Hello everybody. Thank you very much. I'm happy to see that the room is actually crowded. Last time I looked like very few seconds ago, the room was empty. So, happy to have you all here. Um as you know, and thanks for the introduction. Um my name is Tina Klüwer and I'm just currently with the Ministry of Research and Education uh where I started only 5 months ago. So, still quite fresh. And this is a very special special situation for me because I have been in the community of artificial intelligence in Germany for so long in so many different positions. And now I'm quite fresh in this ministry position and it feels a little bit weird, I have to confess, as well, but it also feels good.
And I hope that I can, as I Well, which was my big hope actually to take over this position, that I can bring in all my experience viewpoints that I had into this work that I now do. And today I want to talk about science transfer in artificial intelligence, one of the topics I have been working with for the last years, uh founding my own AI startup, but also supporting others in the last 2 and 1/2 years at the KIC the um artificial intelligence entrepreneurship center of the Berlin Universities where I tried to support as many AI startups as possible in becoming successful. And um now I try to bring this perfect different viewpoints in this new position.
And I will tell you a little bit about what Germany has already done in the last years regarding artificial intelligence and transfer, and also what's still ahead of us. And uh we will start with the well-known AI strategy of the government that I'm pretty sure many of you have already heard about. Um as you may know, we are were one of the first countries in the world to have an AI strategy, which I find quite impressive. So, already in 2018, the government decided to come up with a new paper on how we want to strengthen artificial intelligence in Germany. The first AI strategy had 12 different working areas, such as the research, of course, but also transfer and business, uh training of skills, and um standards, um research in the area of labor.
All important aspects to artificial intelligence already there in 2018. And then in 2020, there was an update to that strategy, which uh concluded it all together to six main pillars, and one of those are heads, meaning talents, uh research, still an open and important um topic, of course, transfer and applications, world of work, frameworks and standards, and the society. Of course, the government does not only want to bring up new regulation or want to force the innovation by building financial forces, but also we want to include you all, yeah? So, to talk about what your opinion is about artificial intelligence and how we want to uh live with AI AI in our society. Um there was also a budget allocated to it.
It wasn't the 40 billion that uh Fabian just mentioned, but still, so the government uh allocated 3 and 1/2 billion euros to artificial intelligence. And as you can see here, these are the biggest um the six main pillars I just showed, and as you can see, um almost um a little bit more than 2 billion go to transfer and application, which is basically everything that deals with industry. Yeah, so it's not only transfer in the meaning of transfer, but it's also everything that supports business and and industry. Then the second biggest uh area is research with almost a billion. Minds and talents the next one uh labor market, society and frameworks. And what came out of that?
As I mentioned in the beginning, I'm now with the Ministry of research, so I'm um explicitly interested in this 760 million um in research. And as I when you check it out, you can see that we do have a long history of artificial intelligence research in Germany. Um only no not even 20 years after the first time that the term artificial intelligence was generated, we also had our first research project um for electronic language research directly followed by a even bigger one, which was called artificial intelligence knowledge-based systems. And after that and in the course of that, we also founded German Research Center of Artificial Intelligence also as one of the countries having such a research institution so early in 1988, uh the it was founded and from then on also uh financed by the government.
And it's still existing and expanding. Um there were several other important projects and decisions in the field of artificial intelligence in Germany and also some very important results. We have just heard in the in the speech of Fabian that um uh Hochreiter mentioned he wants to destroy out of alpha, which we will see, but anyway, he in 1997 he and his colleague Schmidhuber also made a very important um result in research result, which was the long short-term memory, which is now one of the main pillars of neural networks and also of things like ChatGPT and GPT in total, generative AI. So, there is a long history. So, we have done a lot and it was always in this time also publicly funded.
It was financed by the government. Not always, not all of it, but a lot. Which resulted in the situation that we now do have in Germany more than 90 AI research institutions and university departments all distributed around Germany. So, you can see there's a lot going on already. And they all focus on different aspects of artificial intelligence. So, this could be everything all from the technological baseline and foundations to ethical things that you have to think over. And because we thought that this is not yet enough, the ministry also came up with even more artificial intelligence dedicated research competence centers and generated those six new centers for artificial intelligence a few years ago, which receive get more than 120 million euro funding each year.
So, those are centers, the colors indicate that they belong together and they have different focus areas. So, everyone has a specific area where they deal with different questions in artificial intelligence, but they all perform cutting edge research. It's one Those are the six like excellent research centers that were additionally generated. And this all has results. As we know, Germany is a notable AI research country. So, if you check for publications in 2023, you can see, of course, as we already know, China is one is the biggest publication producer so far. The US right after that. This is India down there. And then Great Britain and Germany. So, there is output out of all the research that we do.
But, transfer is a different question. So, bringing all those results that we have from research into applications, using them for the society, for business, for a better world, yeah? So, that's a different question. And here you can see, so those are all numbers from the current AI index report from 2024, so quite fresh. And you can see where we rank when it comes to newly funded AI companies, for example, which is one of the main pillars of research transfer. Bringing companies uh research results into companies to use the uh insights there. So, we rank nine here. And United States, big difference to everybody else. Then China, United Kingdom, Israel, Canada, France, India, Japan. And if you look at that, for example, France, can I get back here as well?
is quite behind us here in the matter of um publications, but still they do have more um new companies started in the field of AI, and I think maybe that's the better ratio, yeah? So, but we will check. So, what do we have else? Um one more aspect that people look at if they find try to select what's the meaning of transfer in country is um AI patterns. And also here, we rank ninth. So, we do have a different situation actually. South Korea at top, Luxembourg, yeah, also quite interesting. Then the US, Japan, China, Singapore, Australia, Canada, and Germany. So, here also the same rank. And you can say that Germany is one of the largest producers of scientific results worldwide, but we only rank 18 out of 70 uh 27 EU member states for transferring these results into practice.
And this is what we see also in the charts that I've just shown. Um and um investigation by the VC fund Earlybird once found out that although a lot of successful research research is being carried out, the results are rarely translated into applications. Only an estimated 25% are used. That's the real low number, but it also means there's 75% potential, yeah? So, always focus on the things that you can improve, and I think we do have some chances here. So, to sum it up, what is the status of transfer AI transfer bringing research results into application in Germany? Currently, we only have 15% of all German companies saying that they use AI. That's low. Yeah, so imagine the case that 50% of all German companies use AI.
What a difference that would make, huh? Well, how we could exploit a lot of new innovative business ideas and new products. Maybe also digitalized ways to live. Yeah, so there's a lot of potential here. We have roughly 500 AI startups founded in Germany, which is also increasing, but still not so much. Uh we have a strong role of research in this area because 40% of startups are spin-offs from academia, which is a super high number compared to all startups. So, if you say all startups have 2.4% research background. And regarding the AI patterns, we routinely rank mid-tier behind USA and China, but also South Korea and Japan. So, what would we do? The government thought, "Okay, we do have a right research base here, but we want to move this research results into the application even more."
So, the Ministry of Research came up with a plan, which is the Action Plan AI. It was published in November 2023, and it's an implementation of the AI strategies that I showed in the beginning. So, it's not a totally different plan with totally different goals or something. It's the same thing, but it makes clear what the Ministry of Research can do to achieve those goals. And we combine 11 fields of action and analyze the status quo and derive goals from that, and transfers one of the main overall goals here throughout the whole idea of things that we could do. Um to give you some examples, to foster AI transfer, there was a new SME specific funding program generated, which is called Kai for Kai Mu, where especially um yeah, so small medium-sized enterprises can apply for funding, and we hope it's a little less bureaucratic, so that's the goal, and it should be available uh for artificial intelligence application in this SMEs.
Uh we've set up a regional competence center for labor research in artificial intelligence to help companies to actually understand what artificial intelligence and applying artificial intelligence means for their workforce. Because, as you already know, um there are many misunderstandings regarding artificial intelligence. Well, and there are also many um employees that are feeling a little a lot endangered about it. So, this is our hope that we can help with that um project to help the uh the companies understand better how they need to apply AI and how they can make their workforce accept it. Uh we have set up the AI competence centers that I just showed, the research centers, and they have the um they they the duty as well to act as seeds for AI transfer to industry and startups, and all players together should create a regional ecosystem with different stakeholders.
We additionally generated four AI service centers. One is in Berlin-Brandenburg, one is in Lower Saxony, one in Hessen, and one in Ostwestfalen, and they offer targeted AI advice and support for SMEs. So, if in case you are an SME, you need help because you're still struggling with how to apply AI, what kind of AI to use, is there any way any use case that is necessary for your business in AI, you could contact you can actually I would love to encourage you all to do that, yeah. So, please contact those AI service centers and ask for advice and support and afterwards tell me how it was. Um and uh we also offer low threshold access to computing power for startups and SMEs.
So, we do have like 1,000 GPUs here, which is still growing. We all know it's not so easy to get GPUs nowadays, so you have to wait for them quite a long time. Um and if you're a small player, which the German government actually is, obviously, then Nvidia may take a while to deliver that to you. Um but we also have a link to the Jülich Supercomputing Center, which is actually just already now a big supercomputing HPC uh oppor- opportunity. It's not only for research, you can use it for commercial um projects as well, and it's still growing. So, we are building here one um one of the biggest supercomputer with GPUs in um Europe.
Um it will be combined and concluded in the next year, beginning March April of 2025, so hopefully, and we'll have 25,000 GPUs there. So, keep it in mind. And we have added 10 to 20 million funding additionally in different programs to SMEs. But this is not everything, as you all know, uh we also need a lot of talent to make it work, yeah. And this is the last aspect of it that I want to talk to you about today. So, academia, startups, and industry, they all need the AI talents, and as the Ministry of Research and Education, we feel responsible for that, too. So, we offered the competency offensive, which is basically um adding more AI professorships all over Germany.
So, you can see not in that chart in that map where we all added more AI departments to universities and institutions. But, we also had several calls for junior scientists, especially women, and we have offered started the AI campus, which is a learning platform for the bright broader public. And I want to say that this actually pays off. Last chart, because if you compare this to the things that I've just shown you in the beginning regarding the transfer, this shows a totally different picture. So, the AI skill penetration rate for countries, which means basically how much AI skills do you have in the country, how many people have AI skills, you can see we rank third after India and the US, and also not with the big uh gap here, but and with a meaningful amount of people, and I think that's a good sign, and it's also something that makes us strong, and that we should not forget that we have that.
Thank you very much.