After going through the many documents assigned for today’s reading, I still cannot help myself from having a negative opinion about the use of AI, especially in history education. I strongly believe that generative AI (creating information, images, and videos) is harmful to education, as it can allow anyone to create false information and spread it online. Additionally, the “ideas” that chatbots create are not their own, but instead generated by aggregated data from all other users of the chatbots, as well as online sources, it quickly analyzed before spitting out a response.
Regarding its use in historical analysis and education, I do think a lot of the cons outweigh the pros. We have already seen in the past decade the intense damage that social media and unregulated content creation can have on people, such as on Wikipedia and other websites. People being able to edit factual information and consider themselves experts on topics without proper academic training and instead utilizing online information, is ultimately more detrimental than beneficial in the far future.
However, I have also begun to see the positive applications of AI in specific aspects of historical analysis. This is mainly its use in transcription. In our group, we used Gemini to transcribe a file from the digital archives of UMW, and it did an impeccable job analyzing and transcribing the handwritten document. It didn’t add any new information that was untrue or unrelated. It completed the task and then asked if there was anything else for it to do.
I think this is the sole way AI should be utilized when it comes to education, because it relies on AI’s main strength, which is pattern recognition. It can analyze the structure and data of anything given to it and then give it back in an easier format to read. If we just asked it what it thought of the document, it would create false information or pull from outside sources. By narrowing an AI’s focus to a specific task, it can be a lot more beneficial to historians in the modern age, rather than being a terrifying concept. Giving AI unlimited access to anything it wants is the wrong way to go, but structuring it to have a sole focus and task, like any other machine, will allow us to continue the process of historical analysis unimpeded by outside sources.