My mission started with research, but I already ran into problems with how historical content can be hard to digest and understand. Historians have a small addiction to throwing more names and dates at the reader than they can handle! Regardless, I did several hours’ worth of reading until I knew everything, but it was all an incoherent mess in my head.
Generally speaking, humans learn best through social interactions, an exchange of input and output (hence why video games are such a potent teaching tool). A conversation would be my best bet for straightening out the details, which is why I turned to ChatGPT. Rather than make it teach me everything directly, I instead tried to teach it about the siege, and then have it correct me where I was wrong. This led to a productive exchange of words, giving me new lines of research and helping to sort the jumbled mess in my head. Pretty soon I was able to describe the Siege of Jerusalem in a variety of ways without violating the filter of ChatGPT or the other documents that I read.
I now had all the information I needed to begin writing. I made a bulleted list of topics I needed to cover and dove straight into it. I tried to make dialogue that felt as organic as possible, and whenever the conversation started to slow down I found a good opportunity to transition to the next chunk of information I needed to deliver. By this point my understanding of the history was so thorough that I could weave historical facts into the dialogue seamlessly without needing to double check my notes every two seconds.
After double checking what I wrote, I sent it off for approval by the history professor…and it was good! He tweaked the wording of some things, and there was a blunder where I said ‘our god’ instead of just ‘god,’ but other than that, my content was solid and historically accurate!
Darren Walsh 