Monday, July 31, 2023

The Evolution of Learning: From College Classrooms to AI Education

I was browsing through YouTube and came across a video from a young woman from Russia. In this video, she discussed how she had applied to university and was not accepted. Then she said she was not done learning. This is not a new idea of learning outside of formal education. I did start thinking about how learning has changed over time. Here in 2023 you can go online and find what you want to learn. This could be learning how to fix your car, a hobby, etc. But, you can learn many things that are traditionally taught in a college classroom. During the Covid-19 pandemic college professors and grade school teachers discovered that many of their lesson topics had online academic articles and YouTube videos. Because they were trying to transition their classroom lessons to a remote learning environment. Why create their own videos or printed material when the material already existed online? There is nothing wrong with that, the student still received the information (in many cases) and the educator did not need to recreate the wheel. A side effect of this is that in many cases this information is sharable. I actually learned several subjects from class material of classes my friends had shared with me. In some instances, I was able to use these resources for my own classes, and I can assume my friends could have used what I shared for their classes. And no these were not the same classes, None of us were enrolled in the same college. But a history class on labor in the United States and a psychology class on industrial psychology have overlapping material. Recently I was discussing organic chemistry with someone, and they asked when I took organic chemistry. I have not taken that subject (or any college-level chemistry class), but a friend of mine did and they shared the class material with me. I continue to learn even when not in school. and there are whole YouTube channels dedicated to educating. Some have been around for years like CrashCourse and others have emerged from the pandemic. These traditional educators who had to create online material for their classes have realized that they could share this material online, and in some cases even make some money off of it.

A college degree is a key to getting past a barrier of entry. In many professions, once you have your first job in your career path, they don't care where you went to college. Some employers don't care if you have a diploma if you have relevant job experience. now this is not to say jobs that require a degree would overlook it, such as lawyers and accountants. A computer programmer with a decent portfolio and a good job history may be more valuable than a computer science graduate looking for their first job. A college degree shows more than that you learned a subject. it shows that you have learned a broad range of subjects. it also shows you have the discipline needed to get a degree. 

If I look back at the history of learning, I wonder how advancements in technology and teaching affected past barriers of entry.  Big ones that come to mind are things like the printing press. Books and the written word becomes less scarce. More people have access to the written word and are able to read. As the written word becomes more accessible, formal grammar and spelling increase. this reduces ambiguity in written communication. with so many languages and regional dialects, this still occurs to this day, but it is reduced. another advancement is transportation. As people are able to move through the world, so does knowledge. The movement of knowledge is the backbone of education. The internet is a prime example of this. Many see the internet as the end-all of this movement of knowledge. But, the internet evolves. The way knowledge is spread on the internet changes. Now the new game changer is AI, or is it?

What AI can do is amazing, but what is new is how assessable it is. What we call AI is not set in stone. Is the spelling and grammar checking I am doing right now with Grammarly AI? Was the spell check I used with Word Perfect 25 years ago AI? What has changed is the data set that these new products have. Another new thing is the level of automation. If I wanted I could take this blog when I am done writing it and feed it into Bard or ChatGPT and have it rewritten to be more SEO-friendly. I do this with my YouTube titles and descriptions. I have written whole blogs with these websites while I tested them.  The next question is can AI produce consumable knowledge? As I have tested these two sites I have discovered that when it does not have the information it could do two different things. it could tell you that it is beyond its capabilities. or it could make stuff up.  If this is not a disturbing human trait that AI has learned, I don't know what is. Sometimes the incorrect information is clearly incorrect and it is clear that the AI misunderstood the request. the biggest thing it makes up is references. it can make formal citations look legit. But, if you don't check the primary source you may be in for a rude awaking.  sometimes it is dead links, other times it points to something that is not relevant.  Will we get to the point where AI is teaching us (reliably)? and will we lose the ability to teach? I am not even discussing the loss of employment of educators. AI can assist in distributing knowledge, but I don't feel it can generate new knowledge.  Will AI search for a new insect and discover how it interacts with its environment? or explore the universe?  I do not doubt that AI can be programmed to use a telescope and find new celestial bodies, but will it be able to analyze what it means?

So how do you learn?

How do you see learning evolving?