sdoughltec5703

[sdoughltec5703]

Final Thoughts

Given all that you learned in the course, how are you feeling about the use of AI in education and training?

Is it the future of education?

I do not think that AI is the whole future of equation. I do think it is a big part of the fitter of education, but there will forever be things that I will hold onto as being human in education, such as science experiments and math (idk why, but I just much prefer math being taught be a teacher).

Where do you expect it to be used in the future?

I highly expect AI to be utilized as a virtual tutoring assistant and to help with personalized assignments for students. Also, NLP can be particularly useful for essay grading or grammar correction. AI has a place in education, but I do not think it should completely replace education as we know it.

What are your hopes and concerns about the use of AI for teaching and learning?

I genuinely hope that people will just use it responsibly. Using it to constantly cheat will result in students never actually learning, or teachers using it as a means to not have to help students may hurt them in the long run. 

[sdoughltec5703]

Week 14 Reflection

Given what you understand about how AI can be used for learning diagnostics and decision making in educational spaces, how do you think educators and administrative staff should employ it and why?

What are potential practical pitfalls with relying on AI for these tasks?

What are potential ethical challenges with relying on AI for these tasks?

AI can be used for learning diagnostics and decision making in educational spaces quite effectively by educators and decision making. I think that educators should employ it to see where students are at on their learning paths and meet them at that point. For example, giving students more problems of the same material if the student doesn’t yet understand it or advancing them to new materials if they are ready for it. AI shouldn’t just be used to help students who are struggling to meet the expectations, but it should also be used to help advanced students learn at the pace they would like to. Personally, one of the most impactful teachers I had from elementary school was my fifth grade math teacher who would let one of my friends and I move onto new topics and use a computer to learn things that we were interested in once we finished our assignments. It was her enabling us to take ownership of our education and further it at our own pace that really encouraged my love of learning, but if I would have had to stay at the pace of my peers my boredom in school would have likely resulted in me losing the passion for learning. AI could accomplish the same task by recommending more topics to students who have already mastered a topic. My friend and I learned some random topics as fifth graders since we were just researching random math topics that we had only heard about (I.e. Pythagorean theorem and factorials). An AI system, however, could have suggested we learn topics that made sense sequentially at that point in our educations (such as multiplying/ dividing fractions, pre-algebra), that way we could have progressed further into another math class at a quicker rate instead of just knowing random topics that wouldn’t be introduced to us for years in classes. Relying on AI to help a student who is struggling can be difficult because AI can’t necessarily hit the root cause of why the student doesn’t understand like a teacher can through conversations with a student or by analyzing their hand written work and the steps that they follow. Thus, the AI may be continually offering more practice problems to a student who will never be able to understand it in the way that it is being presented, thus having a teacher to regulate that and analyze and assist the student is important so that the student can truly progress. If a student is stuck in an infinite loop of getting the problem wrong, they are likely to grow discouraged and start to resent the subject or learning. When I used to tutor middle school students, I had one girl who hated math at the beginning because she always struggled to understand the topics. Her teachers and parents had tried having her use online platforms such as khan academy to help her, but she just couldn’t grasp the subjects. When I came in, I learned to meet her where she was at. If she couldn’t understand a topic the way her teacher or khan academy would teach it, then I would explain it in a  different way to help her understand. As she gained more confidence in her ability to learn and complete the problems without my assistance, she would get more excited to learn and her grades started to improve. She was so proud to tell me when she would get an A on tests or how her peers would ask her for help on practice problems. Where AI failed, human interaction and teaching was able to truly help a student in a sustainable way that she grew more confidence in her education and didn’t need my assistance anymore. Thus, the ethical dilemma of relying on AI for these tasks could be that it isn’t effective and can result in a student growing discouraged from learning overall. 

[sdoughltec5703]

Week 13 Reflection

Reflection Topic: How comfortable are you being tracked by AI? I do not feel comfortable at all with the idea of my employer tracking my body language, movements, and posture while I’m working. My current job is a live in position, so of course they would never try to track my movements, but I do occasionally work desk shifts in the dorm that I live in when our desk clerk calls out sick. Sometimes, those desk shifts are in the middle of the night (like 1AM to 5AM). That is a very dead time to be working a front desk, and I will admit that I have occasionally played my head down for a few minutes while working that shift, and I feel like that is something that an AI would flag when monitoring an employee. The idea of an employer being so nit picky as to monitor eye movement and the employees gaze sounds completely dystopian. If you’re so concerned about an employee having a minute of not being productive, maybe you just need to relax a little bit and recognize that during a log shift, a person is going to need moments of breaks here and there, which means their gaze may fall or wander. That being said, there are purposes that I feel may justify the use of AI to track employee movements. For example, if someone works in a high risk, very hazardous work environment, then having movements tracked to ensure employee safety would be understandable, especially if it’s an environment where small things being off can cause big consequences. There is also the consideration of employee privacy. If an employer is upfront about tracking employees movements, then I would be more inclined to be okay with it, but if an employer does it secretly to try and catch an employee or something, that is when I feel it crosses a line. Informed consent should definitely be obtained before any employer decides to record their employees for the purpose of tracking specific movements via AI. But overall, I would lean towards the side of saying that I am not comfortable with my employee using AI to record and track my movements. There is always the concerns of not knowing what it may be used for. Is the employer recording and keeping the records for a while or are only the “flagged” movements stored while everything else is immediately deleted? Will employers use the data in performance reviews and use it as grounds of firing employees or use it in some way to harm an employee? This is such a grey area and there are many caveats and steps that need to be completed for it to even feel remotely okay, so for the most part I feel that it is better for employers to stay away from using AI to track employee movements, gaze, posture, et cetera. When looking more specifically at education, I tend to agree that student movements shouldn’t be tracked. That being said, of course there are pre existing programs such as Lockdown Browser that use AI to detect and track student movement and students aren’t given a choice to use the program or not, we have to. When I struggle to answer a test questioner comprehend it, I will say the question to myself but I have to refrain from doing that out of fear of my test getting flagged because I spoke during th exam.

[sdoughltec5703]

Week 12 Reflection

AI supported sentiment analysis can be incredibly helpful in determining users and authors opinions towards a topic/ program. Working for the TAMS program, I am around high schoolers quite often and will hear a lot of complaints towards multiple TAMS policies. For example, the requirement to take up to Calc 2, being limited to applying to 15 colleges, and the wifi for our building cutting out at 1AM. These are all valid complaints, but I think that typically we only hear the people who feel negative and those who don’t mind the policies often don’t speak out about them, hence why I only know the negatives. If TAMS admin were to roll out a survey to gather student opinions towards different policies such as those listed above, and assuming a 100% response rate, then the admin could see if students tend to genuinely feel negative towards these different policies or if it’s a loud minority that feels negatively towards the policies. That being said, the results of a sentiment analysis tool could effectively guide the decision making process as to whether or not these policies are beneficial or harmful to the student body. Since I work for TAMS but also regularly have conversations with the students (a side effect of living where I work), I am able to hear both the admin and the student perspective on these things. In regards to calculus, the student complain the their average high school peers don’t have to go up to calc 2, as well as many college make them take calculus again after TAMS. Admin argues that it shows how much stringer the TAMS education is when it’s a standard for their students to be high achieving in their level of math. The 15 schools college application limit is always a major point of contention around October/ November as the seniors/ second year students start wanting to last minute apply to extra schools in fear of being rejected from their top picks. The admin push back is that this limit pushes students to be more mindful of what schools they are choosing to apply to and making sure that they have a healthy balance of safety schools, targets, and reach schools, as well as applying to that many schools can be expensive for families and time consuming for the academic counselors to submit that many transcripts/ letters of recommendation. Lastly, the wifi cutting out in our building at 1AM is a very nuanced issue among students and staff. The students complain that it prevents them from being able to finish assignments, but the staff argue that it promotes a healthier sleep schedule and if students really need extra time to complete an assignment that they can wake up at 6AM when the wifi turns back on. This issue gets more nuanced when you take into consideration that many students have cellular data and hotspots that they can use to continue working, and that this policy can disproportionally affect student from lower socioeconomic backgrounds, so this policy is even getting the criticism of some staff now. Overall, the many TAMS policies that are points of contention could benefit from a survey to gather student opinions and then responses could be analyzed using a sentiment analysis tool. There are even many group chats that students use to talk and those could be analyzed to see what the actual opinions are instead of the refined thoughts that are put onto a survey response. I genuinely believe that this form of analysis would greatly benefit the TAMS students, as well as the admin who are constantly having to field complaints from TAMS students. I feel that this is a rather simple sentiment analysis applications o I definitely think that currently available tools would be able to properly handle the task. 

[sdoughltec5703]

Reflection Week 11 (?)

I touched on this a bit last week, but I feel that learning diagnostics can be implemented well in classes as a means of determining a students understanding of a topic for the most part. Neuro typical students can more easily be tested in traditional ways, including a computer based test that analyzes their data without having to consider too many outside factors. That being said, neuro divergent students would not be able to be effectively analyzed in the same way, and AI based learning diagnostics won’t be able to produce as accurate or relevant of results/ diagnostics due to the more unique minds/ learning styles of those students. For the most part, I do think that AI based diagnostics are useful and should be implemented assuming they can be equally implemented across all schools, as it can better asses if a class has a good understanding of a topic and can help the decision making process as to if students are ready to move on to the next topic or not. In terms of administrative staff, learning diagnostics can be used to see what teachers may yield better test scores, assuming student are given the same test and graded the same. This can help guide decisions in terms of determining what is a better teaching style, but of course there are other factors to be considered, such as it wouldn’t be fair to compare a gifted and talented (GT) class to a regular class, as the teachers wouldn’t be the difference so much as the students abilities and known performance differences. Everything with AI needs to have much initial thought into its implementation in my opinion. You can’t just freely go about saying that everything can be properly analyzed and decided upon via an artificial intelligences decision making skills, as AI only knows how to consider the variables that it is told to consider. Humans must aid the AI by determining what can even be analyzed properly by AI vs. what offers too many confounding variables that an AI wouldn’t know how to properly analyze and make useful and helpful decisions regarding the information/ population. At the administrative level, AI can determine what teachers are most effective and make the decision that maybe most teachers should aim to model their classes in the way that the supposed “more effective” teacher does, since it is seemingly a good teaching/ classroom method. AI can also be used administratively to look at student attendance and maybe find patterns in what days/ times are more common for students to skip classes, and then administration can look at this information to help them guide potential schedule changes for the year. Another perspective to changing schedules is looking at what time of day student perform best in classes, and make the harder/ core classes be during those times and set electives during the less effective times. The using AI, there is always the ethical concern that computers can’t see the full picture of a student/ staff/ administration and may arbitrarily make a decision that isn’t fully informed and thus harmful.

[sdoughltec5703]

Reflection Week 10

Having worked as a virtual learning facilitator for elementary school students at a daycare back in Fall 2020 when Covid forced many schools to start out the semester online, I was able to see a lot about how students were assigned different materials and moved through different topics based off of how well they were learning using different virtual, AI supported systems. Hard decisions had to be made with this information, as teachers had to have conversations with parents that their students weren’t able to keep up and I would often be in the middle of that conversation since the parents were expecting me to be making sure their students kept up in their classes. Students who could read would often be assigned self paced assignments on different AI based learning platform like ImagineMath, Epic, and DreamBox. If a student couldn’t even read then they typically couldn’t be assigned work on these platforms and the teachers would have to spend the whole day going through assignments and teaching these students over zoom (typically kindergarteners and some 1st graders). We even had a second grade student that had to be moved back to first grade because she was continually struggling through her different AI based learning platforms and the teachers felt she just wasn’t well prepared for the learning level of the grade. This decision was guided by AI based learning analytics that showed this student just could not master her topics no matter how hard she tried. One spot where this is a drawback is that, if I was doing what that students parent seemed to want me to do, I would have done all of that students assignments for her so that she would pass and her lack of understanding never would have been caught. So AI was able to catch the students short comings, but AI could have been easily faked  into believing that the student had full mastery of the subjects and topics. This shows how AI analytics data can make decisions about a specific students education, and how easily it could be faked, which is always a concern with any online education. Curriculum decisions can be made and changed by AI supported learning analytics by telling educators what areas students may be struggling with at the current instructional amount so that educators can add in more instruction for students moving forward. AI supported learning analytics are able to find trends in what areas student struggle in and highlight it so that the issue can be addressed. Instructional methods effectiveness can also be mastered if an educator gives instruction via two different methods on the same topics to two different classes, such as using mixed media instruction for one class and just lecturing the other and then seeing how the two classes perform differently AI supported assessing platforms. Then, based off of how the scores differentiate and the results yielded by the AI supported learning analytics system an educator could see what teaching method or method of instruction was more effective. 

[sdoughltec5703]

Week 8 Reflection

I think that AI based adaptive learning is a great concept and should definitely remain an option for students with learning disabilities or those who may need extra help. The ability to slow down and gear the path of learning and activities towards the students needs is very crucial if someone is falling behind and has been identified to need extra help for various potential reasons. However, I feel that for the most part having to put in extra work when you don’t understand something is more beneficial to a student than just being handed an alternative option so that you don’t have to work through the struggle. I whole heartedly believe that adversity makes people stronger, and the class that I feel I took the most away from in high school was a math class to where I would consistently perform below my standards on every single quiz. Then I would put in the extra work and go over every problem before retaking the quiz and getting a 100% almost every time. Adversity is what made me learn the materials more wholly and push myself to perform better. If I would have been given an out through an alternative assignment I may not have learned the materials as well and wouldn’t have had to push myself. I know this opinion can be unpopular, and I do think that adaptive learning can be very beneficial in instances of learning and mental disabilities. For the most part, however, I feel that AI based adaptive learning being the default for everyone will prevent students from being able to grow on their own and become stronger and better learners by persevering through adversity. The teachers role in this situation will also be greatly minimized. If the AI is the one coming up with different ways to teach a student, the teacher isn’t having that interaction with the students to try and help guide them towards an answer. The teacher may do the initial round of teaching but then let students loose to do practice and master content on the computer. The teacher becomes more of an overseer than an active participant in the students education. I used to tutor 5 middle school students in math. I had to develop my own adaptive teaching to figure out how to most effectively help each individual student learn the concept, be able to do simple practice problems, and then master it and have the confidence in their skill to be able to do problems without wanting to check the answer with me every single time. I had one student that would learn best if I let them struggle through the problem at first without guidance so that they could try to learn on their own. Another student, however, would shutdown if I tried to let them figure out a new concept on their own. The second student needed to be taught how to do the problem, but then could do practice problems just fine after being explained to how to do it.

[sdoughltec5703]

Week 7 Reflection

The first commercial product that uses AI-based Natural Language Processing is Amazon’s Alexa. The main features that are highlighted on a quick google search thereof are “smart home, productivity, shopping, entertainment, communications, news, routines, kitchen, fun and games, and bedroom”. There are many features of an Alexa to be excited about. For example, the integration of smart home and routine features allows you to tell Alexa to turn on the lights of a room at a certain time each day, or play music at a specific time. This can greatly help families with kids establish better routines that are easy to follow with very set visual and audio cues for things. Alexa can also “play games” with kids. For example, I know I used to do MadLibs with my Alexa when I would cook. Things to be more worried about is the ease of accessibility of Alexa. To expand on that, I’m going to use an example that I found on social media. A mom would order things off of Amazon by just telling her Alexa to order it. Her child learned that that was a possibility and started to also order things off of the Alexa like toys. Luckily Amazon has a safe guard in place for that so that someone can disable the voice setting, but that general ease of access is frustrating and it would be cool if Amazon implemented a specific voice recognition software into Alexa so that only set voices can order things. But overall, Alexa uses Natural Language Processing to understand what it is being asked or told to complete a task or answer a question. If I asked Alexa what the weather is outside it would search the internet and tell me. If I told Alexa to add Apples to my shopping list, it could. Alexa is able to to do things easily using NLP and AI. In terms of being applied to the scope of education, Alexa is useful in that it is able to gather information for students in a hands free manner. If I were writing a history paper and needed to know the year of a specific battle super quick, I could just ask Alexa instead of having to google it myself. Looking something up is quick and simple, but then I avoid clicking into Google and inevitably getting sidetracked by a million other things on the internet. Also, the Mad Libs feature could be implemented in elementary grammar classes as a way to help students learn parts of speech. Alexa will start a story and then ask for a word in a specific part of speech (for example, a noun or a verb). If you don’t know what the part of speech is, you can ask for an example so you don’t accidentally give a word that won’t make sense. After Alexa has asked for a number of words, it reads the whole story back to you with your silly words inserted, and assuming you followed the parts of speech correctly, the story will be grammatically correct. 

[sdoughltec5703]

Reflection Week 6

One technology interface that I feel could benefit from the use and implementation of Artificial Intelligence is ANKI. My boyfriend is currently studying for the MCAT, and he uses ANKI quite often for that. ANKI is a platform similar to Quizlet to where users can create notecards to study. ANKI is better than Quizlet in that you can add media and set the cards to where you have to click parts of photos to properly identify parts as a means of studying. ANKI also allows you to take other peoples study decks and add them to your account so that you cansptudy them and even add cards to the study deck if there is something you feel is lacking that you need for yourself. For example, since my boyfriend is studying for the MCAT he uses a very detailed study deck that someone has already made that covers as a lot of MCAT topics. This is useful because for something like the MCAT, there are a lot of terms that need to be studied and being able to use someone else’s pre made study deck helps to cut the workload down significantly. One way that ANKI could benefit from the inclusion of AI is by using it to automatically sort  notecards based nohow well the students knew the information. Currently, the platform prompts users today how they felt about a card (they can say it felt good, and the card won’t be shown for 24 hours, they felt okay and the card will be shown again in 8 hours, or they did not feel good and the card will be shown again ASAP). Asking students to self rate themselves doesn’t feel very intuitive as students may say they felt good even if they struggled a little. ANKI could prompt students to type out the answer for a card, and if they get it all correct then it falls under the it felt good category. If they took more than a set time (i.e. a minute) then it could go under the it felt okay category. And lastly if it took more than 5 minutes or they got it wrong it could go under the it did not feel good category). In this date and age technology can easily be taught and trained to do simple tasks like these and I feel it would benefit students. Additionally, it could be trained to identify trends of missed materiel. For example, if someone is constantly getting questions regarding endocrine disorders wrong or taking longer to answer them then ANKI could suggest for the user to review that specific topic or even pull suggested links from the internet for the users to look at so they can more adequately fill the gaps in their learning and knowledge. Anki could also potentially incorporate some kind of feature to where it will suggest note cards/ content based off of what other users have in decks that they’ve created that have similar content or information as your deck. Overall, I think ANKI is well set up as a notecard/ studying platform, however I think that it could incorporate artificial intelligence to leverage ANKI into being even more top notch and more helpful for students by making sure they are learning efficiently and studying all of the information that they could possibly need. 

[sdoughltec5703]

Week 4 Reflection

As I have perceived it and interacted with it, AI is being applied to education in many ways. Such as being used for personalization or to automate otherwise tedious tasks. It is also being used as a resource for students to find answers more easily, which can be a slippery slope into the idea of cheating, particularly in ChatGPT, where a student can have ChatGPT write an essay or short answer response for them and turn it in as their own. This particular application of AI in education may not be good, however for the most part the others are. Automating tedious tasks, such as checking grammar or verifying plagiarism are very beneficial applications of AI in education! Previously, teachers may have not been able to catch plagiarism that wasn’t blatant, but now there are platforms that can catch it as well as tell the teacher where the plagiarized parts came from. That being said, teachers should not just blindly trust what the platform tells them is plagiarism, as I once had an assignment come back with a relatively high plagiarism score because it flagged all of the questions on the assignment as plagiarism (specifically as having been plagiarized from students at my high school who all did the same assignment). This supports teaching as it enables students to be able to be more thorough in their checking of students without having to do an extensive amount of extra work (I.e. just have your students turn their assignment in through TurnitIn instead of the default system). I would say that there is strong evidence that plagiarism checkers are good at their jobs, but just occasionally need to be double checked by teachers as they aren’t always the best at filtering out properly cited sources as not being plagiarized. Another strong application of AI in education is how it can be used to personalize an educational experience for students. It can be used in a test to give students questions that are to their level of understanding, it can be used in the classroom for general teaching to give students materials for learning that are more likely to help them where they are at (if someone is still mastering multiplication, they would be given more multiplication problems, not moved on to division). The platform IXL uses personalization based on skill on their problems. If you get a question wrong, it will give you easier questions until you master that skill and move on to harder skill questions. This supports learning as it allows a student to learn where they are at instead of being forced to keep moving through topics they aren’t ready for, and this helps teachers to be more effective by knowing what students may need more specific help with. There are definitely concerns of AI in education becoming too much, and completely replacing the traditional classroom. I think it is very important for students to still have face to face learning and actions and conversations with humans should be a foundation to education, especially in early years as they develop their social skills. Of course there is also the concern of cheating that was discussed above in regards to ChatGPT. Overall though, the benefits of AI in education outweigh the concerns and just needs to be used wisely so it doesn’t oversaturate the learning experience for students. 

[sdoughltec5703]

Reflection Week 2

Your definition of AI in Education?

I feel it is hard to give Artificial Intelligence in education one set definition. It is a very multifaceted topic being implemented in a multitude of ways to make education more effective as well as more efficient. From making learning more engaging, to personalizations so that a student is not left behind on a topic, all the way to even the academic administration, AI has been applied to all sides of education to maximize what a student can get out of their education. AI in education is, at its base, the application of machine learning to a topic in education that can look at data and identify patterns, and use those patterns to more efficiently complete a task or offer an insight into the data that a human could not easily identify themselves. As explored in the reading, ‘How AI is Transforming the Future of Learning’, I was able to see how AI has been implemented in academic administration to make transferring community college credits a more seamless process, which is an application I would not have previously thought of as being under the scope of AI in Education. 

Can something be described as AI without machine learning done by the technology independent of a human?

I feel that something can still be considered artificial intelligence even if the machine learning requires human assistance or intervention. So long as the findings are still deduced using a machine learning program then it is AI. The technology was used to parse through data and produce whatever product it was created to do. All AI and machine learning projects start with a human being the one to code them, so in theory AI is not completely independent of humans. Many datasets are human created and human supplied datasets, so that is another level of human involvement in the AI. Without human interaction there would be nothing for the machine to be trained on in order to learn and be able to act independently of a human. Thus I feel that something can still be described as AI even if the machines learning was not entirely independent of human interaction/ involvement. 

What qualifies and what does not?

There is not a hard set line of what can be considered AI or not. Any machine/ technology that acts wholly independent of human involvement after its initial coding would blatantly fall under the scope of AI. That being said, as discussed in the previous point, something isnt disqualified from being considered AI if it requires further human involvement. For example, my Photos app has the ability to scan through photos and group together images by person. That being said, it will often ask me if a specific photo is of me and there have been times that the photo was actually my mom or a friend. I would still consider the machine that can sort through and identify and differentiate between faces to be AI, even if it requires my approval and interaction to be accurate. I am furthering its training and that is okay. Just as humans are never done learning, machines are not ever done learning. Thus, I would say any machine trained on data that can properly complete a task without always requiring human correction can be considered AI, but we have to give the machines grace in knowing that sometimes they need guidance and redirection just as humans do, and that need for guidance doesn’t make them any less of an ‘AI’.

[sdoughltec5703]

Reflection Week 1

I am interested in AI as it ties into the field I am going into, computational linguistics. In addition to my interest in computational linguistics, I have worked with children in many different settings. Educationally, I have tutored middle school students in person and virtually, I have facilitated virtual learning for students during Covid, and I currently sub at schools in Denton ISD. Additionally I have worked at summer camps and in a daycare. All of that to say, I have exposure to many sides of K-12 education and children interactions. Through all of that I have been able to see how children positively and negatively interact with technology. AI also has so many benefits to education. AI can help with student personalization, tracking specific student trends and needs by seeing where they are slacking or excelling. AI can also help students current essays before submitting them, telling them what grammatical or spelling issues need to be fixed. AI can also be effective in early childhood education by making learning engaging and exciting for children through educational games. Assistive technology can help special needs students to be able to communicate and learn more effectively. AI has been widely implemented to help identify plagiarism in texts and student submissions, which is something that would have been harder to keep in check previously. Specifically for learning and teaching I think AI can be tailored to help with grading and looking at learning analytics and track specific student improvement. That all being said, AI is not flawless in things such as sentiment analysis or gauging specific rubric checkpoints. So one of the potential concerns would be that if a teacher is relying solely on an AI to grade something, it won’t be the most accurate or reflective grade. AI and technology can have the negative impact in that gaming learning technologies can be addictive to children or make their attention spans smaller so that students find it harder to pay attention to teachers when needed. Another potential concern for the application of AI in learning and teaching is ChatGPT. It is a recent advancement in the field of AI and it is very controversial and brings up many potential ethical dilemmas. Students can use ChatGPT to write their papers and take written tests for them and a professor/ teacher wouldn’t necessarily know right away. ChatGPT is trained off of internet sources but isn’t always correct. When it is trained off a variety of sources it can have conflicting content being brought in. That being said, ChatGPT can offer sources if you ask it to cite its sources. There is also the issue of ChatGPT potentially training itself later on off of other sources that were created from ChatGPT. ChatGPT is just a new and growing AI that’s scope is currently unchecked and will cause a lot of potential ethical issues. Another potential ethical issue of AI in learning and teaching is that students will lack privacy, which can be a huge issue, especially in students under 18. AI typically needs unrestricted access to student information to be most effective, and that is something that can limit a students access to educational privacy. Lastly, there is the fear that a students reliance on AI and technology will limit their own abilities.