- In a consortium of 10 German organisations, Data & AI consultant Alexander Thamm is developing the European answer to GPT-3 from the USA.
- AI language models produce independent texts, support in customer service or convert everyday language into programming code
- The Federal Ministry for Economic Affairs and Climate Protection is funding the OpenGPT-X project with around 15 million euros.
AI Lighthouse Project for Europe
Munich, 20 January 2022 - The OpenGPT-X project is being launched under the leadership of the Fraunhofer Institutes for Intelligent Analysis and Information Systems IAIS and for Integrated Circuits IIS. It is one of the Winners of the Gaia-X funding competition.
The aim of the project is to develop a large AI language model for Europe. "For business, industry and society, the disruptive potential of AI language technologies is enormous, and international competitors have already recognised this. A European AI language model such as OpenGPT-X is therefore imperative to ensure Europe's digital sovereignty and market independence".says Dr Nicolas Flores-Herr, head of the project at Fraunhofer IAIS. During the development, special emphasis is placed on data protection and European language diversity.
Existing large AI language models* such as the American GPT-3 or the Chinese Wu Dao 2.0 use huge amounts of data to create diverse AI models. The problem is that data protection and the integration of European languages are not guaranteed.
With OpenGPT-X, Germany enters the race for the best language AI
"With the development of OpenGPT-X, Germany is entering the race for the best speech AI," says Alexander Thamm, founder and CEO of the eponymous data & AI consultancy. "Large, powerful language models open the door to completely new applications." A special feature of these models is their ability to generalise text. Without special training data, these AI systems can flexibly solve complex "intelligent" tasks, such as writing sophisticated texts, providing support in customer service or generating programming codes from everyday language.
"Whoever masters the training and use of large AI models will have a great supremacy in the market," adds Alexander Thamm. "If we miss this chance, Europe risks becoming dependent on American service providers again, comparable to the search engine market. US companies receive the data of European users and can use it to further improve their language models. A self-reinforcing cycle that leads to overpowering monopolies. So I am very proud that we are involved in an AI project of this magnitude."
"Whoever masters the training and use of large AI models will have a great supremacy in the market."Alexander Thamm, Founder and Managing Director Alexander Thamm GmbH
Due to the high technical requirements, such as computing power, such powerful language models can so far only be implemented by large companies or consortia. For this reason, the project was able to win a variety of well-known partners. Fraunhofer IAIS together with DFKI and the companies Aleph Alpha and Alexander Thamm GmbH will develop the AI models. "A particular challenge for training such large AI models is data management and the establishment of stable and continuous operation models. Here, [at] - Alexander Thamm brings its experience from over 1,000 data and AI projects to the consortium and will implement powerful processes and applications for data management and the automation of model creation," explains Managing Director Alexander Thamm.
In addition, the Jülich Research Centre and the TU Dresden are making their high-performance computing capacities available. 1&1 IONOS SE will supervise the integration into the Gaia-X structure. The project will receive domain-specific data for the development of the models from ControlExpert and WDR, among others. Both partners are developing a use case in the project. The AI Bundesverband is responsible for the communication of the project.
European companies will have the opportunity to further develop innovative voice services without disclosing sensitive company or customer data
"With OpenGPT-X, European companies get the opportunity to access and develop innovative language services beyond English as needed, without exposing sensitive company or customer data."explains Dr Flores-Herr. "This also makes it possible to implement individual solutions for regional, small and medium-sized enterprises.".
*Background: About two years ago, the American company OpenAI revolutionised the market for Artificial intelligence. The model has been trained with over 175 billion parameters, outperforming previous models by a factor of ten. More than 300 applications are already using the language model. These applications do not have free access to the model, but can only access an interface.
About [at] - Alexander Thamm GmbH:
The data & AI consultancy [at] - Alexander Thamm GmbH is a leader in the development and implementation of data-driven innovations and business models in German-speaking countries. The service portfolio covers the entire data journey - from the data strategy to the development of algorithms and the construction of IT architectures to maintenance and operation.The Company's own Data Academy offers training in data science, big data and artificial intelligence an. Founded in 2012 by Alexander Thamm, [at] currently employs over 300 people. The headquarters are in Munich. Other locations are Berlin, Frankfurt, Leipzig, Stuttgart and Cologne. Its clients include more than half of the DAX 30 companies.
You can find more information at www.alexanderthamm.com. Network with us on LinkedIn: https://www.linkedin.com/company/alexander-thamm-gmbh/.
For further information, background discussions or interview requests:
Christian van Alphen
Head of Communications
Alexander Thamm GmbH
Tel: +49 172 - 134 39 10