Oak Weekly Report (2/7/25)

Hey Cougars!

Today, we dived into our first lesson: AI categorizes data through patterns. After sharing what we already knew about the topic, we started looking at how AI models function in a 3-step process: create a set of rules, train the AI model to use those rules on data, and learn from the model’s mistakes to improve the rules.

We also tried thinking like AI in a data training game. We asked students to come up with rules for recognizing dogs versus cats, and then had them sort pictures of random animals based on those categories. Would you categorize a fox as a dog or a cat (if those were the only 2 categories you knew)? How about a horse? We ended the day with a discussion about the limitations of AI’s reliance on predefined rules.

As homework, we’re asking students to experiment with giving AI rules to follow. Using AI (like ChatPGT), try to generate this image of a floating island. Remember, the more specific you are, the better your results will be. For the next meeting, bring a printout of the image you generated. Record what AI you used, what you wrote to create that specific image, and how many tries it took.