University and college professors have been using simulations for teaching and research purposes for some time now, at least since the early 1980’s when personal computers made their first appearance.
Matt, thanks for the response! And yes, you're correct. The models are trained on a lot of historical documents. A historian colleague mentioned to me a couple months ago that while that is indeed the case, massive amounts of data remain locked in the world's many archives. The Venetian archive, for example, contains about 60 km of documents, covering some 1,500 years of history. Little to none of that has even been digitized. The typical situation with my clients is that they will curate a narrow, specific set of documents and it is just those documents they want to interact with. RAG systems now allow us to do that. You make a great point!
Dan, an interesting article! One thing I wonder about is that AI is trained on modern documents. In some ways, that's true, but there are really a lot of historical documents that are also used, bringing with them their own values and concepts. Some of that is how we end up with bias in AI models that produce output that our current norms find objectionable. It's interesting to think about the different mirrors that come into play when all of this is thrown together...
Matt, thanks for the response! And yes, you're correct. The models are trained on a lot of historical documents. A historian colleague mentioned to me a couple months ago that while that is indeed the case, massive amounts of data remain locked in the world's many archives. The Venetian archive, for example, contains about 60 km of documents, covering some 1,500 years of history. Little to none of that has even been digitized. The typical situation with my clients is that they will curate a narrow, specific set of documents and it is just those documents they want to interact with. RAG systems now allow us to do that. You make a great point!
Dan, an interesting article! One thing I wonder about is that AI is trained on modern documents. In some ways, that's true, but there are really a lot of historical documents that are also used, bringing with them their own values and concepts. Some of that is how we end up with bias in AI models that produce output that our current norms find objectionable. It's interesting to think about the different mirrors that come into play when all of this is thrown together...