AI + Ethics Curricula for Middle School Youth: Lessons Learned from Three Project-Based Curricula

Abstract

Artificial Intelligence (AI) is revolutionizing many industries and becoming increasingly ubiquitous in everyday life. To empower children growing up with AI to navigate society's evolving sociotechnical context, we developed three middle school AI literacy curricula: Creative AI, Dancing with AI, and How to Train Your Robot. In this paper we discuss how we leveraged three design principles-active learning, embedded ethics, and low barriers to access - to effectively engage students in learning to create and critique AI artifacts. During the summer of 2020, we recruited and trained in-service, middle school teachers from across the United States to co-instruct online workshops with students from their schools. In the workshops, a combination of hands-on unplugged and programming activities facilitated students' understanding of AI. As students explored technical concepts in tandem with ethical ones, they developed a critical lens to better grasp how AI systems work and how they impact society. We sought to meet the specified needs of students from a range of backgrounds by minimizing the prerequisite knowledge and technology resources students needed to participate. Finally, we conclude with lessons learned and design recommendations for future AI curricula, especially for K-12 in-person and virtual learning.

Keywords: AI Literacy; Artificial Intelligence (AI); Constructionism; Curriculum; Middle-school; Online Learning.

Fig. 1

Visual representation of the student/art teacher analogy to describe the generator and discriminator

Fig. 2

Screenshot of the Teachable Machine interface, which abstracts away technical implementation details, and advanced mathematics from student users. (Image from teachablemachine.withgoogle.com website.)

Fig. 3

Most students across the three workshops were also active users of technologies that leverage AI, especially YouTube, Google Search, Email, Netflix, Tablets and Gaming Systems

Fig. 4

Students’ perceptions of capabilities of AI

Fig. 5

Terms that students used to describe AI

Fig. 6

Students’ self-perception in relation to AI technologies

Fig. 7

Students’ responses when they were asked if certain media were GAN-generated or not (they were all GAN-generated)

Fig. 8

GAN Tools (left to right): 1) This Person Does Not Exist is a website that curates fake faces generated using StyleGan2 that has been trained on human faces to generate fake human faces using GANs (Karras et al., 2019). 2) Developed by Xinhua and the Chinese search engine, these AI-powered news anchors were developed through machine learning to simulate the voice, facial movements, and gestures of real-life broadcasters (Kuo, 2018). 3) Built by Google Creative Lab, Sketch RNN is an interactive web experiment that lets you draw together with a recurrent neural network model (Ha & Eck, 2017). 4) Built by Yotam Mann and Google, this web tool utilizes generative piano music to let users play a duet with the computer. Users press keys to play a music note, and AI Duet adds some notes to form a duet (Karras et al., 2019)

Fig. 9

Students identified benefits (top) and harms (bottom) of the GAN tools that they explored

Fig. 10

A generative story created by a student using the image and text generation tools

Fig. 11

An example of the “Strike a Pose” worksheet students completed to compare image and pose models

Fig. 12

A student final project that had a sprite tell the user to smile if it detected an eyebrow furrow

Fig. 13

An image of a slide used in an introduction of neural networks for image recognition. The slide shows how the image of a cat is broken down into pixels, then pixels are combined to recognize high-level features of an object, such as whiskers

Fig. 14

Screenshot of a student’s final project that used image recognition to detect different injuries and give a user assistance. The students’ stage, where the application runs, shows a micro:bit robot saying, “I am your Mini Medical Robot!” with two images of fingers next to it. One image is of a finger with a cut, the stage shows a classifier labeling that image as “Cut” with 100% confidence. The other image shows a finger that is healthy; it is classified as “Healthy Skin” with 99% confidence. In the toolbox you can see some of the students’ code which uses ‘Event’ blocks to have the robot give the user instructions on what to do if their skin is cut or not