The Auschwitz Project

    The Auschwitz Project was made possible by applied generative artificial intelligence. Here's what that means:

  • Generative AI is a type of machine learning that can generate new content based on a given input. For example, a generative AI model can be trained to write a new sentence based on a given prompt. The model will then generate a new sentence that is similar to the prompt.
  • Generative AI models are trained on large amounts of data. The more data a model is trained on, the better it will be able to generate new content. For example, a model trained on the entire works of Shakespeare will be able to generate new sentences that are similar to Shakespeare's writing.
  • Applied generative AI is a type of generative AI that is used to solve real-world problems. For example, a generative AI model can be trained to write a new sentence based on a given prompt. The model will then generate a new sentence that is similar to the prompt.
  • The Auschwitz Project uses applied generative AI to generate new content based on the primary accounts of Auschwitz. The model is trained on the provided corpus of primary accounts of Auschwitz. The model is then used to generate new content that is mathematically similar to the primary accounts of Auschwitz to provide an accurate, 360° picture of the commandant's testimony coupled with primary accounts to back it up and sources. The model was ordered not to use sources that it was not trained on (primary testimony), and most of its answers are verifiably accurate.
  • This is powered by GPT-4 API from OpenAI. GPT-4 is a large language model that is trained on a variety of data. Embeddings are a type of representation of text that is used to represent the meaning of a given text. The embeddings are then used to generate new content that is similar to the primary accounts of Auschwitz.
  • The list above, along with much of the project's code, was largely generated by AI.

    • The use of applied AI in history is intriguing and has the potential to revolutionize the field. Here's why:

  • AI can analyze large amounts of content in any language and provide summaries, answer questions, and generate new content. In The Auschwitz Project, this allowed us to analyze around 50 primary accounts accurately without reading many.
  • AI can find and fix errors in text and translate documents. AI can also write code, teach students, generate images, automate tasks, and solve disputes
  • AI can discover new information that is hidden in large amounts of data and find patterns. In scientific research, AI is being used to discover drugs.

    • The Auschwitz Project is a work in progress. We are constantly improving the project and adding new features. If you have any questions, comments, or concerns, please add an issue on GitHub. For more information regarding acknowledgements, licencing, and the primary sources, go to the README on GitHub. This project is not affiliated with the Auschwitz Memorial and Museum, the United States Holocaust Memorial Museum, or David Boder. The project is not intended to be used for research or academic purposes without fact-checking. Please be aware that the content of the AI generated prompts may be disturbing.

This website was made by Luke Kosner and is available for use under the MIT License. View the source code here.