ChatGPT Makes Everything Easier

The Ancient Maya App incorporates data from four different sources: a drawing by Schele, glyphs and definitions from Carter, glyphs and definitions from Tokovinine and definitions from Kettunen and Helmke. I only ask researchers for permission to use their material. I never ask for data a specific format, I use what they have previously made public. In the case of material from Tokovinine and Kettunen and Helmke, that was their pdf files. Converting information from a pdf into structured data useable by an app too a lot of time. Most of the time to create a new feature in the app was spent converting the pdf into some structured data. Actually writing some code for the feature was always the easy part.

With ChatGPT-4 that is no longer true. For example, I created a 39 page pdf containing just the dictionary from Kettunen and Helmke starting on page 88. Now I just upload that to ChatGPT and say:

This file contains a large, multi-page table. Can you covert the large table into a single json object containing an array of dictionaries? The keys to each dictionary should match the table headings.

ChatGPT just completes the request as if the task was trivial. (I expect I could have uploaded the 130 page pdf and specified the page numbers the containing the table. I test that when I get a chance.) Creating endless digital artifacts and converting the format of digital artifacts is now trivial. My ability to add new resources to the Ancient Maya App suddenly became so much easier. And it will also help me write the code for the app. We all need to reset our expectations regarding how quickly we can extend our app.

Leave a Comment

Your email address will not be published. Required fields are marked *