Artificial Intelligence is here to stay. It has yet to be determined what the full impact will be on humanity. Similarly, no one foresaw the impact social media, precisely Facebook, Google Search, Twitter, and related technology advances, would have on society\u2014more people complaining of increased negative engagement, cyberbullying, depression, social isolationism, etc.<\/p>\n
The question is, even if we did know what might happen, would we still push ahead for the sake of the thrill of discoveries and, of course, the money? In this post, we are to answer what are the dangers, limits, and possibilities of AI for our studies. I will address the question and then take the liberty to point out what one opportunity cost of AI will be for society.<\/p>\n
Let’s start with the three dangers that will most impact my studies. One significant shortfall of AI that Michael Webb pointed out in the ChatGPT Impact on Education video (as did the other videos and articles) was the probability of biases inherent in the data.[1]<\/a> In addition, knowing what data sets the specific AI app is using can also cause problems.[2]<\/a> And lastly, the inability to accurately recognize what algorithm produced the text response.[3]<\/a><\/p>\n I am not an expert on AI and all the research surrounding the topic. What is concerning is that research is confirming there are biases – but I need to see the corresponding examples or research that points out how to begin to recognize the biases. As I research topics in ChatGPT, the AI-produced results are convincing. It will require someone more highly skilled than myself to pinpoint the biases and how they influence outcomes, which means my data might need to be revised. Secondly, the data sets could reside anywhere worldwide – on anyone’s server farm or kitchen table. Again, not being an expert, I’d like to know where and by whom the data is aggregated for the truth’s sake. The untruths are so much more sophisticated these days that it is hard to discern the truth from a fact. My brain sometimes becomes overloaded, trying to plow through every bit of nuanced statements. The inability to recognize the algorithms is in the same category as the first two – although I will return to the algorithms later.<\/p>\n I’ve identified two limits in AI regarding how it impacts my studies. This list will change as more AI apps are produced. The first limit is that AI-produced results require one to verify and verify. Boud mentioned that the results must be trustworthy.[4]<\/a> One cannot use ChatGPT, type ahead, or word suggestions without going back and performing an in-depth review. Just today, my text type ahead took the word I was trying to type, Hhmm, and changed it to him. Different.<\/p>\n The second limitation is that there are still gaps in the knowledge base of AI. As a result, one must still use old, reliable Google Search. For example, the free version of ChatGPT is lacking in local demographic information or US government statistics.<\/p>\n I see two benefits in the current AI tools I use. Specifically, ChatGPT is very good at getting me past a “brain freeze” – those times I cannot think of another creative idea. It is advantageous to help me generate new ideas. I’ve used it to start producing outlines for online course content, seminar topics, blog content, and lesson plans. The other helpful feature is to drill down on specific research topics. The results must be verified, and sometimes the statements duplicate one another – but the AI results always spark a creative thread for me to follow. I enjoy the writing process, so I do not want any AI tool or app to take that away. But I’m grateful for the assistance.<\/p>\n Now, I will briefly discuss one of the opportunity costs associated with AI – particularly how it might be used in education. I am sold on Sal Khan’s proposal for how AI can supplement the classrooms. Although his proposal has tremendous potential to improve student learning, it also has tremendous potential to increase the digital divide. More importantly, according to a review article on AI in education, the research confirmed that AI can predict what students might be failing or on the verge of quitting and the areas of competencies the student will most likely have.[5]<\/a> On the surface, this sounds like a fantastic tool for teachers. The opportunity cost – just like with Khan’s proposal – is that there will be students who will benefit, and there will be others who will be penalized. I’m reminded of how the 3rd and 5th-grade end-of-grade tests are used to predict the prison pipeline.[6]<\/a> The ability to predict a student’s potential competencies through unknown algorithms might also lead to segregating and steering students down a career path that removes career choice from the equation.<\/p>\n One of my favorite movie trilogies, Hunger Games, is prophetic. People were assigned districts (a place to live) based on trade or what the District was responsible for manufacturing, i.e., masonry or defense, general electronics, etc.\u00a0 Too much. Maybe. But when new technologies are discovered – the benefits leave so many out of the equation. I’d like to know if an algorithm can be built to help solve this worsening digital divide.<\/p>\n [1]<\/a> Michael Webb, Chat GPT-3 and its impact on Education\ua789 Michael Webb.mp4<\/p>\n [2]<\/a> David Boud, Assessment AI 27032023.mp4<\/p>\n [3]<\/a> Lucinda McKnight, Engaging with AI writers in HE| THE Campus Learn, Share, Connect.pdf<\/p>\n [4]<\/a> David Boud, Assessment AI 27032023.mp4<\/p>\n [5]<\/a> L. Chen, P. Chen and Z. Lin, “Artificial Intelligence in Education: A Review,” in IEEE Access, vol. 8, pp. 75264-75278, 2020, doi: 10.1109\/ACCESS.2020.2988510.<\/p>\n [6]<\/a> https:\/\/www.aclu.org\/issues\/juvenile-justice\/school-prison-pipeline\/school-prison-pipeline-infographic<\/a><\/p>\n","protected":false},"excerpt":{"rendered":" Artificial Intelligence is here to stay. It has yet to be determined what the full impact will be on humanity. Similarly, no one foresaw the impact social media, precisely Facebook, Google Search, Twitter, and related technology advances, would have on society\u2014more people complaining of increased negative engagement, cyberbullying, depression, social isolationism, etc. The question is, […]<\/p>\n","protected":false},"author":167,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"_monsterinsights_skip_tracking":false,"_monsterinsights_sitenote_active":false,"_monsterinsights_sitenote_note":"","_monsterinsights_sitenote_category":0,"footnotes":""},"categories":[2309],"tags":[2817],"class_list":["post-32943","post","type-post","status-publish","format-standard","hentry","category-doctor-of-leadership","tag-ai-dlgp01","cohort-dlgp01"],"acf":[],"_links":{"self":[{"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/posts\/32943","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/users\/167"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/comments?post=32943"}],"version-history":[{"count":2,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/posts\/32943\/revisions"}],"predecessor-version":[{"id":32945,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/posts\/32943\/revisions\/32945"}],"wp:attachment":[{"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/media?parent=32943"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/categories?post=32943"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.georgefox.edu\/dlgp\/wp-json\/wp\/v2\/tags?post=32943"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}