CDU/CSU on artificial intelligence: key technology for Germany
The CDU and CSU in the German Bundestag are demanding more commitment from the federal government to artificial intelligence. They failed to further develop the AI strategy that had existed for five years and was developed by the previous government. However, “active political will, transparency about the financial resources provided and rapid implementation measures are urgently needed in order not to fall behind in international competition,” says a motion from the parliamentary group, which is to be debated in the plenary session of the Bundestag next Thursday.
Last August, the Federal Ministry of Research presented an action plan according to which the state should in future provide tools, skills and infrastructure related to AI as “part of basic services”. The Research Ministry plans to invest 1.6 billion euros in this field over the next two years. The action plan “shows many areas of action,” write the CDU and CSU in their application, but the financing of the individual implementation steps remains unclear.
The opposition faction accuses the federal government of saying that its draft budget envisages savings of around a quarter in the area of AI. The item for IT infrastructure in the AI area will be reduced from 24 million to 18 million euros. The up to 30 Humboldt professorships in the field of AI have not yet all been awarded by 2024.
The federal government should now create the conditions so that existing supercomputing infrastructure can be expanded and expanded, with which large AI models can be created. Start-ups, SMEs and open source developers should have their own access to the expanded capacities at FZ Jülich or LRZ Garching, for example. There should be special funding for the promotion of open source AI.
In mid-September, a position paper from the CDU became known, according to which the German judiciary and administration should rely extensively on artificial intelligence in the future. This should not only keep the administration functioning, but also relieve employees and simplify official procedures. The current proposal from the CDU and CSU (PDF) goes beyond the positions described therein.
Development and application based on European values
The CDU and CSU believe that large AI models called Foundation Models offer great opportunities for Germany and Europe. These are crucial for further technological development as they help transform the way AI systems are built. Since they would be trained with a large amount of data, they could later be adapted to tasks. There are already such large AI models in the USA, but Europe needs to become competitive quickly in order not to be dependent on US technology. “The development and application of the technology should be based on European values,” the application states.
Quantum computing could greatly advance the development of AI, the group continues. The potential lies, among other things, in optimizing and accelerating complex processes such as logistics, drug development, cybersecurity, financial risk management, network planning or machine learning. With AI it is also possible “to precisely control quantum particles much faster and more efficiently than before using tailor-made electromagnetic fields.”
Another keyword that appears in the application is “neuromorphic computing,” which attempts to emulate the way the human brain works in order to optimize machine learning and data processing. The approach can significantly increase the performance of AI and is particularly useful for complex tasks such as visual and auditory signal processing, decision-making and autonomous robotics.
“Knowledge can flow away”
The CDU/CSU refer to media reports according to which employees of educational institutions, organizations and companies used chat models for official tasks, in particular ChatGPT. Here the group lists software development and debugging, but also many other knowledge activities. This means that knowledge within the organization can flow away, which is why large companies are turning to providing their own “on-premise systems” in order to use AI in their own, protected environment. State institutions would also have to act in the same way, the application states.
AI should be strengthened as a central learning content in educational offerings. Knowledge about AI models, including their possible socio-cultural effects, must be anchored “across society”. Last but not least, qualified AI specialists must be recruited beyond the domestic potential and international talent in the field of AI research must be specifically recruited, write the CDU and CSU in the application.
To the homepage
#CDUCSU #artificial #intelligence #key #technology #Germany