Week Two: Artificial Intelligence and Implications for Digital History

Within the past two decades, the complexity and ability presented to the academic and the rising scholar by the evolving capabilities of artificial intelligence starts a discussion on the appropriate nature in which chat algorithms and text generating programs find themselves in academia. This week I would like to take a look at the growing nature and usage of artificial intelligence in the world of digital history and how it has been discussed within the pursuit of the professionalization of history in  the digital realm. 

What is AI capable of?

Firstly, understanding the power and capabilities presented by text generating artificial intelligence is important as it defines the basis of our understanding of how the programming affects the digital history community on a larger scale. Wulf Kansteiner discussed deeply how historical theory is simply unsettled knowledge, with AI finding it's ow place and formation in the scholarly community. Kansteiner believed that AI could be beneficial as it allowed historians to test large scale hypothesis by correlating subjects and finding commonalities. The issue, however, was in the idea that AI often comes up short in terms of detailed, layered information that can be cross examined and discussed; There is no hard truth and detailed analysis (Kansteiner, 2022). J. David Bolter pointed the same details out in his analysis of methodology used by AI in creating responses to prompts imputed by the user. Bolter discovered that AI stresses results over methodology, reducing ambiguity and complexity in its historical prompts (Bolter, 9). Kansteiner and Bolter, both in their own way, show how the limitations of AI demonstrate it's ability to predict and repeat, rather than develop complex thought like the human brain.  

How does this affect historical research?

Scholars have widely debated on how the usage of Artificial Intelligence could benefit and hinder work in the academic field of education and research. Most notably, its important to understand how it impacts the digital archives and digital media record. In a roundtable discussion led by the American Historical Association, Kate Crawford from USC Annenberg noted the incompleteness of the digital archives, noting how the prominence of funded, larger media sources in comparison to smaller organizations creates a bias because of it's overwhelming presence; Larger news outlets easily overwhelm smaller sources, leading AI to focus on the prominent and abundantly available sources offered, ignoring the holes still present in the digital world (American Historical Association, 2021).  Continuing on during the discussion, Matthew Jones from Columbia University comments that AI works in a way that affirms pre-conceived ideas based on the sources they pull from; Computers are non-binary, meaning that AI is only predicting how they'll finish the sentence rather than creating unique ideas (American Historical Association, 2021). Ultimately, AI poses a threat in the sense that it does not do the work for us, but rather it disrupts our work by promoting bias and misinformation in the digital realm. 

How is it being treated in the classroom?

Its no surprise that AI has found itself in the classroom as a tool that is easily misused by students on all levels of education. Multiple approaches are being taken to not only understand it's impact on the classroom, but also understand how the use of AI can be properly handled in an appropriate manner. UCF has made multiple statements on how ChatGPT is bound to be used by students and even thought its discouraged, students will be unable to help themselves. In response, teachers and faculty are asked to promote their curriculum by asking harder questions, breaking up assignments, and even discussing with students why its crucial that they learn the material (UCF, 2023). Lauren M.E. Goodlad and Samuel Baker both wrote in a separate article how they believed that the correct usage of AI in drafting and essay planning with actually help students think deeper and not rely on proper syntax; By understanding what AI can and cannot do correctly, we can use it in a way the pushes our academic goals further and higher (Baker and Goodlad, 2023). 

Ultimately, Artificial Intelligence is a force and a tool that will eventually find itself in our world of academics whether we like it or not. Knowing this, its our job to understand it's capabilities and the risk and reward factors it poses to our research and discussion in the classroom. AI may seem like a threat, but it's our understanding of it and response that truly gives it's threat credibility. 

References

"ChatGPT (and other AI) at UCF - What Faculty Need to Know," Faculty Center for Teaching and Learning, August 18, 2023

Historical Association (AHA) Roundtable), “Artificial Intelligence and Its Implications for the Present and Future of Historical Research,” June 8, 2021. Closed Captions (CC) available; check user settings. Duration: 90 Minutes.

https://www.youtube.com/watch?v=QIiuqf7dvJA

Lauren M.E. Goodlad and Samuel Baker, “How the Humanities Can Disrupt ‘AI’,” Public Books, Feb. 20, 2023

https://www.publicbooks.org/now-the-humanities-can-disrupt-ai/

THE VIEW FROM 1984: J. David Bolter, "Artificial Intelligence," Daedalus 113, no. 3 (1984): 1–18. http://www.jstor.org/stable/20024925

THE VIEW FROM 2022: Wulf Kansteiner, "Digital Doping for Historians: Can History, Memory, and Historical Theory Be Rendered Artificially Intelligent?" History and Theory 61 (2022): 119-133. https://doi.org/10.1111/hith.12282Links to an external site.