Artificial intelligence is still an enigma for the majority of Germans. According to a study by the Bavarian Research Institute for Digital Transformation (bidt) conducted in late 2019, around three quarters of internet users in Germany say they only have a vague idea of what AI actually is. More than one in five have either no knowledge of what lies behind artificial intelligence or have never heard of the term. And this, despite the fact that AI is already an integral part of our everyday life. Machine learning is now used for internet searches as well as for voice assistance on smartphones; algorithms also perform numerous tasks in medical diagnostics and mechanical engineering. But this is only the tip of its potential. In the coming decades, AI systems will play a role in almost all economic and social sectors.
Vision: AI-competent society
With the increasing use of AI, the need for skilled workers also increases. According to a study by the Stifterverband on Future Skills, by 2023 around 450,000 employees with specialist IT skills will be required in the business sector alone. Not everyone has to become a computer scientist, says Florian Rampelt, Head of the AI Campus office at Stifterverband. What is needed is "a broad range of essential competences and skills." Everyone has to understand how the data-based decision-making processes work behind AI systems in order to be able to interpret and use them properly. This applies to a doctor who has been working in the practice for 20 years and is suddenly confronted with diagnoses from mobile phone apps, as well as to an engineer working with autonomous driving. Trainee teachers also have to understand AI basics today in order to be able to teach them in the classroom tomorrow. “Our vision is an AI-competent society”, adds Rampelt. “We must aim to enable as many people as possible to understand the essence of artificial intelligence and the related technologies. If we can then build on this and inspire even more people to enter exciting AI professional fields, then we have achieved our most important goals." The AI Campus offers a range of innovative learning opportunities in different AI areas and imparts essential AI skills to students, professionals and lifelong learners.
Florian Rampelt, Head of AI Campus © Stifterverband
Beta launch with multiplier effect
The online course “Schule macht KI” (“AI for Schools”) presents the subject artificial intelligence to teachers and trainee teachers in a way that brings fun to the classroom, while the interactive podcast course “Dr. med. KI” (“Dr. med. AI”), with Kerstin Ritter, junior professor for Computational Neuroscience at Charité Berlin, shows in an entertaining way how AI is used in medicine - these are just two of the first online courses launched on the beta version of the AI Campus' learning platform in July 2020. Three quarters of a year after the start of the project, which is funded by the Federal Ministry of Education and Research (BMBF) until 2022, and nine months earlier than planned. "We made a conscious decision to launch the beta early in order to help strengthen self-organised learning in general and AI competences in particular by providing free online courses during the Coronavirus period," says Rampelt. The project partners, which include Stifterverband, German Research Center for Artificial Intelligence (DFKI), Hasso Plattner Institute (HPI), NEOCOSMO and mmb Institute, are delighted with positive response: since the soft launch, several hundred users have already been participating in courses on the platform. “Our goal is to reach a four-digit number of learners in the near future,” explains Rampelt, taking into account the multiplier effect. If, for example, Kerstin Ritter from Charité integrates the podcast into her own teaching lecture and the start-up Junge Tüftler brings its own community of teachers and trainees with it, this would immediately boost the numbers of learners on the platform.
Primacy of openness: Copy as much as you like
It is also welcome that the first English-language course developed by the Berlin social start-up Kiron for the AI Campus will be copied onto the Society for International Cooperation's (GIZ) learning platform. “Copy as much as you like, that's the first thing I tell our partners,” says the AI Campus head. “We want to strengthen AI competences, but we know that we cannot achieve this alone. This can only be achieved if we work together.” It is not about “reinventing the wheel” or replacing other platforms, but about achieving greater visibility and networking for AI. Therefore, in addition to the so-called AI Campus Originals - i.e. courses, videos, podcasts etc. specially developed for the platform - users can also find German and English-language courses from other platforms on the website. Courses from the University of TU Eindhoven and Harvard University are available as well as content from partners DFKI and Hasso Plattner Institute and the Hamburg start-up AppCamps.
Digital hub in Berlin-Kreuzberg
"Our cooperation partners come from different regions, our consortium is spread across Germany - and that is important," explains Florian Rampelt, "but the heart of the AI Campus is our digital hub in Berlin-Kreuzberg." The choice of location was a conscious one. On the one hand, the Stifterverband's digital department with over 80 employees has long been based in Berlin. The DFKI also has an office in the German capital. “Berlin is a central location for innovative AI development and scientific excellence. I really admire Berlin's innovative spirit," says the University of Passau graduate, who has been working in Berlin for almost five years. “Many start-ups, universities and other initiatives are doing great educational work here. There are just so many inspiring stakeholders. We benefit enormously from this."
Quality for all
The cooperation partners and learning opportunities selected for the AI Campus are determined according to specific quality criteria. This includes an academic level. However, the platform should not only be attractive for students, but also for the general public. "The most important thing is that the courses are open and available free of charge, and that it only requires a quick registration," says Rampelt, in favour of a low-threshold educational programme. “AI skills shouldn't be a privilege,” he emphasizes. “Payment barriers or other formal hurdles are not included. And that's important. We don't want to serve an elite that can already afford it.” "The primacy of openness” takes the philosophy of the AI Campus one step further: All AI Campus Originals are subject to open licensing, which means that the courses can be republished free of charge and also be integrated into any institution's teaching module. The technologies used also follow the principle of openness: The Hasso Plattner Institute's openHPI platform for massive open online courses (MOOC), on which the AI Campus' learning environment is based, is also an open source project.
"Everyone can copy, use and integrate the courses into their own curriculum or into their own platform," says Rampelt of the topic close to his heart.
From micro-degree to chatbots
The project team at the AI Campus has three years to prove that the concept works and that there is demand for these courses. It will be three very eventful years with strong development: “The focus is on testing learning opportunities in several areas of artificial intelligence,” he explains in a podcast interview. Which operating models are suitable for an AI learning platform? How can income be generated in the future? The AI campus deals with these and many other questions on an ongoing basis. And last but not least, the platform itself must be expanded and further developed. Over 20 new AI campus originals on various topics are currently being created. The ideas and formats for this are provided by scientists from universities in Germany and Luxembourg as well as experts from the DFKI. "These are not just courses. They include learning through a wide variety of modes such as further podcast series, learning videos and micro-degrees," says the Berlin resident. In addition, with a little luck, the first courses with ECTS credits will already be available for the winter semester 2020/21, and will be recognised or credited in coordination with the respective university. They are not intended to replace existing analogue courses. Instead Rampelt believes that the online courses can complement lectures or be recognised as part of a seminar. But the AI Campus is not only continually developing in terms of content, it is also being optimised on the technological level, too. AI systems are expected to play an increasingly important role on the platform. "The DFKI is working on the development of chatbots for the AI campus, and recommendation systems are also planned," says the Campus head. In the first phase, anonymised data will be collected in order to determine which courses are clicked on most frequently and where the online students stay longest. From this data, recommendations can be derived. “In the medium term, the dashboard will make personalised recommendations for learners,” he adds, looking to the future. “However, data protection is our top priority.” Because the responsible use of personal data is also an important part of an AI-competent society.