 Okay. All right. Welcome back everyone. Hope you had a good lunch. Did anyone go into town or stay here? Anyway, okay. I'm just going to give a quick introduction to this new initiative that was launched last year. It's called Open Source Science. It lives at NumFocus, which was mentioned earlier as a nonprofit in the U.S. That is the home to many well-known multi-Python projects, including Jupiter, and pandas, and NumPy, and SciPy. The goal behind this initiative is to accelerate scientific research by improving the way open source software gets done in science. There are a few surprising challenges that I'm going to get into. This is me. I'm at IBM Research. I do community. You can connect with me on LinkedIn if you'd like. So, big picture. We're facing some pretty pressing issues as a planet, as a society, including climate and diseases and many other things. And we're going to need a lot of science over the next few years, if not decades. And without a really more robust, much more robust infrastructure and foundation of open source, it's going to be much harder to make progress. And some of the challenges that we're facing include this phenomenon that people are constantly reinventing the wheel, solving the same challenges over and over. Another challenge is that we are seeing people build cool software that then just dies on the vine in some, you know, abandoned GitHub repo. There's not just in science, but in science in particular, lots of challenges around sustainability and funding of open source projects. We see that often people don't have a consistent way of getting trained and on-boarded into doing good open source. Unlike in the private sector where the people developing open source are usually software people, in science, it's usually science people. And they can have a hard time making this part of their career paths. We see gaps between domain experts and people with the computer science and software engineering skills. And the question becomes, how do you bridge those gaps? It still can be hard to pitch open source to the people that hold the purse strings. Overall, we see a lot of incentives misaligned, visa-visi, you know, ideal open source culture, and I'm sure there are many more. And so what this open source science initiative tries to do is improve things by bringing together scientists and open source developers and other stakeholders to share those practices, to identify common pain points, and to explore solutions together. So as I mentioned, we're an unfocused program launched last July. And how we work is we've set up a number of interest groups. We have three domain-specific groups, chemistry, material science, and climate sustainability, and then very soon also on life sciences and health care. And because it sits under an unfocus, anyone can suggest additional groups that should be added to the mix. We also have a couple of horizontal groups one is focused on the question of reproducible science and the other is building a map that will let you explore the interrelations between existing tools and the published research and the people involved behind both. The map of science is a pretty cool project. We're starting to actually do some prototyping and this is some data we collected at a recent conference in Greece where we just show orange here are the projects that are used by these the people attending this conference and then blue are the contributors on GitHub and GitLab. I'm just starting to see some very basic non-earth-shattering yet, but interesting relations between those projects. So you could imagine if you were to start a new science initiative, you could maybe maybe want to talk to some of the people that are involved in several of these open source projects. Other events, other activities that we do, we do events. In person we are starting to do virtual meetups probably later this year as well as new content that we're starting to publish focusing on learning about these various projects and how they operate. And so that is a very high level intro and the reason we're interested to be here is because AI of course has come up many times in these domains that we cover, including generative AI. And as I mentioned earlier, we're very interested in connecting this emerging community with the latest trends and skills in open source generative AI and to bridge those two worlds. That was a quick intro. Any questions? If you're interested, you can follow us. We're on LinkedIn, we have a newsletter, there's a website, open source.science, and expect to hear some things emerge over the next few months as people bring up these AI topics in the interest groups. I'll ask a question. Sure. Because you guys have not been there but Tim was there and I think it's really interesting. So I guess kind of what is the most typical friction points you see between scientists and open source people and how can we as developers here at the most, how can we help scientists advance solve these problems? What's a good way for people here to engage and help scientists? So one point of friction is definitely the skill gap. So you have people developing code I call them science people versus software people but of course it's a simplification but who are developing code for the purpose of let's say publishing a paper or getting a PhD. But then not having the time and resources to really make sure the code can live on in some or is set up in a way that could be for it to be continued to add value in some community or in collaboration with others beyond the publishing date of that paper. And so really trying to attract not just scientists but also people who have the software engineering, the coding skills to share their best practices and help scientists you know overcome their challenges. There are, we also are looking for examples of where people in the field are innovating and trying to find new ways of doing things. So there's with regard to managing the skill gap between or this this gap between the domain experts and the people understand software there are different approaches emerging and we're looking to you know highlight success stories and and share insights that people gather from these experiments whether it's like upskilling scientists of some degrees or putting people in the middle that can kind of mediate between those those two groups right some cases they're called research software engineers whether it's standing up a new unit at a university that can have software engineering jobs to help you know certain projects advance. So we're interested in bringing together those experiments hopefully successful experiments. Thank you. I just have a few feedback. I think I think uh scientists kind of a application you need a lot of domain knowledge for most of the software engineer they probably either forgot a long time or they did not actually um focus on that because every software engineer is very busy for their current job and things like that so I don't know I mean maybe school academic is probably having more people I mean student interest to gain to the scientific unless uh unless the software engineer is focused on this vertical area certainly they will yeah that's that's exactly the challenge right so so either you try to get the scientists to become more of a of someone who can do software things or you bring in people who understand software but then they have to first understand what the science wants to accomplish like in the specific domain right so it's not an easy easy fix but yeah so with this initiative we're trying to bring these two worlds together right so open source culture and and science and uh we'll see how that goes anyway and we just wanted to mention here because I think we're going to see more things come out specifically with regard to open source AI because that's we see that kind of starting to gain more traction yeah the I think the focus of a lot of these tools is AI so like scientists who do material science they look how can generate if I generate new materials people who do cancer treatment look how can we make new drugs and understand what's going on right so like it's it's a lot of it is AI so in scientists generally across the disciplines they want AI it's a life hate relation but they doubt it can help because like they have all the knowledge right so it's a very interesting tension they want the AI to help them but they're afraid is gonna first of all is going to waste their time right now it's very inefficient and second like nobody knows what's going to happen if they actually AI starts to behave like a scientist right so I think it's everywhere because the board so it's very very urgent all right that was it