The Swiss AI Initiative aims to develop and train transparent, open-source large language models (LLMs) that meet legal, ethical, and scientific criteria. The initiative will use ten million GPU hours on the Alps computer over the next year to develop these models. The initiative also aims to bring together science, industry, and politics to advance the use of AI in Switzerland, with a focus on areas such as robotics, medicine, climate sciences, and diagnostics. The initiative will also support start-ups in the AI sector and aims to transfer the developed software infrastructure and expertise to society and industry.
Key takeaways:
- ETH Zurich and EPFL have launched the “Swiss AI Initiative” to position Switzerland as a global leader in the development and implementation of transparent and reliable AI. The initiative is supported by the new Alps supercomputer, one of the world's most powerful computers, based at the Swiss National Supercomputing Centre.
- The initiative aims to develop and train new large language models (LLM) that are transparent, comprehensible, and meet legal, ethical, and scientific criteria. The Swiss AI Initiative emphasizes transparency and Open Source, allowing everyone to understand how the models were trained and how results are recovered.
- The Swiss AI Initiative will use the Alps supercomputer to develop new, industry-specific AI base models for use in different areas such as robotics, medicine, climate sciences, or diagnostics. The initiative also aims to explore fundamental questions in the development and use of LLM models.
- The Swiss AI Initiative aims to bring together science, industry, and politics to collaborate in shaping and driving forward the development and use of AI in Switzerland. The software infrastructure, expertise, and base models developed should be transferable to society and industry, with a particular focus on supporting start-ups in the area of artificial intelligence.