 Egin! Silydd is research programme manager at the protocol labs. I was amus to see in her biography likes to believe she's a rising star golfer as do we all. Silydd is going to talk to us about research road mapping, permissionless discourse for improved foresight coordination. Delighted to be here to talk to you about research road mapping. It's fun to talk about the importance of network and collective knowledge after this talk. So research road mapping is a special project within network codes as protocol apps. Network codes is a team focused on identifying, fostering and rewarding the creation of public goods. Is this moving? Okay. And how can we accelerate and optimize the creation of scientific public goods? And we claim research road maps are valuable coordination models to optimize the creation of those codes. They are dynamic and shareable instruments through which an inclusive set of stakeholders coordinate and align on the most impactful research paths, match problems and capabilities and embed incentives to maximize positive impact. And this is precisely the hypothesis we want to test. And to do so, we want to create a portfolio of actionable road maps, the ones we use to make informed decisions, and a reusable infrastructure for research road mapping that enables decentralized peer-to-peer incentives alignment between stakeholders. Hopefully, in doing so, we'll gather enough evidence to support our claims and establish research road maps as revolutionary scientific coordination systems. But let's look at what research road mapping actually is. So research road mapping is both a knowledge sharing activity and a tool for scientific coordination problems. The road mapping process requires research stakeholders to engage in a conversation where they share latent knowledge and identify leverage points in a research network, openly and transparently sharing assumptions and does it knowledge to quantify costs, impact and uncertainty of alternative paths and aligning on the creation of incentives, requests for experiments or investment thesis. A strategic planning and vision alignment exercise that results in a map. And that map is a goal-centered graph representation of the alternative path enriched with quantitative data to support decisions. A shared network of research goals, milestones and dependencies to bootstrap research collaborations and foster alignment between researchers and research funders, hopefully filling the gaps in research, fostering scientific discoverability and pushing the scientific frontier forward. And although it's easy to recognize the value of road mapping for coordination, there's much to be done to drive that adoption among researchers, R&D labs or industry labs and revive the openness mindset of science. We need to make it easier to create, share, navigate and update roadmaps, guarantee value attribution to those that contribute the map or use the map for decisions and incentivize high impact research path both prospectively and retrospectively. So let's start with the knowledge sharing part. Road mapping research requires knowledge exchange models. A data model, a common language to translate and combine knowledge from different fields, experts and views into an iterative and parsable graph that makes research paths are certainable. Multidisciplinary and interoperable data models that track knowledge across the sciences, enabling evidence-based decisions and scoping of research initiatives. And those models should be expandable, querable and personalizable to facilitate semantic merging of knowledge graphs, enabling different levels of precision and granularity without compromising the mergeability of graphs or the use of AI assisted tools either to synthesize knowledge or to make predictions. Besides, establishing priorities of ideas and aligning research needs require a substrate where this conversation can happen. Coordination protocols and instruments that allow crowdsourcing of ideas, gap analysis and prediction of research costs, risks and impact as well as the prediction of risks and impact so we can scope down anti-risk research initiatives. However, the adoption of research road map is hindered by a subpar user experience. In terms of tooling, we are missing good graph navigation tools, the capability of querying graphs or link nodes or whatever is a graph in the case, although we represent a road map to start with. We are missing collaborative features for decision making. In all these challenges, we are limiting the production of collective maps that would boost science discoverability. Speaking about collective road maps, we need research infrastructure to make it easy to create, collaborate and share knowledge. In fact, what we are missing is a playbook on how to road map a specific area of knowledge. How do I get started? What are the tools that I can use? What kind of models should I use? After I'm done with it, after I made my decision or after I wrote my grant proposal, what do I do with the map? Do I just write another PDF and publish it as a vision paper so that any other person needs to go to the process of breaking it down in parts again, creating a graph, putting it together, making decisions, write another PDF? How can we break this cycle? How can we make these instruments citable? We also need to have incentives in a way that researchers, funders, politicians or any other research stakeholders feels comfortable sharing their ideas in the open without the fear of being scoped. We need to track those contributions in a way that we can establish value attribution, either if they create impact or if others build on top of the knowledge they created. For instance, as I mentioned before, network codes is a team that works a lot on these revolutionary coordination models. We have innovative mechanisms to fund public codes, including research public codes such as iPerser's that are linked at the end of the presentation. These incentives are not necessarily just financial incentives. They also need to account for the metrics that you value in research like reputation. Is it collaborating or making a research map public something that you value on the metrics that you use to evaluate our own researchers? Or is the research risking their reputation by contributing to these instruments? If until this far this seems quite intuitive or useful as I sure hope so, I claim it's because I'm using a set of knowledge exchanges models that were developing with our extended network of writers and partners. On my right, we have discourse graphs. This is a schema that maps scientific arguments into their constituent parts. So we have the research question that we are trying to solve, all the claims that we made that would provide a solution to that question. And then we come up with hypotheses. We generate results, hopefully gathering evidence that will either support or oppose the claims we did. On the left, and pardon for looking because I'm terrible at mixing up right and left, we have what I call a research road mapping schema. When you are mapping a new area of knowledge, usually there's a goal or a question that you want to answer. And in order to answer that question, there are a certain set of desired data properties or user needs that we need to account for so we can actually accomplish that goal. And usually those properties are enabled by core technologies or approaches, and they can be limited somehow. So there are a set of new capabilities in green that we need to add to those core technologies again that will enable desired data and will answer our research question. And finally, here we have in blue the open problems that are all the questions that are problems that haven't been addressed yet. And there are somehow blocking the progress here. So without going into too much detail, and because road maps are all about collaboration, our impact is perceived through our network of impact that includes cross institutional and multidisciplinary scholars, school builders, or practitioners either developing those protocols, developing the support infrastructure, or stress testing this in practice to plan for research initiatives or the risk science. So if you have suggestions, a topic to run an experiment with, or you went to somehow support the development of the space, please reach out to us. Or if you want to know more about the initiative and innovative work we do at network codes, there are some links here. And this is a QR code that lets you download the presentation so you can have all the links and emails. Thank you for your attention. Great. Thank you Sylvia, very much. Sorry we're a bit tight for time. We've got a minute and a half for questions if anyone would like to come forward. We've got one colleague there. Yes, far away. Thank you. Jeff Alexander with RTI, the Research Triangle Institute. So I'm trying to operationalize what you're talking about because I'm sympathetic to your ontology. But road mapping, I'm coming at this from the industrial technology road mapping kind of background, usually has some consensus around the end goal. For example, it's coming back to us. It was staying on Moore's law. I think one issue we run into with science is that especially in a multidisciplinary environment, one, we use different terminology for the same questions. And two, we have a lot of disagreement about it's a somewhat adversarial kind of process of, well, which way to go is really based a lot on some theory that may not have a common body of like shared understanding or shared agreement. How does road mapping overcome that kind of adversarial disagreement and actually get consensus on a path forward? That's actually a great question because one of the things that we've been claiming as well is that we need some kind of government system or voting system to attach to these kind of mechanisms. So how do we actually shape collective roadmaps? You have your own road map. It's your vision of the Swiss air mine. How can I merge them? And of course, we need a common language, but we need to disambigwate if we are using the same term for different things. So this is one of the open problems that we've been facing. I'll do a merge them semantically. And fortunately, there's a lot of AISS tools that already happen helping the system. But until we come up with this portfolio of roadmaps that we can actually study and operate on to make decisions, we won't know for sure. It's just something that we guess. We assume that it's hard to combine them because of the implicit knowledge. But a road map, I will also claim that a road map is not a good road map if you can't navigate it without a lot of implicit knowledge. So I don't know. We can discuss it further if you want afterwards. Thank you. Great. Thank you, Sylvia, again. And thanks also for the links for people to follow up. Thanks again.