What are some of the major limitations of GenAI?
The biggest limitation of GenAI is its weak ethical and moral decision making. AI does not understand right from wrong as it relies on training data, this data could be manipulated by human biases which can lead to problematic outputs. Another limitation I find is that it lacks originality and creativity since it just recombines existing patterns rather than producing new ideas.
This video also highlights how AI still struggles with creativity and ethical decision-making. It shows that while AI can generate answers quickly, it doesn’t actually understand meaning the way humans do, which reinforces the limitations I discussed above.
Possible use cases for GenAI in school settings
I have tutored Calculus and Pre-Calculus to high school students so my answers would be catered towards them. Students could use AI to get step by step explanations for topics such as derivatives and integrals when they’re stuck, especially outside classroom hours. It can also generate extra practice problems at different difficulty levels, which helps with exam prep. Here is gemini solving a typical derivative problem at the 12th grade level perfectly with explanation too.
From a teacher’s perspective, it could help create worksheets, quizzes, or even visual explanations of graphs and functions. That being said, it shouldn’t replace the problem solving process, because struggling through math is how students actually learn in my experience.
Thanks, Hazza. Another good post on AI. You are catching up well!