Introduction to AI

There are several concerns mentioned in the video and I appreciate the article that has links to resources for the classroom. I hear about the concern of plagiarism the most and I echo the question of how well students can be trained to think when AI is doing some of the thinking for them. We don't know what the impact will be yet- or if there will be an impact. I often worry about bias. It can be really easy to assume AI productions are objective and miss any present nuance, especially if it is about people or groups. 

One important ethical consideration is making sure students understand how AI works and how to use it responsibly. Because of student privacy rules, I can’t give them access to many of the most popular AI tools at school. I know they often try them at home, but without any real guidance on the ethical side or how to use them appropriately. That gap worries me, because students need more than just access to technology; they need the skills to use it thoughtfully and responsibly.

LLMs (AI Large Language Models) tend to project a high degree of confidence even when they are completely wrong. As a science teacher, that worries me because I want my students to learn how to think critically rather than accept explanation authoritatively. I’ve started using AI in class as a way to model how to interface with its different tools. Sometimes I’d show them an AI-generated answer and ask, “Does this actually make sense?” It’s can offer an engaging way to get them thinking both about the question itself and how answers are generating by these different tools.

One ethical consideration that stood out to me instantly when watching the video is the idea of how these AI models are trained.  As teachers, we teach our students that they need to be mindful of what they read on the internet.  We teach them that not everything they read or see is evidence based or factual, especially now that AI is so heavily used.  What took me aback in the video was this, "they start with how these [AI] models are trained.  They're trained on everything on the internet."  This opened a whole can of worms for me.  AI models are using millions of sources - some evidence based, some copywritten, some written by Jane Doe who has a very strong bias.  As an educator, I encourage my students to look at both sides of an argument, a situation, a topic and evaluate without a bias using text or sources that are evidence based and factual.  However, using AI could hinder this.  AI could provide sources that are not evidence based and factual with a heavily one sided bias without students being able to realize this. 

In previous years AI was really just used for a behavior predictor. Now AI is used for all sorts of things like information, art, video and audio. In 2022 Chat GPT actually became the first large model that was able to do tests. AI is going to have a big impact on educators. 

There are ethical considerations that educators must use when implementing AI. First off they need to realize that AI has to watch for copyright, so sometimes it cannot give you the exact correct information because that information is copyrighted. Also, AI is full of biases. For example if you ask it to create an image of an entrepreneur, it is more likely to produce an intelligent looking man, than a woman and this is encouraging stereotyping and biases. Also, we need to think about AI doing the work for us, and at what point is it no longer a tool to help us, but to just do all of the thinking for us.