{"version": "1.0.0", "segments": [{"startTime": 0.18, "endTime": 4.92, "body": "It will not be the artificial intelligence which will take your job very soon.", "speaker": "Speaker 1"}, {"startTime": 5.219, "endTime": 10.159, "body": "It will be another developer which is using artificial intelligence which is taking your job.", "speaker": "Speaker 1"}, {"startTime": 23.67, "endTime": 28.909, "body": "So welcome to a brand new episode of our podcast, Human and AI Minds Machine in the gradient descent.", "speaker": "Speaker 2"}, {"startTime": 29.53, "endTime": 39.009, "body": "Thanks for tuning in to our geeky podcast where we discuss the fascinating field of AI and machine learning, corporate craziness, passion for technology, and the role of humans.", "speaker": "Speaker 3"}, {"startTime": 39.159, "endTime": 50.52, "body": "We are Uli and Avri, your host for this episode, and today we're diving deep into a topic that's reshaping software development, generative AI and in particular, the GitHub co-pilot.", "speaker": "Speaker 3"}, {"startTime": 51.09, "endTime": 51.45, "body": "That's right.", "speaker": "Speaker 2"}, {"startTime": 51.49, "endTime": 66.69, "body": "And to help us navigate this kind of conversation, we have two exceptional guests for Siemens, who have been spearheading actually the rollout of the solution that may see as a game changer, actually at least a pretty cool JAI use case for Siemens for sure, right?", "speaker": "Speaker 2"}, {"startTime": 66.809, "endTime": 78.55, "body": "As it changes the day to day activities of software development, changing workflows, maybe even possibly to, you know, streamline um and take more and a different approach to the software development.", "speaker": "Speaker 2"}, {"startTime": 79.0, "endTime": 79.01, "body": "culture.", "speaker": "Speaker 2"}, {"startTime": 80.12, "endTime": 92.709, "body": "So please join us in welcoming none other than Thomas Rittenmeyer, the GitHub co-pilot, product owner, and Adrian Dischinger, a seasoned GitHub enterprise DevOps engineer, both from DIIT at Siemens.", "speaker": "Speaker 3"}, {"startTime": 93.0, "endTime": 106.059, "body": "And with their leadership, creative approach, and meticulous dedication, Thomas and Adrian are not just enhancing today's productivity, but are also paving the way for continuous innovation and efficiency in software development.", "speaker": "Speaker 3"}, {"startTime": 106.68, "endTime": 110.629, "body": "Today's discussion promises to uncover the layers of this transformation.", "speaker": "Speaker 3"}, {"startTime": 111.01, "endTime": 119.889, "body": "And throughout this episode, we're eager to dive into this journey, the strategies behind the adoption, and of course the challenges and opportunities that you've encountered.", "speaker": "Speaker 3"}, {"startTime": 121.029, "endTime": 123.05, "body": "Thomas Adrian, happy to have you today.", "speaker": "Speaker 2"}, {"startTime": 123.309, "endTime": 124.769, "body": "So welcome to the show.", "speaker": "Speaker 2"}, {"startTime": 125.23, "endTime": 126.22, "body": "Yeah, thanks for having us.", "speaker": "Adrian Dischinger"}, {"startTime": 126.55, "endTime": 126.75, "body": "Thanks", "speaker": "Adrian Dischinger"}, {"startTime": 126.75, "endTime": 127.489, "body": "for being here.", "speaker": "Speaker 1"}, {"startTime": 128.429, "endTime": 131.669, "body": "So maybe to break the ice, it's pretty cold today in Munich as well.", "speaker": "Speaker 3"}, {"startTime": 132.19, "endTime": 141.44, "body": "So maybe for those out there in the audience that don't know you yet, what are two hashtags, or should I rather say good commits that describe you best?", "speaker": "Speaker 3"}, {"startTime": 141.83, "endTime": 143.33, "body": "Maybe Thomas, do you want to start?", "speaker": "Speaker 3"}, {"startTime": 143.589, "endTime": 143.91, "body": "Let's", "speaker": "Speaker 3"}, {"startTime": 143.91, "endTime": 147.99, "body": "start with hashtag stand up again and curious.", "speaker": "Speaker 1"}, {"startTime": 148.86, "endTime": 149.949, "body": "Stand up again, you know.", "speaker": "Speaker 2"}, {"startTime": 151.97, "endTime": 153.559, "body": "Yeah, great, great input from Thomas.", "speaker": "Adrian Dischinger"}, {"startTime": 154.039, "endTime": 157.539, "body": "For myself be innovation and communication, two driving topics I maintained for a while.", "speaker": "Adrian Dischinger"}, {"startTime": 158.339, "endTime": 159.029, "body": "Uh, super cool.", "speaker": "Speaker 2"}, {"startTime": 159.119, "endTime": 162.97, "body": "Um, so, uh, innovation, what does innovation mean for you", "speaker": "Speaker 2"}, {"startTime": 162.97, "endTime": 163.119, "body": "then?", "speaker": "Adrian Dischinger"}, {"startTime": 163.21, "endTime": 172.009, "body": "Innovation is, is mainly to um innovate or the daily workflows and process at Siemens to enhance our daily tasks and basically just to reduce time to market for developers.", "speaker": "Adrian Dischinger"}, {"startTime": 172.21, "endTime": 174.19, "body": "OK, what does standing, standing up mean?", "speaker": "Speaker 2"}, {"startTime": 174.72, "endTime": 183.789, "body": "It means that in the life of a project manager, there are many obstacles to overcome and you have to stand up again after every problem you have faced.", "speaker": "Speaker 1"}, {"startTime": 184.0, "endTime": 186.369, "body": "Alright, staying resilient in the age of AI.", "speaker": "Speaker 2"}, {"startTime": 187.08, "endTime": 191.919, "body": "But let's talk about a bit of, you know, the AI specifically in software in IT environment.", "speaker": "Speaker 2"}, {"startTime": 192.07, "endTime": 194.729, "body": "Generative AI has taken the world by storm, right?", "speaker": "Speaker 2"}, {"startTime": 194.85, "endTime": 196.119, "body": "Shop floor to top floor.", "speaker": "Speaker 2"}, {"startTime": 196.199, "endTime": 201.66, "body": "Everybody's talking about co-pilots, everybody's talking about how to augment their own task, right?", "speaker": "Speaker 2"}, {"startTime": 202.35, "endTime": 209.88, "body": "What's, how do you see the role of of Gen AI and generative AI technology in software development in the life cycle?", "speaker": "Speaker 2"}, {"startTime": 210.19, "endTime": 220.649, "body": "How do you see it currently in the software process maybe and where do you see greater opportunities in involving the large language models, the LLM stack more, more increasingly?", "speaker": "Speaker 2"}, {"startTime": 221.44, "endTime": 239.08, "body": "Um, yeah, currently I think that, um, software development is maybe one of the areas from Gen AI where it shows the most potential and also seeing at the latest developments, um, in Gen AI for coding systems, it's very promising.", "speaker": "Speaker 1"}, {"startTime": 239.16, "endTime": 243.52, "body": "There are new reasoning models out there which um have.", "speaker": "Speaker 1"}, {"startTime": 244.139, "endTime": 247.55, "body": "Similar capabilities as humans already.", "speaker": "Speaker 1"}, {"startTime": 247.82, "endTime": 253.369, "body": "The electricity is the electricity needs for those reasoning models is very high.", "speaker": "Speaker 1"}, {"startTime": 253.699, "endTime": 261.76, "body": "So currently the costs of those models is far bigger than the humans, but there's a big potential in that area.", "speaker": "Speaker 1"}, {"startTime": 262.72, "endTime": 283.38, "body": "And also it's not only the writing the code where the Chen AI can support, it's also the whole environment around it, where I think like from the infrastructure and getting, getting bigger, leaving just the the the coding stack from my perspective.", "speaker": "Speaker 1"}, {"startTime": 284.029, "endTime": 299.109, "body": "So if you, if you look on a bit of the on the software development process, I guess, you know, you can see that there is a bit of the requirement specification aspects, I guess, then, then you have to actually maybe the softer, softer boiler plating, I, I guess, which can be done, right?", "speaker": "Speaker 2"}, {"startTime": 299.19, "endTime": 306.149, "body": "There should be a bit of a testing regime, I guess, which should be, you know, um, and maybe then a bit of the CIC.", "speaker": "Speaker 2"}, {"startTime": 306.274, "endTime": 308.195, "body": "The pipeline along the line, right?", "speaker": "Speaker 2"}, {"startTime": 308.295, "endTime": 318.994, "body": "Um, do you see, do you feel already that, you know, each part of that step has been, has been somehow innovated, pushed through, throughout, um, using AI technology?", "speaker": "Speaker 2"}, {"startTime": 319.165, "endTime": 319.315, "body": "Yeah,", "speaker": "Speaker 2"}, {"startTime": 319.394, "endTime": 332.029, "body": "definitely from my own perspective on experience as a developer, engineer, using the type of tools and mainly we do see those enhancements, basically the um The overarching picture is the the growth of the context windows.", "speaker": "Adrian Dischinger"}, {"startTime": 332.119, "endTime": 338.64, "body": "That means the inclusion of more and more context of more and more as you mean inputs basically um goes over time.", "speaker": "Adrian Dischinger"}, {"startTime": 338.739, "endTime": 347.739, "body": "You had the source code, as Thomas mentioned, as a as a start, and then when we grow from there to include CIC tool inputs, their results basically, and to just enhance.", "speaker": "Adrian Dischinger"}, {"startTime": 348.01, "endTime": 352.019, "body": "The um the inputs that you give to the models your prompts are enhanced by those inputs.", "speaker": "Adrian Dischinger"}, {"startTime": 352.609, "endTime": 354.16, "body": "This happens mostly in the IDE.", "speaker": "Adrian Dischinger"}, {"startTime": 354.269, "endTime": 357.22, "body": "So for example, digital code that we use as a primary target.", "speaker": "Adrian Dischinger"}, {"startTime": 357.769, "endTime": 362.01, "body": "There has been a lot of hype around the notion of the industrial co-pilot for engineering.", "speaker": "Speaker 3"}, {"startTime": 362.25, "endTime": 366.869, "body": "So it's, it's everywhere at every fair, we we see it and we hear lots about it.", "speaker": "Speaker 3"}, {"startTime": 367.32, "endTime": 373.98, "body": "And what it does is it leverages AI to help automation experts crank out code way faster.", "speaker": "Speaker 3"}, {"startTime": 374.369, "endTime": 376.73, "body": "So Thomas and Adrian, I'm curious to hear your thoughts.", "speaker": "Speaker 3"}, {"startTime": 376.809, "endTime": 382.97, "body": "So how do you stack up the industrial co-pilot for engineering against something like the GitHub copilot?", "speaker": "Speaker 3"}, {"startTime": 383.049, "endTime": 385.89, "body": "So what are the overlaps or the differences that you see?", "speaker": "Speaker 3"}, {"startTime": 386.94, "endTime": 398.059, "body": "Yeah, so from my own perspective, as I mentioned, I'm a development engineer, so I provide tools for um developers basically, their platforms and apps, and we do see that the co-pilot is a strong focus on the software development.", "speaker": "Adrian Dischinger"}, {"startTime": 398.179, "endTime": 404.38, "body": "So that means in IDEs in those in development environments, and the industrial co-pilot is a different focus.", "speaker": "Adrian Dischinger"}, {"startTime": 404.42, "endTime": 406.279, "body": "It's more on the shop floor.", "speaker": "Adrian Dischinger"}, {"startTime": 406.829, "endTime": 421.559, "body": "Enhancing your workflows, your, your products basically by um combining insights from Siemens and also our partner Microsoft, which is the provider of the GitHubtrial industrial co-pilot and also the owner of GitHub, um, but it's a different focus, that's that's that's for sure, yeah.", "speaker": "Adrian Dischinger"}, {"startTime": 422.929, "endTime": 447.69, "body": "So it's, there's a lot of co-pilots in the current times and you know, it's speaking on, on typical capabilities that generative I provides and one can see that in a lot of cases, you know, the translation of, you know, let's say natural language input to specific output, right, in terms of I guess encoding environment, you know, saying what do you want to code and then it produces.", "speaker": "Speaker 2"}, {"startTime": 448.13, "endTime": 450.519, "body": "I don't know, Python, C++, Java, right?", "speaker": "Speaker 2"}, {"startTime": 450.529, "endTime": 457.149, "body": "And we see the copilots for, I don't know, industrial corpolets for the ST SCL for the automation code.", "speaker": "Speaker 2"}, {"startTime": 457.29, "endTime": 462.769, "body": "We see the copilot for in teams, I don't know, summarizing our messages, I guess.", "speaker": "Speaker 2"}, {"startTime": 462.809, "endTime": 469.41, "body": "So we see it in, in, I don't know, in in Power BI and others visualization and creating dashboards, right?", "speaker": "Speaker 2"}, {"startTime": 469.769, "endTime": 476.179, "body": "And so what makes it, what makes it really somehow unique in terms of in the software development?", "speaker": "Speaker 2"}, {"startTime": 476.42, "endTime": 486.17, "body": "Why, why should, should Siemens care, you know, using these kinds of capabilities in, in, in Python, C++ and stuff, right?", "speaker": "Speaker 2"}, {"startTime": 486.29, "endTime": 486.299, "body": "Yeah", "speaker": "Speaker 2"}, {"startTime": 486.94, "endTime": 504.109, "body": "I think because Siemens is one of the biggest software development companies out there and for that is crucial for us to Their lead, lead, lead the topic and implement those changes in a, in a sensible way.", "speaker": "Speaker 1"}, {"startTime": 504.269, "endTime": 508.269, "body": "So having all the rules, having all the rules checked and so on.", "speaker": "Speaker 1"}, {"startTime": 508.63, "endTime": 518.33, "body": "But yeah, Siemens is a software development company and I think for us it's crucial to optimize our process to deliver the best products also to our customers.", "speaker": "Speaker 1"}, {"startTime": 519.818, "endTime": 523.689, "body": "So then there is this, this notion, everybody's talked about GitHub, right?", "speaker": "Speaker 2"}, {"startTime": 523.778, "endTime": 527.689, "body": "GitHubb is the community, um I think acquired by Microsoft, but it hasn't been, right?", "speaker": "Speaker 2"}, {"startTime": 527.789, "endTime": 538.439, "body": "So uh with, you know, I guess everyone maybe outside in the listeners here, right, if you, you probably have your own GitHub, you know, store with private, you know, code base in there.", "speaker": "Speaker 2"}, {"startTime": 538.848, "endTime": 545.888, "body": "So on top of that lies then a certain kinds of capability which is called GitHub Copilot, right?", "speaker": "Speaker 2"}, {"startTime": 545.968, "endTime": 548.109, "body": "Can you a bit of summarize what, what is this?", "speaker": "Speaker 2"}, {"startTime": 548.669, "endTime": 549.989, "body": "Yeah, it's good to mention that.", "speaker": "Adrian Dischinger"}, {"startTime": 550.07, "endTime": 552.229, "body": "Um, so both products have a very similar name.", "speaker": "Adrian Dischinger"}, {"startTime": 552.39, "endTime": 567.57, "body": "They have both the GitHub, the company name and the brand and the product name, but Github Copilot as a product is in our version, Gitb Copied for business or the client-based version is independent from GitHb Enterprise server or Gitup.com, which is the public one used by most of our listeners probably.", "speaker": "Adrian Dischinger"}, {"startTime": 568.34, "endTime": 572.099, "body": "So it's understand that both are independent, but um they interconnect very well.", "speaker": "Adrian Dischinger"}, {"startTime": 572.229, "endTime": 581.77, "body": "So, um, as they're both from the same company, they are very linked together to, for example, Let's say you would create a pull request, so a change request on your GitHub corporate repository.", "speaker": "Adrian Dischinger"}, {"startTime": 582.049, "endTime": 588.45, "body": "This would fetch information from GitHub into your local IDE to enhance your workflow, your creation of the contents basically.", "speaker": "Adrian Dischinger"}, {"startTime": 588.53, "endTime": 595.849, "body": "So it's, it's connected by from the maintenance from the for the product, it's independent, but the, the interaction is there, of course, yeah.", "speaker": "Adrian Dischinger"}, {"startTime": 596.21, "endTime": 596.219, "body": "Yeah.", "speaker": "Speaker 1"}, {"startTime": 597.33, "endTime": 602.789, "body": "And how did, I mean, probably not everybody out there is a software developer.", "speaker": "Speaker 1"}, {"startTime": 603.13, "endTime": 606.63, "body": "So the IDE is the tool where the code is written.", "speaker": "Speaker 1"}, {"startTime": 606.89, "endTime": 613.83, "body": "And um so there are many, many other GPT models which can probably support with software development.", "speaker": "Speaker 1"}, {"startTime": 614.95, "endTime": 623.309, "body": "But the GitHub co-pilot offers the possibility as it's directly integrated in the development tool to give a seamless experience.", "speaker": "Speaker 1"}, {"startTime": 623.469, "endTime": 631.609, "body": "So in the past, the developer, for example, he was checking for a solution on Google or Stack Overflow or somewhere.", "speaker": "Speaker 1"}, {"startTime": 632.929, "endTime": 638.809, "body": "And this takes time and brings you out of the loop, but now the developer can stay in his IDE.", "speaker": "Speaker 1"}, {"startTime": 638.929, "endTime": 651.169, "body": "He can do the questions to the GitHub co-pilot chat functionality and at the same time have a has a functionality to autocomplete like the code he's writing in the moment.", "speaker": "Speaker 1"}, {"startTime": 651.53, "endTime": 656.5, "body": "So Thomas, that mean no 20 tops open and 7 tops on stack overflow.", "speaker": "Speaker 2"}, {"startTime": 656.71, "endTime": 661.609, "body": "How, how to write the UI 911.", "speaker": "Speaker 2"}, {"startTime": 661.719, "endTime": 666.989, "body": "Point to one important thing to understand, um, of course you could still use, let's say for example, chat.", "speaker": "Adrian Dischinger"}, {"startTime": 667.13, "endTime": 672.59, "body": "Siemens.com, the internal chat platform at Siemens, but you would miss the, the filtering basically.", "speaker": "Adrian Dischinger"}, {"startTime": 672.69, "endTime": 678.229, "body": "So we employ a filter on those results to make sure that the results are compliant with Siemens policy.", "speaker": "Adrian Dischinger"}, {"startTime": 678.309, "endTime": 684.539, "body": "So for example, You might not be able to use open source um results um based on your your coding product.", "speaker": "Adrian Dischinger"}, {"startTime": 684.739, "endTime": 685.919, "body": "So you would have to filter those.", "speaker": "Adrian Dischinger"}, {"startTime": 685.979, "endTime": 691.469, "body": "That's something that's very important to understand this component is employed on the co-pilot before you get the results.", "speaker": "Adrian Dischinger"}, {"startTime": 693.719, "endTime": 695.559, "body": "Maybe a bit of, you know, expectation.", "speaker": "Speaker 2"}, {"startTime": 695.869, "endTime": 706.229, "body": "Why should a company utilize Generative I in terms of either Codex or let's say the GitHub copilot, right, for the software development?", "speaker": "Speaker 2"}, {"startTime": 706.479, "endTime": 710.76, "body": "What was the main trigger point that someone said like in Siemens, let's do that.", "speaker": "Speaker 2"}, {"startTime": 711.58, "endTime": 725.25, "body": "I mean, when I, when I spoke to the to the managers, um we always said, we can provide you immediately with improved documentation, for example, so in, in the, in the past let's say.", "speaker": "Speaker 1"}, {"startTime": 725.929, "endTime": 741.179, "body": "Documentation was always the topic which, oh, we are close to production ready and we need to fix our stuff and we do the documentation later, then new, new issues arise and probably the documentation was not, uh, was never proper finished.", "speaker": "Speaker 1"}, {"startTime": 741.609, "endTime": 747.969, "body": "And with GitHub co-pilot, you could start like supporting the creation of the documentation.", "speaker": "Speaker 1"}, {"startTime": 748.359, "endTime": 755.409, "body": "Also the unit tests, like automated tests to check if your functionality still works after some, some tests have implied.", "speaker": "Speaker 1"}, {"startTime": 755.9, "endTime": 762.33, "body": "So we started like with functionality, which raises the quality, but um.", "speaker": "Speaker 1"}, {"startTime": 763.969, "endTime": 768.01, "body": "Which, which raises the quality of the product and has not been done in the past.", "speaker": "Speaker 1"}, {"startTime": 768.159, "endTime": 775.33, "body": "So this is like an easy, easy, easy to calculate and additional productivity when you have not done it in the past.", "speaker": "Speaker 1"}, {"startTime": 776.46, "endTime": 781.979, "body": "So the current software development by, you know, utilizing, you know, automatically.", "speaker": "Speaker 2"}, {"startTime": 782.109, "endTime": 784.429, "body": "Documenting, I love documenting my code, right?", "speaker": "Speaker 2"}, {"startTime": 784.63, "endTime": 785.789, "body": "Probably not, right?", "speaker": "Speaker 2"}, {"startTime": 785.869, "endTime": 787.429, "body": "And everybody hates it, right?", "speaker": "Speaker 2"}, {"startTime": 787.549, "endTime": 790.27, "body": "But it's automatically generating the documentation of the code.", "speaker": "Speaker 2"}, {"startTime": 790.349, "endTime": 792.02, "body": "What does, does it actually do?", "speaker": "Speaker 2"}, {"startTime": 792.07, "endTime": 793.82, "body": "What, you know, what's the purpose for that?", "speaker": "Speaker 2"}, {"startTime": 794.03, "endTime": 799.77, "body": "What's the role of the function and what are the parameters maybe, right, to be set up as well as for the unit testing.", "speaker": "Speaker 2"}, {"startTime": 799.82, "endTime": 800.21, "body": "OK.", "speaker": "Speaker 2"}, {"startTime": 800.549, "endTime": 803.58, "body": "That sounds about easy, OK, just install it, is it?", "speaker": "Speaker 2"}, {"startTime": 803.919, "endTime": 817.15, "body": "It is really plug and play, yes, so you just um log in with your account and then you use one of the general IDEs Visual Studio or something, and then it comes as a plug-in.", "speaker": "Speaker 1"}, {"startTime": 817.27, "endTime": 819.039, "body": "So it's very easy to set up.", "speaker": "Speaker 1"}, {"startTime": 819.109, "endTime": 824.539, "body": "There is no, no big setup involved except of username, password or.", "speaker": "Speaker 1"}, {"startTime": 825.369, "endTime": 827.159, "body": "Like the Siemens ID.", "speaker": "Speaker 1"}, {"startTime": 827.52, "endTime": 847.549, "body": "But uh in terms of, let's say, you know, OK, understood there is a cool capabilities, we want to augment our folks, obviously, hopefully every code is now, you know, developed in Siemens with some kinds of Gin AI capability, which is pretty cool, faster and more quality, but rolling out a software tool like that.", "speaker": "Speaker 2"}, {"startTime": 848.08, "endTime": 851.309, "body": "In a large corporation isn't, isn't that easy, is it?", "speaker": "Speaker 2"}, {"startTime": 851.44, "endTime": 851.52, "body": "Right?", "speaker": "Speaker 2"}, {"startTime": 851.64, "endTime": 857.239, "body": "Can you guide us a bit of the journey, right, in rolling out the GitHub Copolate, how do you approach it?", "speaker": "Speaker 2"}, {"startTime": 857.39, "endTime": 863.08, "body": "Where are some reactions, where there some risks or with some blockers, you know.", "speaker": "Speaker 2"}, {"startTime": 864.229, "endTime": 872.88, "body": "Yeah, I, I also, I, I took over the topic from Ron Davin Stetzer, so he, he did one of the tasks in the beginning.", "speaker": "Speaker 1"}, {"startTime": 873.5, "endTime": 894.2, "body": "But yeah, how our approach was first checking if all the, all the different governance topics are fulfilled, if the if the legal side, the the export control topics, the data protection, if all those things, if, if the solution fulfills the demands in that area.", "speaker": "Speaker 1"}, {"startTime": 895.2, "endTime": 904.4, "body": "Then we did many different pilot phases, first with like random code, then internal code, and the latest than the productive code.", "speaker": "Speaker 1"}, {"startTime": 904.71, "endTime": 911.299, "body": "You don't want to start with your productive code, of course, you want to make sure that everything is fine before.", "speaker": "Speaker 1"}, {"startTime": 911.679, "endTime": 920.859, "body": "Then we checked the code for duplicates, so we, we, we, we set up one pilot phase where all the code which was created.", "speaker": "Speaker 1"}, {"startTime": 921.15, "endTime": 950.809, "body": "Is going to um check if there are code applications or if there are um copies or if somebody copied like code from stack Overflow which we which we just mentioned, and yeah, then we also did a scientific study together with foundational technologies, they did a deep investigation and after we saw it gives us additional productivity and all the governance topics are clarified, then we set up like.", "speaker": "Speaker 1"}, {"startTime": 951.619, "endTime": 959.84, "body": "An environment where we could have up to 2000 users and like those 2000 users have been allocated overnight.", "speaker": "Speaker 1"}, {"startTime": 960.84, "endTime": 978.419, "body": "The demand was very huge and then we stick to those 2000 and at one point we presented everything to the DI management, what you had, the additional productivity and the DI management, they did a bold move and decided to get licenses for all software developers within DI.", "speaker": "Speaker 1"}, {"startTime": 979.239, "endTime": 980.599, "body": "That's a huge achievement.", "speaker": "Speaker 3"}, {"startTime": 980.88, "endTime": 982.5, "body": "Congratulations on that.", "speaker": "Speaker 3"}, {"startTime": 982.96, "endTime": 989.419, "body": "And it seems like it was a pretty big endeavor that you started there to roll that out and you started small with the pilots and.", "speaker": "Speaker 3"}, {"startTime": 989.78, "endTime": 995.95, "body": "And not with the productive code and doing all the data cleansing at first until you really rolled it out.", "speaker": "Speaker 3"}, {"startTime": 996.239, "endTime": 1005.02, "body": "And I guess once you encounter on such a journey, there must be also quite some obstacles on the way and must be sometimes tricky and there must be some setbacks.", "speaker": "Speaker 3"}, {"startTime": 1005.08, "endTime": 1009.75, "body": "So maybe you want to share a little bit what what hurdles did you experience along the way.", "speaker": "Speaker 3"}, {"startTime": 1010.039, "endTime": 1015.64, "body": "I think one of one of the biggest hurdles was regarding the the legal clarifications.", "speaker": "Speaker 1"}, {"startTime": 1016.08, "endTime": 1024.709, "body": "So the AI, the technology has evolved very quickly and the legislation to that not so um.", "speaker": "Speaker 1"}, {"startTime": 1025.489, "endTime": 1039.55, "body": "There is, I mean, it's, it's no legal advice, but so the outcome of, of the generative AI is most, there, there are some issues with copywriting that, and of course, as a company, you want to save your copyright.", "speaker": "Speaker 1"}, {"startTime": 1040.05, "endTime": 1058.489, "body": "So we needed to set up a rule set to make sure we can keep that copyright, and yeah, the users were not happy with it because like the sea the seamless work which was promised was not possible with that strict rule set.", "speaker": "Speaker 1"}, {"startTime": 1059.01, "endTime": 1067.53, "body": "So like we did many iterations with the legal team to improve our rules to still make sure that we can have the the copyright.", "speaker": "Speaker 1"}, {"startTime": 1067.859, "endTime": 1086.88, "body": "Um, we have been in discussions with politicians and from the working group of artificial intelligence, and they said they also see the need that on the copyright side there needs to be some change, even though they don't want to make very hard restrictions within AI.", "speaker": "Speaker 1"}, {"startTime": 1087.219, "endTime": 1093.04, "body": "They see the copyright point it's something where they need to work on and.", "speaker": "Speaker 1"}, {"startTime": 1093.619, "endTime": 1105.42, "body": "Yeah, we are still waiting for the adoption of the legislation, but now we have set a rule set where like the developers can work and the legal stuff is addressed.", "speaker": "Speaker 1"}, {"startTime": 1105.739, "endTime": 1111.969, "body": "So that means in principle, right, so the the the model being used is trained on data.", "speaker": "Speaker 2"}, {"startTime": 1112.02, "endTime": 1117.51, "body": "The data is being used from from the GitHub repositories basically on on.", "speaker": "Speaker 2"}, {"startTime": 1117.569, "endTime": 1119.28, "body": "Code base, right, that is contributed.", "speaker": "Speaker 2"}, {"startTime": 1119.489, "endTime": 1139.89, "body": "Based on that, the model has learned certain kinds of vocabulary, certain kinds of grammars along the line of C, Python, and so on, whatever, you know, program language and with that, obviously as a corporate, we, we gain into a risk because obviously we don't want to run into an IP infringement, copyright infringement, right?", "speaker": "Speaker 2"}, {"startTime": 1140.239, "endTime": 1141.349, "body": "Is that so.", "speaker": "Speaker 2"}, {"startTime": 1141.989, "endTime": 1151.79, "body": "What, how do you see the role of, of a provider of a solution, in this case, I guess Github and what's, why do we need to also have it on our side somehow?", "speaker": "Speaker 2"}, {"startTime": 1152.109, "endTime": 1153.829, "body": "And what can you, can you give it a bit of example?", "speaker": "Speaker 2"}, {"startTime": 1153.91, "endTime": 1158.26, "body": "How do we test that things that we're not running in this, in this IP infringements?", "speaker": "Speaker 2"}, {"startTime": 1158.39, "endTime": 1158.55, "body": "Yeah,", "speaker": "Speaker 2"}, {"startTime": 1158.869, "endTime": 1171.569, "body": "let's let me just, um, I mentioned those filters that we have those filters are mainly on GitHub, but additional layers are also on Sieman's side to scan code, just as an additional layer before we ship the product, basically before the market.", "speaker": "Adrian Dischinger"}, {"startTime": 1171.989, "endTime": 1174.729, "body": "We do have those code scans, um, Thomas mentioned them before.", "speaker": "Adrian Dischinger"}, {"startTime": 1175.319, "endTime": 1180.069, "body": "Just keep in mind those, those, those scans for now in the current market state, they don't really run in real time.", "speaker": "Adrian Dischinger"}, {"startTime": 1180.15, "endTime": 1187.93, "body": "So we, we, um, we try to run them as often as possible, at least before we ship, of course, but as often as possible during during development.", "speaker": "Adrian Dischinger"}, {"startTime": 1188.469, "endTime": 1194.54, "body": "And in the long term we of course test and and um and talk to vendors and to have those tools running in real time in the IDE.", "speaker": "Adrian Dischinger"}, {"startTime": 1194.63, "endTime": 1195.79, "body": "So in the, in the tooling.", "speaker": "Adrian Dischinger"}, {"startTime": 1196.329, "endTime": 1199.729, "body": "To just reduce the time to market by scaling those on a higher basis.", "speaker": "Adrian Dischinger"}, {"startTime": 1199.88, "endTime": 1208.189, "body": "But right now, the, the most common vendors can't provide this higher scalability to do this on-prem, on, on all the time on demand.", "speaker": "Adrian Dischinger"}, {"startTime": 1209.319, "endTime": 1214.0, "body": "And there's, there's like the, the two sides of the metal.", "speaker": "Speaker 1"}, {"startTime": 1214.64, "endTime": 1224.319, "body": "So one is that we don't infringe the copyright of somebody else, and one is that we can really save the copyright for ourselves.", "speaker": "Speaker 1"}, {"startTime": 1224.959, "endTime": 1230.02, "body": "So it's a small explanation, when you have an AI which generates a picture, then you.", "speaker": "Speaker 1"}, {"startTime": 1230.829, "endTime": 1246.449, "body": "Give there a prompt and then you get a picture and of course you don't have the copyright to it, you just enter it like a very short prompt, but then um in GitHub copilot, it can be like a very sophisticated prompt, a very long prompt and.", "speaker": "Speaker 1"}, {"startTime": 1247.27, "endTime": 1265.79, "body": "It's interesting that in the current current legislation, it doesn't make a difference because it's always the output which you want to save the copyright to and so you need to somehow change the, the result so that it becomes yours and", "speaker": "Speaker 1"}, {"startTime": 1266.109, "endTime": 1269.339, "body": "this is like in addition to that.", "speaker": "Speaker 2"}, {"startTime": 1270.209, "endTime": 1275.729, "body": "And I, I read that Github also has a bit of a let's say a check box and like, you know, they.", "speaker": "Speaker 2"}, {"startTime": 1276.05, "endTime": 1283.91, "body": "They, they agreed or they, they, they commit to that this code is free of IP, right?", "speaker": "Speaker 2"}, {"startTime": 1284.369, "endTime": 1290.449, "body": "Um, so that means Siemens embraces a double check, a triple check before going into productive somehow.", "speaker": "Speaker 2"}, {"startTime": 1290.53, "endTime": 1290.81, "body": "Is that", "speaker": "Speaker 2"}, {"startTime": 1290.81, "endTime": 1293.869, "body": "it's it's, it's um it's.", "speaker": "Speaker 1"}, {"startTime": 1294.025, "endTime": 1301.055, "body": "That we don't infringe the copyright of others and this is where that functionality which you just mentioned comes into place.", "speaker": "Speaker 1"}, {"startTime": 1301.385, "endTime": 1308.415, "body": "But like having, having the, the results being copyrighted as yourself is something different.", "speaker": "Speaker 1"}, {"startTime": 1308.625, "endTime": 1311.125, "body": "So their GitHubb cannot really help you.", "speaker": "Speaker 1"}, {"startTime": 1313.02, "endTime": 1321.29, "body": "But GitHub makes a lot of efforts to make sure that your code has not parts of other codes included, yeah.", "speaker": "Speaker 1"}, {"startTime": 1322.31, "endTime": 1329.609, "body": "Yeah, it's also good to understand, basically, as, as Thomas has mentioned, the um the later part of the copyright that happens on the client on the IEDE by the by the user.", "speaker": "Adrian Dischinger"}, {"startTime": 1329.67, "endTime": 1338.869, "body": "So basically by modifying those, those things, we then are compliant with the legal framework to say, OK, modification is done now you're the one who drives, but also you have the responsibility.", "speaker": "Adrian Dischinger"}, {"startTime": 1338.91, "endTime": 1343.91, "body": "It's also important to note that the user has to stay in in the pilot's seat in the in the driver's seat to um.", "speaker": "Adrian Dischinger"}, {"startTime": 1344.53, "endTime": 1353.839, "body": "You can't just lean back and kick kick off the the the audio and do the co-pilot and does the work for you, you have to stay in the driver's seat as of course, the main responsibility for it.", "speaker": "Adrian Dischinger"}, {"startTime": 1354.579, "endTime": 1373.099, "body": "You just mentioned a very important aspect, so the user and I, I wonder about like 1st, 1st of all, did you like conduct any like user research before you um rolled out out the tool to better understand like what, what is it that the users really need and how do they want to use it?", "speaker": "Speaker 3"}, {"startTime": 1373.28, "endTime": 1380.609, "body": "And then on the other hand, as as you Adrian just mentioned, it's also about that the user remains in the driver's seat and is responsible.", "speaker": "Speaker 3"}, {"startTime": 1380.719, "endTime": 1389.02, "body": "So how do you facilitate that adoption and did you feel any pushback at first or how, how did that, did that go?", "speaker": "Speaker 3"}, {"startTime": 1390.4, "endTime": 1417.469, "body": "So yeah, we, we had different ways about how the how the um How the user could reach out to us in case they have a problem, then we set up a business partner support structure, so like from FA from every business unit or from every business, there have been a contact person, so the the users went then to their business partners, they collected the information, gave the information to us.", "speaker": "Speaker 1"}, {"startTime": 1418.569, "endTime": 1438.969, "body": "And um yeah, during that scientific study which I, which I just mentioned, yeah, we had, we got very deep insights, we had 6 teams which we, which we interviewed over 8 weeks and we make really in-depth interviews every week with every single person and there we got most of our insights.", "speaker": "Speaker 1"}, {"startTime": 1439.02, "endTime": 1442.729, "body": "And later on, yeah, now we have set up our champions community.", "speaker": "Speaker 1"}, {"startTime": 1443.109, "endTime": 1465.91, "body": "So everybody can volunteer and be a champion, gets the latest information from us um regarding GitHubb co-pilot, and we are hoping that they act as an ambassador and also yeah, spread, spread the news and also give like small help and if, if, if the small help is not enough, giving us the information so that we can support there.", "speaker": "Speaker 1"}, {"startTime": 1467.439, "endTime": 1478.64, "body": "Can you, can you share a bit of, you know, in the early, early days of, you know, the, the test, the test period, let's say, you know, a bit of systematics, what kind of insights have been drawn.", "speaker": "Speaker 2"}, {"startTime": 1478.8, "endTime": 1484.56, "body": "I remember it also because we were also in my organization part as we have also back in the days, 10,000 developers.", "speaker": "Speaker 2"}, {"startTime": 1484.64, "endTime": 1494.88, "body": "So, you know, I was driving and we were driving as well and and trying to support, you know, the first testing out there and I remember, you know, we had to commit back and say like, OK, AI generated.", "speaker": "Speaker 2"}, {"startTime": 1495.75, "endTime": 1496.829, "body": "in these kinds of aspects.", "speaker": "Speaker 2"}, {"startTime": 1497.06, "endTime": 1506.04, "body": "Can you, can you guide the audience a bit through from, you know, how is the setup, what are the typical focus use cases maybe were, right?", "speaker": "Speaker 2"}, {"startTime": 1506.65, "endTime": 1522.829, "body": "And then what kind of impact criteria you were looking for, asking for, and, and obviously one thing obviously the audience is interested, what you know, any number would be nice to hear, you know, what's what's the impact on the business side or on the on the on the people side or on the personal side, maybe, right?", "speaker": "Speaker 2"}, {"startTime": 1523.78, "endTime": 1525.67, "body": "Yeah, so we just start from the beginning.", "speaker": "Adrian Dischinger"}, {"startTime": 1525.75, "endTime": 1542.0, "body": "So the whole process, the on off boarding is is highly automated, which is a big step for us as a Siemens to roll out to be highly automated and then have just basically a form to use um the automation process that cycles back to focus on business partners if required for further clarification.", "speaker": "Adrian Dischinger"}, {"startTime": 1542.31, "endTime": 1549.829, "body": "So that that means if someone is interested in testing it out, there is a workflow that can be triggered and then, you know.", "speaker": "Speaker 2"}, {"startTime": 1550.145, "endTime": 1552.814, "body": "It's not easy in large corporation to get a software, right?", "speaker": "Speaker 2"}, {"startTime": 1552.944, "endTime": 1552.954, "body": "It,", "speaker": "Speaker 2"}, {"startTime": 1554.824, "endTime": 1559.094, "body": "use just the overhead costs and doing rollout and and offboarding, so user management.", "speaker": "Adrian Dischinger"}, {"startTime": 1559.515, "endTime": 1568.984, "body": "And from there on, once you're onboarded, we have, of course, surveys and and and and forums to to provide feedback and um have those those chances to provide insights.", "speaker": "Adrian Dischinger"}, {"startTime": 1568.994, "endTime": 1577.775, "body": "And based on numbers you just mentioned, we in doing this study, we see we've seen about 10% increase in productivity during our study that Thomas mentioned earlier in the talk.", "speaker": "Adrian Dischinger"}, {"startTime": 1578.31, "endTime": 1589.719, "body": "And um we still see very positive feedback in terms of users providing us the um They just reduce the time in learning and research and of course have um more time for creativity.", "speaker": "Adrian Dischinger"}, {"startTime": 1589.829, "endTime": 1607.89, "body": "So um the, the daily tasks, the documentation we mentioned earlier, those repetitive tasks, they are very um very much faster than before and you have more time for your your structure, your which is having driver's seat basically to drive the the development without having to spend too much time on the site for your your jaws basically.", "speaker": "Adrian Dischinger"}, {"startTime": 1609.28, "endTime": 1620.949, "body": "So that That means the, the task of boilerplate generation and documentation generation has been contributing to good quality, I guess then, right?", "speaker": "Speaker 2"}, {"startTime": 1621.349, "endTime": 1624.51, "body": "Um, and to, to productivity by roughly 10%.", "speaker": "Speaker 2"}, {"startTime": 1624.56, "endTime": 1626.869, "body": "That's, that's a big rocket, is it?", "speaker": "Speaker 2"}, {"startTime": 1627.569, "endTime": 1631.489, "body": "That's for, you know, we have a couple of thousands of software developers,", "speaker": "Speaker 2"}, {"startTime": 1631.56, "endTime": 1631.699, "body": "right?", "speaker": "Speaker 1"}, {"startTime": 1631.79, "endTime": 1639.91, "body": "It's it's a, it's a huge leap and also um during, during the, during the study, you saw like that there is a learn curve.", "speaker": "Speaker 1"}, {"startTime": 1640.05, "endTime": 1648.68, "body": "So in the, in the beginning, probably starting with 5% in the end, ending up with 15% and then.", "speaker": "Speaker 1"}, {"startTime": 1648.859, "endTime": 1651.02, "body": "You come to an average of 10%.", "speaker": "Speaker 1"}, {"startTime": 1651.42, "endTime": 1656.849, "body": "But if you would do that study for much longer than probably the percentage would also raise.", "speaker": "Speaker 1"}, {"startTime": 1657.06, "endTime": 1669.8, "body": "Important is that to say to that 10% is always the developer doesn't spend 100% of his time developing, and those 10% time saving is with development tasks.", "speaker": "Speaker 1"}, {"startTime": 1670.469, "endTime": 1677.439, "body": "And in those surveys, so you, you measured productivity, how do the developers, how did they accept that new tool?", "speaker": "Speaker 3"}, {"startTime": 1677.55, "endTime": 1687.15, "body": "So are there, do they feel more, more satisfied now that they're using it it was there a lot of resistance at first or did they have difficulties to really on board on it?", "speaker": "Speaker 3"}, {"startTime": 1687.219, "endTime": 1690.81, "body": "Did you have like particular training strategies that you employed?", "speaker": "Speaker 3"}, {"startTime": 1691.31, "endTime": 1700.05, "body": "I think for for 80, 80% or something where the tool works without any restrictions, the.", "speaker": "Speaker 1"}, {"startTime": 1700.489, "endTime": 1703.479, "body": "Developers are in general very happy.", "speaker": "Speaker 1"}, {"startTime": 1704.359, "endTime": 1718.939, "body": "So that those tasks are taken from them and yeah, for some reason, once we needed to remove some licenses, and yeah, we got immediately feedback, oh I need it again, how do I get it back?", "speaker": "Speaker 1"}, {"startTime": 1719.16, "endTime": 1727.64, "body": "And this was like not good that we took in the licenses, but it was good to get that feedback that it really helps to the people.", "speaker": "Speaker 1"}, {"startTime": 1728.02, "endTime": 1749.5, "body": "And for that other 20% where the solution probably does not really work as expected, so it can be specialized languages or really huge, huge code bases or something like this, and we are collecting the feedback and also discussing it with Microsoft, where we have a very strong partnership together.", "speaker": "Speaker 1"}, {"startTime": 1749.959, "endTime": 1755.349, "body": "And yeah, so that we can reach them closer to the 100%.", "speaker": "Speaker 1"}, {"startTime": 1755.93, "endTime": 1766.869, "body": "Is there a difference between if you're a very senior developer and or a very junior developer, who, who gets more value", "speaker": "Speaker 2"}, {"startTime": 1766.869, "endTime": 1766.969, "body": "out", "speaker": "Adrian Dischinger"}, {"startTime": 1766.969, "endTime": 1767.089, "body": "of", "speaker": "Speaker 2"}, {"startTime": 1767.089, "endTime": 1767.199, "body": "it?", "speaker": "Adrian Dischinger"}, {"startTime": 1767.209, "endTime": 1768.369, "body": "It's a good point to mention.", "speaker": "Adrian Dischinger"}, {"startTime": 1768.53, "endTime": 1773.25, "body": "So um basically the main idea is, imagine we have a junior de sitting next to you.", "speaker": "Adrian Dischinger"}, {"startTime": 1773.369, "endTime": 1783.25, "body": "Let's say it's somebody who is, Relatively new to the organization but has lots of background from the public, so lots of, let's say he's from university or something else, provides lots of new insights, new ideas.", "speaker": "Adrian Dischinger"}, {"startTime": 1783.8, "endTime": 1789.51, "body": "But the more senior experience, so the, let's say the driving force in the project has to be available for now.", "speaker": "Adrian Dischinger"}, {"startTime": 1789.589, "endTime": 1798.55, "body": "But for all the, the input, all the initiative is, is driven now from the project, which is very, very good for our senior partners as well to have those, let's say newer ideas.", "speaker": "Adrian Dischinger"}, {"startTime": 1798.859, "endTime": 1810.949, "body": "You've been working on your project for the last decade or so, and now somebody comes in and say, well, why don't we just look into this or this, maybe we, we do a refactoring or do a new database in the background that may decide to that time or money.", "speaker": "Adrian Dischinger"}, {"startTime": 1811.219, "endTime": 1812.77, "body": "So that's new ideas basically.", "speaker": "Adrian Dischinger"}, {"startTime": 1813.42, "endTime": 1830.5, "body": "The the the study which we which we conducted came to a clear answer and they said that it helps more to the senior developers than to the junior developers and the junior developer can probably produce much more code, but he cannot.", "speaker": "Speaker 1"}, {"startTime": 1831.17, "endTime": 1834.55, "body": "Distinguish immediately if that code suggestion is correct.", "speaker": "Speaker 1"}, {"startTime": 1835.13, "endTime": 1840.92, "body": "And so for the senior developers, they know exactly how to phrase the questions and so on.", "speaker": "Speaker 1"}, {"startTime": 1841.209, "endTime": 1853.839, "body": "So for them, the, the productivity increase was bigger and what was not subject of the of the study was how a junior developer can also benefit from it.", "speaker": "Speaker 1"}, {"startTime": 1854.209, "endTime": 1854.79, "body": "So.", "speaker": "Speaker 1"}, {"startTime": 1855.329, "endTime": 1866.819, "body": "If it's like learning a develop a new language or something, then it can for sure be extremely helpful for a junior developer, but as he's working on productive code.", "speaker": "Speaker 1"}, {"startTime": 1867.63, "endTime": 1868.89, "body": "There is some.", "speaker": "Speaker 1"}, {"startTime": 1870.089, "endTime": 1879.29, "body": "Some respect from the tool or something to make sure that I really understand what the code suggestion was and that it really fits to the solution I want to have.", "speaker": "Speaker 1"}, {"startTime": 1879.65, "endTime": 1883.369, "body": "And this is just working better with a senior developer currently.", "speaker": "Speaker 1"}, {"startTime": 1883.839, "endTime": 1891.739, "body": "Um, you mentioned earlier that you rolled out the tool or you created more licenses and now for, for 20,000 developers.", "speaker": "Speaker 3"}, {"startTime": 1892.16, "endTime": 1909.859, "body": "Um, so are you also planning to conduct a more representative study and maybe also incorporate aspects that you just mentioned that I don't know, does it, does it help me to, to learn a new programming language or how can maybe Github copilot also help me train so I, I get on boarded way, way quicker to that coding environment.", "speaker": "Speaker 3"}, {"startTime": 1911.77, "endTime": 1933.31, "body": "So what, what, what, what we are trying to do is if he, if he can somehow find a correlation between the pro or the engagement of the of the user with GitHub co-pilot, so if he uses GitHub co-pilot very often, if then the output is higher, like some, some KPIs, some KPIs which we have defined.", "speaker": "Speaker 1"}, {"startTime": 1933.839, "endTime": 1945.439, "body": "So what we are looking right now is um if we could somehow get the data from those source code management systems and link it to the teams from GitHubb Copilot.", "speaker": "Speaker 1"}, {"startTime": 1945.79, "endTime": 1951.51, "body": "It's always important to not go on an individual level and have everything aligned with the workers' council and so on.", "speaker": "Speaker 1"}, {"startTime": 1951.79, "endTime": 1961.42, "body": "We don't, we are not interested in the performance of an individual developer, but we are interested in, is there a real correlation in the outcome?", "speaker": "Speaker 1"}, {"startTime": 1961.589, "endTime": 1964.31, "body": "Has the invest, is it, is it really valuable?", "speaker": "Speaker 1"}, {"startTime": 1964.4, "endTime": 1967.3, "body": "Yeah, that's one of our next steps.", "speaker": "Speaker 1"}, {"startTime": 1967.31, "endTime": 1969.79, "body": "We're there in discussions with the workers' council.", "speaker": "Speaker 1"}, {"startTime": 1970.18, "endTime": 1973.67, "body": "And really want to make sure to make it right to see there.", "speaker": "Speaker 1"}, {"startTime": 1973.949, "endTime": 1977.969, "body": "I wonder if the number of accesses to stack overflow has been reduced.", "speaker": "Speaker 2"}, {"startTime": 1978.91, "endTime": 1979.89, "body": "Have you measured that.", "speaker": "Speaker 2"}, {"startTime": 1982.29, "endTime": 1990.01, "body": "The, the, the, the, the funny, the funny thing is like, yeah, many, many people don't know that they are not allowed to copy things on the stack overflow.", "speaker": "Speaker 1"}, {"startTime": 1990.27, "endTime": 2000.17, "body": "And in the, in the, in the very beginning when we conducted our investigation with um scanning all the code is, is are some duplicates there.", "speaker": "Speaker 1"}, {"startTime": 2000.829, "endTime": 2001.229, "body": "um.", "speaker": "Speaker 1"}, {"startTime": 2001.68, "endTime": 2010.56, "body": "Yeah, we could find some stack overflow copies where we could inform them the colleague that this is not the way how it should be done.", "speaker": "Speaker 1"}, {"startTime": 2011.479, "endTime": 2017.469, "body": "Then this initiative not only made our developer faster, but also our code base more, you know, legible, right?", "speaker": "Speaker 2"}, {"startTime": 2017.719, "endTime": 2021.599, "body": "I think, I think this is, this is, this is really an important, important part.", "speaker": "Speaker 1"}, {"startTime": 2021.714, "endTime": 2039.385, "body": "I mean, there are also different tools outside, not just GitHubb co-pilot, um, that the developer has the awareness that they should only use the tools which we have checked from the, from the governance side, so that not our know-how is is leaving the company.", "speaker": "Speaker 1"}, {"startTime": 2039.925, "endTime": 2048.419, "body": "So if you put like Somewhere in a in a public G AI tool, our, our source code, you don't know what's happening with that.", "speaker": "Speaker 1"}, {"startTime": 2049.1, "endTime": 2059.159, "body": "And so I think here it can help us a lot with um using, using the official tools and make, make sure that our know-how stays in the company.", "speaker": "Speaker 1"}, {"startTime": 2059.839, "endTime": 2065.878, "body": "Speaking of know-how, so we, we do have our own inner source, right, called Code.", "speaker": "Speaker 2"}, {"startTime": 2066.069, "endTime": 2066.759, "body": "Siemens.com.", "speaker": "Speaker 2"}, {"startTime": 2066.799, "endTime": 2068.279, "body": "It's like a Gitlab, right?", "speaker": "Speaker 2"}, {"startTime": 2068.398, "endTime": 2071.799, "body": "Um, quite, quite nice, uh, kudos to the work, right?", "speaker": "Speaker 2"}, {"startTime": 2071.898, "endTime": 2075.799, "body": "Obviously also from IT and Roger Meyer and the folks, right?", "speaker": "Speaker 2"}, {"startTime": 2075.989, "endTime": 2076.349, "body": "Pretty cool.", "speaker": "Speaker 2"}, {"startTime": 2076.438, "endTime": 2078.358, "body": "Actually, Roger, greets to you to Swiss, right?", "speaker": "Speaker 2"}, {"startTime": 2078.479, "endTime": 2079.779, "body": "You've been on the podcast anyway.", "speaker": "Speaker 2"}, {"startTime": 2080.08, "endTime": 2083.36, "body": "You know, we talked about that, but would it make sense then combine it?", "speaker": "Speaker 2"}, {"startTime": 2083.439, "endTime": 2093.59, "body": "Have you thought about, you know, how do we, how do you a, let's say the capability of an outside, you know, code base which is significant with the internals?", "speaker": "Speaker 2"}, {"startTime": 2096.28, "endTime": 2108.62, "body": "Topic, um, Thomaas mentioned one feature topic is of course the integration of the of the correlation data basically and the second topic feature is that was internal code base to use them to find fine tune the models.", "speaker": "Adrian Dischinger"}, {"startTime": 2108.929, "endTime": 2112.379, "body": "There is also some active work on, on Rouxi's side and his team.", "speaker": "Adrian Dischinger"}, {"startTime": 2112.729, "endTime": 2115.35, "body": "But the important thing to understand is the scale is the issue.", "speaker": "Adrian Dischinger"}, {"startTime": 2115.459, "endTime": 2126.379, "body": "So, um, fine tuning as on a technical level is, is always, is always possible, but the Siemens scale, the big one, the big scale gigabytes of code of text files, basically, that's the main, main issue.", "speaker": "Adrian Dischinger"}, {"startTime": 2126.5, "endTime": 2139.1, "body": "So if we do it on our own, on our own, we have the, of course, just the the the challenge of the costs and and the resources, but we do speak of GitHubb on this topic on a regular basis and, and, and see if we can use it in the future as well there.", "speaker": "Adrian Dischinger"}, {"startTime": 2139.78, "endTime": 2142.09, "body": "And, and uh.", "speaker": "Speaker 1"}, {"startTime": 2142.61, "endTime": 2151.919, "body": "Git GitHub co-pilot is very agnostic, so even with code.siemens.com or something, you can, you can also use GitHub copilot.", "speaker": "Speaker 1"}, {"startTime": 2152.169, "endTime": 2156.189, "body": "But we are very eager to join our forces there and work together.", "speaker": "Speaker 1"}, {"startTime": 2156.489, "endTime": 2168.389, "body": "I know there are many different GAI code assisting tools which are currently in evaluation, and I mean, It's our duty to do it for our developers to provide them with the best tools which we have.", "speaker": "Speaker 1"}, {"startTime": 2168.699, "endTime": 2178.939, "body": "And if that tool is then from Mistral, if it's from Microsoft or whomever, probably you need the solution which fits right to your problem and I think it's important.", "speaker": "Speaker 1"}, {"startTime": 2179.05, "endTime": 2180.399, "body": "That we worked there together, like,", "speaker": "Speaker 1"}, {"startTime": 2181.01, "endTime": 2185.36, "body": "can you give a number indication what is scale in terms of number of software development in the Siemens?", "speaker": "Speaker 2"}, {"startTime": 2185.379, "endTime": 2185.75, "body": "What does it mean", "speaker": "Speaker 2"}, {"startTime": 2185.75, "endTime": 2187.37, "body": "that's an important thing to understand.", "speaker": "Adrian Dischinger"}, {"startTime": 2187.489, "endTime": 2192.25, "body": "So we mentioned before those 200 licenses, that's just a rough ballpark number.", "speaker": "Adrian Dischinger"}, {"startTime": 2192.33, "endTime": 2194.35, "body": "We do have more in the meantime, but um.", "speaker": "Adrian Dischinger"}, {"startTime": 2194.979, "endTime": 2200.32, "body": "To to explain that, it's one of the biggest customers um of Gitta copilot here in in the EMA region.", "speaker": "Adrian Dischinger"}, {"startTime": 2200.899, "endTime": 2203.979, "body": "So that's one of the main challenges of this project is do the scaling basically.", "speaker": "Adrian Dischinger"}, {"startTime": 2204.06, "endTime": 2213.6, "body": "Um, we do have a very small and and and team on Siemens side, just a few people that do the maintenance for this big user group that is very diverse all over the company.", "speaker": "Adrian Dischinger"}, {"startTime": 2214.07, "endTime": 2224.479, "body": "And so as I mentioned before, the models, of course, doing it on-prems would be also possibly do it for other use cases, but then you would have to invest more in terms of manpower to do the the whole maintenance and operations.", "speaker": "Adrian Dischinger"}, {"startTime": 2225.429, "endTime": 2240.159, "body": "And yeah, those, those this this figure which you just asked, I think it's Not 100% existing within its events and 11 additional value which we can bring there is more transparency", "speaker": "Speaker 1"}, {"startTime": 2243.51, "endTime": 2244.459, "body": "software we actually work with.", "speaker": "Speaker 2"}, {"startTime": 2244.85, "endTime": 2254.26, "body": "Uh, including offshore, onshore, partnering, you know, and that it's tens of thousands at least, right, on this kind of scale internationally, right?", "speaker": "Speaker 2"}, {"startTime": 2254.61, "endTime": 2259.669, "body": "Across region, across continents, which makes it, I guess for export control also not that easy, right?", "speaker": "Speaker 2"}, {"startTime": 2261.459, "endTime": 2261.62, "body": "Yeah,", "speaker": "Speaker 2"}, {"startTime": 2261.689, "endTime": 2269.81, "body": "and, and then you probably have some hardware developers which work with a, with a language which is not fully supported.", "speaker": "Speaker 1"}, {"startTime": 2270.139, "endTime": 2274.12, "body": "So there are some obstacles in the way and yeah.", "speaker": "Speaker 1"}, {"startTime": 2275.78, "endTime": 2278.12, "body": "We we're getting more transparency there in the future.", "speaker": "Speaker 1"}, {"startTime": 2279.629, "endTime": 2283.78, "body": "Every project is always, always an opportunity to learn more, right?", "speaker": "Speaker 2"}, {"startTime": 2283.989, "endTime": 2288.25, "body": "So you've been now the hustle, ups and downs, forwards and backs, right?", "speaker": "Speaker 2"}, {"startTime": 2288.379, "endTime": 2292.5, "body": "Feedback and you know, from working from small teams to large scale, right?", "speaker": "Speaker 2"}, {"startTime": 2292.909, "endTime": 2293.75, "body": "What are the learnings?", "speaker": "Speaker 2"}, {"startTime": 2293.83, "endTime": 2296.34, "body": "What would you do differently in the next rollout?", "speaker": "Speaker 2"}, {"startTime": 2296.409, "endTime": 2302.27, "body": "Maybe, you know, the 2.0, right, software augmentation, testing regimes and or any other rollout.", "speaker": "Speaker 2"}, {"startTime": 2302.35, "endTime": 2304.07, "body": "Is there any learnings you can draw?", "speaker": "Speaker 2"}, {"startTime": 2304.215, "endTime": 2304.645, "body": "And", "speaker": "Speaker 2"}, {"startTime": 2304.645, "endTime": 2307.784, "body": "and I think I mentioned before that Siemens is a very diverse company.", "speaker": "Adrian Dischinger"}, {"startTime": 2307.864, "endTime": 2328.604, "body": "We do have many subsidiaries, many, as mentioned, very international in the meantime, and that's, I think it's very, very much known, but we've seen that as a big learning to include as as early as possible, um, all those different regulations, all those local um regulations to make sure that afterwards when you're scaling, you don't have the showstoppers there on on the country basis.", "speaker": "Adrian Dischinger"}, {"startTime": 2329.199, "endTime": 2330.85, "body": "So we did a lot of learning there.", "speaker": "Adrian Dischinger"}, {"startTime": 2331.25, "endTime": 2339.6, "body": "For example, China is a very popular example that has very strict regulations and cybersecurity and so on, India as well and other countries.", "speaker": "Adrian Dischinger"}, {"startTime": 2339.679, "endTime": 2352.739, "body": "So, and that's why we have a lot of learning there and also on the building level and that's also one that we learned a lot lot of insights there, how to build those costs back to the departments on an automated basis, which hasn't been happened so far on a per user basis.", "speaker": "Adrian Dischinger"}, {"startTime": 2353.79, "endTime": 2357.05, "body": "So the, the time was always very critical.", "speaker": "Speaker 1"}, {"startTime": 2357.31, "endTime": 2362.959, "body": "So first we had 2000 users, then by end of April, the, the information came.", "speaker": "Speaker 1"}, {"startTime": 2363.28, "endTime": 2383.83, "body": "Um, that the user base will raised by 18,000 users to 20,000 users, but yeah, like the, the back end was not super prepared for it, and I think it, it was up to the team there to to manage those issues if you would have had more time to set up that organization.", "speaker": "Speaker 1"}, {"startTime": 2384.8, "endTime": 2401.54, "body": "It is more stable in the long run and so currently we are trying to get that that fundament ready and to like clean up, clean up what what we have already done, yeah to to to set it up more more stable.", "speaker": "Speaker 1"}, {"startTime": 2401.879, "endTime": 2409.959, "body": "But honestly saying um with the with the time pressure, we also did a lot of things right and.", "speaker": "Speaker 1"}, {"startTime": 2410.5, "endTime": 2419.76, "body": "Where we also um are in the, are in contact with the PO department to share those insights to make it available for different G AI projects.", "speaker": "Speaker 1"}, {"startTime": 2421.379, "endTime": 2423.12, "body": "Super cool, but it, I wonder.", "speaker": "Speaker 2"}, {"startTime": 2424.139, "endTime": 2448.09, "body": "On a personal level, a lot of fears that one have in utilizing generative AI is about the notion of hydrogenization that means coming up with results that looks cool and relevant and maybe precise, but which aren't because they're made up, made up, maybe the expression not even exist, the variable may not exist.", "speaker": "Speaker 2"}, {"startTime": 2449.239, "endTime": 2458.53, "body": "How, what kind of best practices would you have in terms of mitigating, you know, these spurious results of GR models which maybe exist also in GitHub, right?", "speaker": "Speaker 2"}, {"startTime": 2458.719, "endTime": 2458.82, "body": "Yeah,", "speaker": "Speaker 2"}, {"startTime": 2458.87, "endTime": 2459.62, "body": "that's a good point.", "speaker": "Adrian Dischinger"}, {"startTime": 2459.79, "endTime": 2463.57, "body": "Also from my own experience in the IDE in the environment.", "speaker": "Adrian Dischinger"}, {"startTime": 2463.959, "endTime": 2466.25, "body": "Um, basically a combination of tools.", "speaker": "Adrian Dischinger"}, {"startTime": 2466.37, "endTime": 2471.929, "body": "So, um, the copilot is just one cornerstone in your tooling and also one that can integrate other tools as well.", "speaker": "Adrian Dischinger"}, {"startTime": 2472.04, "endTime": 2479.37, "body": "So for example, if I do Python development on a daily basis, I integrate results for my Python tooling back into the copilot to enhance the results.", "speaker": "Adrian Dischinger"}, {"startTime": 2479.449, "endTime": 2483.929, "body": "For example, as you mentioned, to say, OK, this variable didn't show up in your code, it's probably wrong.", "speaker": "Adrian Dischinger"}, {"startTime": 2484.05, "endTime": 2489.37, "body": "And then which would be in result would be a hint to say, OK, this one looks like a typo, please check it again.", "speaker": "Adrian Dischinger"}, {"startTime": 2489.649, "endTime": 2501.469, "body": "So this sort of the, the, the circle of tools is, is there to, to include all your your insights into one, which is one big benefit of the copilot compared to other tools, which we don't really have this interaction with the tooling in your IDE.", "speaker": "Adrian Dischinger"}, {"startTime": 2502.85, "endTime": 2509.149, "body": "So diversifying the tools and maybe also adding um a little bit of expertise is the secret sauce.", "speaker": "Speaker 3"}, {"startTime": 2509.659, "endTime": 2510.379, "body": "Definitely, yeah.", "speaker": "Adrian Dischinger"}, {"startTime": 2511.35, "endTime": 2513.129, "body": "From from from the.", "speaker": "Speaker 1"}, {"startTime": 2514.379, "endTime": 2529.989, "body": "Rule set which we have, like everybody who's using the GitHub co-pilot needs to understand that it's a co-pilot, it's not the pilot, so the responsibility um still stays with the developers and that's very crucial.", "speaker": "Speaker 1"}, {"startTime": 2530.35, "endTime": 2542.409, "body": "So the developer in the end is responsible and he needs to know that he's responsible and if he doesn't understand the the suggestion he cannot use it so.", "speaker": "Speaker 1"}, {"startTime": 2543.3, "endTime": 2554.26, "body": "Currently it's also like on that process level where we're trying to capture it, and that's why we have seen a bigger productivity increase with the seniors because they're faster in doing that.", "speaker": "Speaker 1"}, {"startTime": 2555.919, "endTime": 2571.77, "body": "Yeah, that's like getting going back again to the users and now you you mentioned the the senior developers as well and I'm, I'm wondering how, how the introduction of this new tool really changed the daily grind for the software and uh developers and do you.", "speaker": "Speaker 3"}, {"startTime": 2571.975, "endTime": 2577.294, "body": "Also perceive maybe a change in the culture, like, do they like do they work differently?", "speaker": "Speaker 3"}, {"startTime": 2577.334, "endTime": 2578.465, "body": "Do they talk differently?", "speaker": "Speaker 3"}, {"startTime": 2578.534, "endTime": 2581.675, "body": "Are they, I don't know, more enthusiastic about their work?", "speaker": "Speaker 3"}, {"startTime": 2581.895, "endTime": 2584.635, "body": "Did the coding process itself completely change?", "speaker": "Speaker 3"}, {"startTime": 2584.814, "endTime": 2587.735, "body": "Maybe you can give us a glimpse into what's the new normal.", "speaker": "Speaker 3"}, {"startTime": 2588.254, "endTime": 2589.375, "body": "Yeah, definitely there's a change.", "speaker": "Adrian Dischinger"}, {"startTime": 2589.415, "endTime": 2591.895, "body": "I mean, I think it's a big change in the industry overall.", "speaker": "Adrian Dischinger"}, {"startTime": 2592.135, "endTime": 2593.594, "body": "You see that before.", "speaker": "Adrian Dischinger"}, {"startTime": 2594.179, "endTime": 2610.699, "body": "Most of the senior developers with the most experience, they had their code base, let be it CEO Java or something else, some legacy code base for the last almost two decades worked on, and more or less a tunnel vision at some point to say I have my knowledge, my insight, my tooling, and they don't really have to change anything.", "speaker": "Adrian Dischinger"}, {"startTime": 2610.739, "endTime": 2612.62, "body": "It works for me, that's my process, that's it.", "speaker": "Adrian Dischinger"}, {"startTime": 2613.139, "endTime": 2617.5, "body": "But I mean software develops, it, it evolves over time, so it's sometimes input from the outside.", "speaker": "Adrian Dischinger"}, {"startTime": 2617.62, "endTime": 2623.02, "body": "So say OK, maybe some change might be good to really um save some time, and resource in the next release.", "speaker": "Adrian Dischinger"}, {"startTime": 2623.139, "endTime": 2633.12, "body": "That sort of inputs we've we've heard from developers from the users that they have some innovation, some, some inspiration is there and helps them to evolve their their their products.", "speaker": "Adrian Dischinger"}, {"startTime": 2634.219, "endTime": 2646.669, "body": "So that was one big thing that we heard from from the users, and those, those insights and um of course just the overall certification, the how satisfied they are with their work has has improved as well from our response to what we heard.", "speaker": "Adrian Dischinger"}, {"startTime": 2649.199, "endTime": 2655.6, "body": "So if you're looking a bit more in the future direction, are there any, you know, the field of AI is fascinating, right?", "speaker": "Speaker 2"}, {"startTime": 2655.709, "endTime": 2659.32, "body": "And we're curious to, to maybe to learn and hear from you as well.", "speaker": "Speaker 2"}, {"startTime": 2659.429, "endTime": 2664.56, "body": "What are there specific innovations, expectations or cool things you love in AI?", "speaker": "Speaker 2"}, {"startTime": 2664.57, "endTime": 2670.169, "body": "Is there a specific application, maybe in software space or a software development space or other, right?", "speaker": "Speaker 2"}, {"startTime": 2670.56, "endTime": 2678.6, "body": "What, what is, what is for you a pretty cool hot ambition level where AI should be, you know, tangled or being pushed for?", "speaker": "Speaker 2"}, {"startTime": 2679.08, "endTime": 2700.149, "body": "One interesting thing, like very recently the CEO from Microsoft, um, I saw a talk with him, and he mentioned like that he in the future probably doesn't really see the need of the software developer because the AI can access directly to the database and changes the data as you want.", "speaker": "Speaker 1"}, {"startTime": 2700.5, "endTime": 2711.06, "body": "And then at the same time, the, the CEO of GitHub, he has given, given spreading his ideas that he's targeting 1 billion software developers.", "speaker": "Speaker 1"}, {"startTime": 2711.54, "endTime": 2715.659, "body": "So within one company they're not really aligned, um, but.", "speaker": "Speaker 1"}, {"startTime": 2716.445, "endTime": 2725.645, "body": "Definitely all those reasoning, reasoning models which, which we, which we see, can they be improved to use less energy, for example.", "speaker": "Speaker 1"}, {"startTime": 2726.114, "endTime": 2730.875, "body": "So currently it's 01, we see some benchmarks from 03, where is the future?", "speaker": "Speaker 1"}, {"startTime": 2731.274, "endTime": 2733.574, "body": "I think that is very interesting.", "speaker": "Speaker 1"}, {"startTime": 2734.889, "endTime": 2736.649, "body": "Yeah, I mean, we mentioned a few trends before.", "speaker": "Adrian Dischinger"}, {"startTime": 2736.77, "endTime": 2742.929, "body": "Fine, fine tuning is big one big one, also the retrieval of of documents or work and retrieval of augmented generation.", "speaker": "Adrian Dischinger"}, {"startTime": 2743.439, "endTime": 2755.669, "body": "So you have your, your documents, your, your documentation, your, your inputs from all over the place that are then um fetched and included into your, your prompts is one big trend that is, has been more and more growing over the year and um.", "speaker": "Adrian Dischinger"}, {"startTime": 2756.09, "endTime": 2777.429, "body": "I think in general, as as I mentioned, the just the the context or the inclusion of more and more say hardware insights maybe or bigger environments would be one big trend that we see in the future to basically move from just your software to let's say, for example, infrastructure as code, to say, OK, I want to increase my my whole workflow and create the infrastructure as well using the same workflows.", "speaker": "Adrian Dischinger"}, {"startTime": 2779.58, "endTime": 2792.82, "body": "So, very exciting times ahead and we will follow you closely guys and maybe also record a second follow up episode once we're larger in scale and uh maybe even even more precise.", "speaker": "Speaker 3"}, {"startTime": 2793.419, "endTime": 2801.209, "body": "So before we wrap things up, maybe do you have some advice for the for the people out there listening in right now?", "speaker": "Speaker 3"}, {"startTime": 2801.5, "endTime": 2807.31, "body": "I guess at Siemens but also beyond the Siemens boundaries, people are introducing new tools and.", "speaker": "Speaker 3"}, {"startTime": 2807.729, "endTime": 2810.37, "body": "Yeah, coming up with innovation, how do you really drive that?", "speaker": "Speaker 3"}, {"startTime": 2810.439, "endTime": 2815.949, "body": "Is there something um that you took away from your journey that you would like to share with the audience out there?", "speaker": "Speaker 3"}, {"startTime": 2817.29, "endTime": 2822.239, "body": "Um, yeah, of course, I mean, um, as I mentioned before, it has big benefits, so do take those tools out.", "speaker": "Adrian Dischinger"}, {"startTime": 2822.29, "endTime": 2828.919, "body": "I mean, we see them being more and more important for, for your daily tasks, so make sure to, to have some experience, on some tests.", "speaker": "Adrian Dischinger"}, {"startTime": 2829.209, "endTime": 2831.729, "body": "I think that would be one big um advice from my side.", "speaker": "Adrian Dischinger"}, {"startTime": 2832.399, "endTime": 2858.59, "body": "So from one of the podium discussions, I was one advice was one hint was it will not be the artificial intelligence which will take your job very soon, it will be another developer which is using artificial intelligence which is taking your job and um even though there is currently fear, I think it's unavoidable and we should be open to that change and.", "speaker": "Speaker 1"}, {"startTime": 2859.04, "endTime": 2865.139, "body": "If we can improve our output with the help of AI then it helps Siemens and it helps us all.", "speaker": "Speaker 1"}, {"startTime": 2867.239, "endTime": 2869.699, "body": "Right, that's, that's good advice for sure.", "speaker": "Speaker 3"}, {"startTime": 2870.01, "endTime": 2873.26, "body": "And one last thing before we jump to the final game.", "speaker": "Speaker 3"}, {"startTime": 2873.679, "endTime": 2887.08, "body": "So, I feel that you have so much passion for this Github co-pilot and this was like you walked such a long journey and there were so many hurdles with the legal people like jumping in your way and like so many considerations, so many people to take along.", "speaker": "Speaker 3"}, {"startTime": 2887.12, "endTime": 2895.0, "body": "Like what like what is it that that that like that maybe that one thing that that makes you so passionate about the GitHubb co-pilot.", "speaker": "Speaker 3"}, {"startTime": 2896.28, "endTime": 2910.399, "body": "I think it's currently you can define the future a little bit or you can, you can spearheading, heading the direction and this is like so so motivating and like.", "speaker": "Speaker 1"}, {"startTime": 2911.149, "endTime": 2925.139, "body": "When you, when you work on a, on a different project with PLM or something, then you use Google and it's very hard to get any information, but that topic is really currently in the focus, you find so many information and there is.", "speaker": "Speaker 1"}, {"startTime": 2925.899, "endTime": 2935.52, "body": "Yeah, you have, you have, yeah, you have basically the chance to get so, so much information and also um define, define it within Siemens, maybe um partly.", "speaker": "Speaker 1"}, {"startTime": 2936.25, "endTime": 2937.949, "body": "That's for me very interesting.", "speaker": "Speaker 1"}, {"startTime": 2938.389, "endTime": 2941.409, "body": "Giving you the creation of freedom or.", "speaker": "Speaker 1"}, {"startTime": 2942.11, "endTime": 2953.179, "body": "Yes, as I mentioned, it's a big chance to uh basically participate, to contribute to the future of the company, to drive soft development at Siemens, to say how does our future look like as at Siemens.", "speaker": "Adrian Dischinger"}, {"startTime": 2953.5, "endTime": 2959.229, "body": "We do focus more and more on software than on hardware as in the past, so I think it's just a big chance to be part of this movement.", "speaker": "Adrian Dischinger"}, {"startTime": 2960.32, "endTime": 2961.31, "body": "That's true,", "speaker": "Speaker 3"}, {"startTime": 2961.32, "endTime": 2967.149, "body": "and there are not many cases where, you know, it's it's leapfrogging the software company Siemens, which we want to be.", "speaker": "Speaker 2"}, {"startTime": 2967.32, "endTime": 2988.419, "body": "And the second uh at at scale double digit impact on, you know, on productivity is not in many cases, you know, uh, so much tangible, so much geeky, so much value actually also, you know, with a positive feedback from each individual, and that's, I think it's a good, you know, showcase and lighthouse, I think initiative, you know, for Siemens to stand out.", "speaker": "Speaker 2"}, {"startTime": 2988.84, "endTime": 2992.52, "body": "The usage of TNAI as a positive tool for for ourselves.", "speaker": "Speaker 2"}, {"startTime": 2993.959, "endTime": 3007.06, "body": "So, uh, Thomas and Adrian, thank you so much for taking the time for coming over, taking the long, long way with the unstable trains here in Germany to come to the Ilab, and we've reached the very end of the session.", "speaker": "Speaker 3"}, {"startTime": 3007.36, "endTime": 3013.149, "body": "But, uh, before we'll let you go, we want to play our signature game called Authentic Autocomplete.", "speaker": "Speaker 3"}, {"startTime": 3013.439, "endTime": 3016.459, "body": "Also a little bit of a co-pilot, right?", "speaker": "Speaker 3"}, {"startTime": 3016.919, "endTime": 3019.219, "body": "Um, and trust me, it will be lots of fun.", "speaker": "Speaker 3"}, {"startTime": 3019.57, "endTime": 3026.08, "body": "Um, but this time I'll give you the sentence starters and you'll bring them home with your finishes.", "speaker": "Speaker 3"}, {"startTime": 3026.56, "endTime": 3028.219, "body": "So ready to give it a go.", "speaker": "Speaker 3"}, {"startTime": 3028.719, "endTime": 3029.159, "body": "Yeah, that's good.", "speaker": "Adrian Dischinger"}, {"startTime": 3029.6, "endTime": 3031.62, "body": "OK, Adrian, why don't we start with you?", "speaker": "Speaker 3"}, {"startTime": 3031.83, "endTime": 3032.659, "body": "Siemens is.", "speaker": "Speaker 3"}, {"startTime": 3033.61, "endTime": 3039.429, "body": "An innovative company that has lots of history but it's constantly evolving and big future.", "speaker": "Adrian Dischinger"}, {"startTime": 3040.34, "endTime": 3041.58, "body": "Hopefully, yeah, a big future.", "speaker": "Speaker 2"}, {"startTime": 3041.82, "endTime": 3044.34, "body": "See that Siemens, colleagues, big future ahead.", "speaker": "Speaker 2"}, {"startTime": 3045.169, "endTime": 3046.31, "body": "And Thomas, AI", "speaker": "Speaker 3"}, {"startTime": 3046.31, "endTime": 3054.6, "body": "is a great opportunity and it's and I and I think it's not possible to be stopped anymore no matter what you think about it.", "speaker": "Speaker 1"}, {"startTime": 3055.06, "endTime": 3056.1, "body": "Unstoppable.", "speaker": "Speaker 3"}, {"startTime": 3057.909, "endTime": 3059.75, "body": "Adrian's software 2.0", "speaker": "Speaker 3"}, {"startTime": 3059.75, "endTime": 3062.989, "body": "is definitely based on tooling and using AI.", "speaker": "Adrian Dischinger"}, {"startTime": 3063.35, "endTime": 3067.76, "body": "We see this in all our vendors and our partners in the industry, so that's one big thing from my side.", "speaker": "Adrian Dischinger"}, {"startTime": 3068.86, "endTime": 3074.959, "body": "And Thomas, humans and AI humans and AI are a match made in heaven because.", "speaker": "Speaker 3"}, {"startTime": 3076.25, "endTime": 3078.07, "body": "Humans are sometimes lazy.", "speaker": "Speaker 1"}, {"startTime": 3080.62, "endTime": 3082.139, "body": "Yeah, sure.", "speaker": "Speaker 3"}, {"startTime": 3082.75, "endTime": 3084.75, "body": "And Julie, why don't you take the last one?", "speaker": "Speaker 3"}, {"startTime": 3085.62, "endTime": 3092.199, "body": "Um, so, uh, Adrian, if I could invent one rule for everyone in the world to follow, it would be.", "speaker": "Speaker 2"}, {"startTime": 3093.27, "endTime": 3101.08, "body": "To um be more communicative, to be more innovative, to enhance your your, your interactions with your teams, that will be a big thing for all of us.", "speaker": "Adrian Dischinger"}, {"startTime": 3101.76, "endTime": 3101.79, "body": "Yeah.", "speaker": "Adrian Dischinger"}, {"startTime": 3102.199, "endTime": 3105.479, "body": "So let's all comply to to Adrian's new rule in the world.", "speaker": "Speaker 2"}, {"startTime": 3105.59, "endTime": 3106.199, "body": "Here we go.", "speaker": "Speaker 2"}, {"startTime": 3106.56, "endTime": 3114.55, "body": "Thanks so much, Adrian and Thomas for being open and minded and and, you know, sharing a bit of the journey of what it means, you know, unleashing.", "speaker": "Speaker 2"}, {"startTime": 3114.879, "endTime": 3118.389, "body": "Generate FAI technology in Siemens and and the tool stack.", "speaker": "Speaker 2"}, {"startTime": 3118.449, "endTime": 3121.639, "body": "So thanks so much for, you know, taking, taking the time and sharing", "speaker": "Speaker 2"}, {"startTime": 3121.639, "endTime": 3121.979, "body": "this.", "speaker": "Adrian Dischinger"}, {"startTime": 3122.469, "endTime": 3122.679, "body": "Yeah, thanks for", "speaker": "Adrian Dischinger"}, {"startTime": 3122.679, "endTime": 3123.04, "body": "having us.", "speaker": "Speaker 1"}, {"startTime": 3123.28, "endTime": 3124.26, "body": "Thanks for the invitation.", "speaker": "Speaker 1"}, {"startTime": 3124.989, "endTime": 3133.959, "body": "And folks out there, stay bold, committed and open-minded, you know, connect with Thomas and Adrian on LinkedIn, um, you know, or comment on, under this post, right?", "speaker": "Speaker 2"}, {"startTime": 3134.07, "endTime": 3136.889, "body": "What do you like, what you don't like, what we should improve, right?", "speaker": "Speaker 2"}, {"startTime": 3136.979, "endTime": 3147.01, "body": "Obviously, we want to be every episode even better, um, you know, and most importantly, stay bold, committed and open-minded, and we hear us at the next Siemens AIL podcast.", "speaker": "Speaker 2"}, {"startTime": 3147.54, "endTime": 3147.939, "body": "Cheers.", "speaker": "Speaker 2"}], "language": "en-US"}