and more we try to avoid to slice the solutions according to what we are working on but we try to work much more on a system base and to look at that system, how the system would work and what are the consequences for the crop management itself starting basically from the planting to post harvest as it has been mentioned. So that's basically what we try to achieve. Of course, even though we are sometimes working particularly on a specific area, like your breeding for instance, we cannot cover all the other aspects of the crop management. So that's why we are also working on partnership with you indeed, but we try to understand also the consequence of that. So that's how we do it. It's a new exercise as well. We do not believe that we have the answer alone and that's also why we want to partner and I'm pretty glad that you launched this Global Rise Science Partnership initiative. Thank you. Any other comments? We still have some time. All you have not. It's very important. Was it not important or has what I hear that sent it to not belong to the top of birth because there is no house solution for it? I think I want to show too slight because I think it's relevant now. I quickly want to show where are we now to give a more complete overview so that we're not missing out things. So these are the type of things that we have identified as solutions, potential, so we talked about improved in-breds hybrid C4, but here are some of the other things that in red we still need to do. So just to make complete, here are some of the quality aspects that Melissa talked about, but even we think about by-products, biochar, feed straw, etc. Resource use efficiency, the whole post harvest thing. I didn't have time to mention that but we're looking at drying, milling from small scale village milling to large scale business models, learning allowances and hermetic ring storage, laser leveling. So there are still a number of technologies, potential solutions in red here that we are going through. We didn't talk at all about these abiotic constraints or the potential solutions for droughts, emergence, salinity, interaction, high temperature, low temperature, these problem soils that are you mentioned coming out of your review. And finally, the biotic ones, this is in black what we have covered, the diseases, the pests, we have rodents, weeds, and here in here you see nematodes, snail and birds. So we haven't tackled them yet because A, we don't have a lot of in-house knowledge on them, so we need to think how are we going to tackle them, get some sense of the importance of these constraints, their occurrences, and then what are potential solutions. And all these ones in red, I'm sure are not even exhaustive yet. So I hope this provides a more complete overview of some of the technologies and at least my apologies for not having shown them before. Good, thanks for this clarification. So this is your final chance to ask a question or to make a comment. Okay, Robert. Thank you again, Robert Habib, Sierra de France. I would like to add a comment on the comment made by our colleague from the private sector. On the difference between the two exercises, as far as I understood this exercise you show us is an assessment exercise and you consider factors by factors to try to estimate what could be the future with a given probability, et cetera. So in the forth side exercise the future cannot be known. And what you do, you build scenarios in a systemic way and so you combine factors all together. For instance, is it possible that breathing innovation can be applied in the field and what are the all technical problems that has to be solved including all social and cultural problems all together. In the forth side exercise you do not try to predict the future but you try to build scenarios to be able to build the future according to what you want to appear. So I think it's two very different exercises and they can gain by combining. Want to comment on this, too? Yeah, in this we have some use of scenarios. We will do it in terms of parameters taking into account uncertainty about specific parameters. So we will do it, for example, in terms of the different higher and lower expected adoption rates and things like this. So we have scenarios for comparing specific aspects. But in the end we really do need to have some sort of an idea about the central tendency in order to make decisions. So we don't want to get too elaborate in terms of scenarios because we do, in the end, want to be able to say something about what is most likely, what is the best information upon which the basic decision is. But we will have some use of scenarios. We may have some use of scenarios in terms of, for example, actual yield trends as well. We have those trends projected out to 2035 to what degree we think we will.