Back when #BigData was the fashionable buzz word, I repeatedly had to explain to enthusiasts that archaeological data are not just Big, they are Confused and Patchy and Hairy.
I can't really see how the current generative algorithms could make me obsolete or even speed up much of the work I do. Because I'm in this really niche activity with no commercial potential that demands constant engagement with wildly non-standardised data as well as creative writing about them.
@mrundkvist AI could help you write about the data. It's great at summarizing stuff. E.g. on the US Amazon site, AI provides a summary of the reviews at the top. When I've checked, the summaries have usually been pretty accurate.
Train it on your collected writing up to this point, then give it data on your latest excavations and tell it to write something similar. Then you become the editor, rather than the author.
@mjausson "Give it data" is a huge part of my job that is hard to automate. Finding them, selecting them and making them uniform and comparable. But the most important part is the creative writing of novel arguments for and against interpretations of those data. That part does not follow from the data themselves or from text the model was trained on. Finally, being the author is way more fun than being the editor.
Add comment