Preparation

For the AI training, you actually only need to bring your work laptop and a charger.

During the training, we will use and sometimes compare several language models from academic and commercial providers. To prepare for the course, it would therefore be advantageous if participants already had accounts with some of these providers:

• Helmholtz Blablador (academic provider)

• GWDG / Chat AI / Kisski of the Academic Cloud (academic provider)

• OpenAI / ChatGPT (Commercial provider)

• Anthropic / Claude (Commercial provider)

• You.com (Commercial provider)

• Google Gemini + NotebookLM (Commercial provider)

• Perplexity (Commercial provider)

If you are unsure about which providers you want to create an account with and which ones you don’t, you can gladly postpone this decision until the training.

Data Analysis at Helmholtz Compute Center in Jülich

If you want to execute AI-generated Python code as part of one exercise, consider logging in to the compute center in Jülich via its Jupyter Platform.

Optional additional steps

For technically interested participants, these optional steps could additionally be useful:

• Running local language models with Ollama: Install ollama on your computer. After it is installed, download the language model llama3.2 by executing this command on the command line: ollama run llama3.2:1b.