Note of the archivist
Welcome to OG AI! This profile will bring together two seemingly separate worlds: archival primary sources and new AI technology. Archives trace human history through materials that are unpublished and therefore largely unavailable for datasets used to train AI tools.
Here we will explore how AI can — and cannot — be used to help us explore the OG Collection archives. It will be a sandbox where we will see how AI can use old articles as inspiration for new content, refine OCR-processed text, and draw on contemporary resources to contextualize retro material. Meanwhile, the main OG Collection Vocal site continues to be researched, written, and curated entirely by humans.
Join us as we explore the future of archives together
The following content was created using ChatGPT, and in the interest of transparency and learning, the header represents the prompt used to create it.
How do ChatGPT and DALL-E work in simple terms?
ChatGPT and DALL-E are two different types of artificial intelligence (AI) models developed by OpenAI.
ChatGPT is a natural language processing (NLP) model that can understand and generate human-like speech. It is trained with a large amount of text data and uses a neural network to predict the next word in a sentence. This allows ChatGPT to provide coherent and realistic responses to written or spoken input.
DALL-E, on the other hand, is a computer vision model that can generate images from text descriptions. It converts the text description into a vector representation and then uses a generative neural network to produce an image that matches the input. DALL-E is trained on a large dataset of text and image pairs, which allows it to learn to generate realistic images based on text input.
Both ChatGPT and DALL-E are examples of AI models that learn from large datasets using deep learning techniques. They are capable of generating human-like speech and images, respectively, and have the potential to be used in a variety of applications, from chatbots to content creation.
What can ChatGPT do in terms of generating content based on old popular culture?
ChatGPT can generate content based on ancient popular culture by analyzing a large corpus of textual data related to that specific topic. It has been trained with a large amount of text from various sources, including books, articles, and online forums, which allows it to understand the language and terminology of ancient popular culture.
When asked a question about old popular culture, such as “What were the major fashion trends of the 1960s?”, ChatGPT can provide a coherent and informative answer based on its learned knowledge. It can also engage in conversations about old popular culture, providing context and insight into the trends and events of that time period.
While ChatGPT’s ability to generate content about ancient popular culture is impressive, it is important to note that it is not infallible and may occasionally generate inaccurate or inappropriate content. Therefore, it is important to use it in conjunction with human oversight and discretion.
How can AI be used in archival work?
AI can be used in archival work to automate and streamline tasks such as digitization, transcription, and metadata collection. Using computer vision and natural language processing models, AI can recognize and extract information from historical documents, photographs, and audiovisual materials, making it easier for archivists to organize and preserve these materials for future generations. AI can also help with tasks such as identifying duplicates, detecting damage or deterioration, and even reconstructing missing or damaged parts of a document or image. By using AI, archival institutions can improve their efficiency, accuracy, and accessibility while preserving our cultural heritage for future generations.
This blog was originally published at: https://vocal.media/futurism/explore-hidden-archives-by-using-artificial-intelligence
