As a part of pre-class preparation, the instructor will bring several items of technology; some of these items use Artificial Intelligence to operate (Amazon Echo, Google Home) while others don’t (mini-robot, iPhone game, etc).

• Instructor poses the following question to students to open the class: What is the first word that comes to mind when you hear the term AI? (students respond on: https://www.mentimeter.com/)

• There is a class-wide conversation, discussing the following questions: what is AI? And/or what are some examples of it?

• Instructor explains what AI is at its core and notes some different applications of it. The instructor then introduces students to 4 different types of AI systems (Natural Language Processing, Machine Learning, etc) and provides real-life examples of AI technologies.

• Students play a game in which they identify whether certain technologies (the items brought in at the beginning of class) would be classified as AI, with bonus points if they can distinguish between types of AI systems (Machine Learning, Natural Language Processing, etc). Before guessing whether an item employs Artificial Intelligence, students will interact with it (talking to the Amazon Echo, playing with the robot, etc).

• Instructor shares a document with students with links to interesting websites that utilize different forms of Artificial Intelligence and applications of AI in different industries

• As class ends, each student will be asked to note down a question or something they’d like to learn about in the Artificial Intelligence field. They will do this by entering their response on https://www.mentimeter.com/.

• The instructor leads a quick activity in which students write down 2 things they’ve learned recently - these “things” can be new skills or interesting facts - and how they came to learn these things. This activity allows the instructor to segway into a quick discussion about how we, as humans, learn.

• Next, the instructor poses a question to students: How do you think a computer learns?

• Students break into groups and attempt to answer the question. They use their whiteboards to draw images and come up with steps as to how they think a computer learns.

• After they respond, the instructor goes into a discussion about how Machine Learning models work at a basic level (machine learning models use data, lecturing about the difference between traditional code and machine learning)

• Students proceed to open Teachable Machine on their computers (Teachable Machine is a website in which students can train their own Machine Learning models) and train their own basic models.

• As a class this activity is debriefed, and students explain the correlation between this exercise and the lecture that the instructor gave.

• The instructor leads students through a game in which they are given x and y values for 10 distinct data points and tasked with predicting the 11th y-value given the 11th x-value. For this game, the x-value is 'hours of sleep' and the y-value is 'height in inches.' This exercise helps bring the concept of linear regression from the abstract into an application that the students might care about. Students perform this activity in randomized groups.

• The instructor poses the following question to the students: How does this activity relate to Supervised Learning and Linear Regression?

• After students respond to the prompt, the instructor gives a brief presentation about Linear Regression.

• Finally, the class explores real-world applications of Linear Regression, such as weather predictions and box office predictions.

• Between the 3rd and 4th classes, the instructor suggests that it might be helpful for students to do 10-15 minutes of research on Decision Trees.

• As a part of pre-class preparation, the instructor will bring several items such as a wallet, an index card, a plastic pumpkin, a colored pencil a MetroCard and a ball.

• Instructor shows 6 unique objects (a wallet, an index card, a plastic pumpkin, a colored pencil, a metro card, and a ball) and asks the students how computers might classify these objects into different categories.

• Instructor then gives a brief explanation of how decision trees work and walks students through a mini-example

• In groups of 3, students now classify the 6 unique objects by building their own decision trees.

• After the activity, the instructor asks the students to debrief. Instructor poses the question: how did you decide on what basis to classify the objects? How does a computer decide on what basis to classify the information it receives?

• Instructor explains the answer to the second question

• Students and instructor discuss potential current applications of classification models and decision trees.

• Similar to the way students wrote down a list of questions at the beginning of the program, students will again write down a question and submit it to the website https://www.mentimeter.com/.

• Between the 4th and 5th classes, the instructor suggests that it might be helpful for students to do 10-15 minutes of research on what Naive Bayes classification models are .

• Write the following problem on the board:

  • “If I play basketball 6 times the last 8 Tuesdays, and if I play basketball 5 times out of the last 10 Wednesdays, am I more likely to play basketball next Tuesday or Wednesday?”

• Instructor explains how this relates to Naive Bayes classification models and what Naive Bayes classification models are briefly.

• Instructor poses the following question to students: how does a computer determine the sentiment (positive or negative) of a message that says “I am super happy!”

• Students break into groups of 4, and attempt to solve this problem . After 5 minutes, the instructor will walk through this example with the class.

• The class will discuss further applications of Naive Bayes classification systems.

• For the last 15 minutes of class, instructor shifts to a game of “AI Jeopardy. “ Students are randomly split up into groups of 4, and they compete to get the most number of points in jeopardy format. This game is designed to help students recall the information that they’ve learned in the first half of the semester (Sessions 1-6).

• The instructor presents students with the details of the case study

• The class splits into 2 groups: The Amazon group and the U.S. government group. Each group must decide how their party should move forward given the incident, and make a couple of solid proposals for the parties that they represent.

• Each group shares what they discussed with the class while the instructor asks questions about their proposals. The instructor will also ask about the technical aspect of the algorithm (does it involve linear regression or classification? If it involves classification, what type?)

• Next, the Amazon group and US government group adapt based on the other parties' proposals.

• This final question is offered to students: How do you as individuals feel about a case study like this?

• Between the 6th and 7th classes, the instructor suggests that it might be helpful for students to do 10-15 minutes of research on what Neural Networks are and/or find an application of Neural Networks that interests them.

• Instructor poses the question to students: What function of humans do neural networks replicate?

• Instructor segways into a discussion about what neurons are (in-put/out-put) and what a neural network is. In doing so, the instructor walks through a simplified example of a neural network system.

• Students are split into the same groups of 4 as they were in in Class 2 and re-open Teachable Machine. Now, students attempt to explain how the Machine Learning model used neural networks to perform image recognition.

• Instructor leads students through this process of discovering how image recognition works and what the model is doing.

• Between the 7th and 8th classes, the instructor suggests that it might be helpful for students to spend 10-15 minutes researching what Natural Language Processing is

• Students interact with chatbots/devices that use natural language processing (CleverBot, ChatGPT, myAI, Amazon Echo)

• Students are prompted to provide explanations for how this might work, and then the explanation is provided by the instructor (along with examples of NLP technologies in everyday life).

• Students assign emotion to a sentence and then observe as a computer does it as well (there is a brief demonstration of code).

• Afterward, this exercise is briefly discussed, and students attempt to explain how the programmed algorithm may have identified emotion.

• Students spend a couple of minutes considering the question: What sorts of chatbots would you create? What would you use NLP for?

• Between the 8th and 9th classes, the instructor suggests that it might be helpful for students to spend 10-15 minutes researching what Generative AI is

• Students listen to a song produced by AI that is meant to mimic Drake: https://www.youtube.com/watch?v=qguj0Ku2CJE

• Instructor poses a question to students: What function of humans does generative AI strives to replicate? Generative AI strives to replicate human creativity.

• Should people be able to take some credit for the work of generative AI? Should Drake get to take credit for that? Should a student get to take credit for an essay written by ChatGPT? This leads to a brief student-led discussion/debate.

• What do you think are some flaws with Generative AI? What are some positive elements of Generative AI? As students list out the benefits and harms of Generative AI, the instructor will simultaneously type it up. This is relevant for the next step.

• On balance, do you think generative AI will benefit our society more than it harms it? A student-focused discussion will occur here.

• At the end of the lesson, students are introduced to the Capstone Project that they will

• Students are presented with a deepfake of Morgan Freeman (https://www.youtube.com/watch?v=oxXpB9pSETo)

• Instructor poses the following question to students and splits them up into groups of 2 to answer it: What would you do with this technology?

• What if someone created a deepfake and fooled you? How would you know it was a deepfake? Students attempt to answer this question, guessing which features distinguish a deepfake from a real image.

• Instructor talks about some common facial features that often indicate whether an image is real or fake.

• The instructor shows several deep fake images of people and students notice distinctions between real images and deep fake images (however, this is often harder than it seems)

• As a class, the instructor leads students through several rounds of a game in which students distinguish between a real image of a person and a fake one. This game is called “Which face is real” (https://www.whichfaceisreal.com/index.php).

• After 3 rounds, students are split into 2 competing teams. Each team takes turns guessing if an image is real or fake. The team with the most correct answers wins.

• Students spend the last 15 minutes of class working on their Capstone Project.

• Instructor sends students a survey to complete on the course.

• Students spend the first 20 minutes working on their Capstone Project

• In the last 15 minutes, students will participate in “AI Jeopardy.”

• If needed, students continue filling out the short survey from last class and instructor talks to students about what they got out of the class.

• Each student will have 3 minutes to present their project and the work that they have done.

• If needed, students complete filling out the short survey from last class.

• Instructor and students engage in a final discussion about future projects/avenues students may want to pursue in AI-related fields