Main menu


What is artificial intelligence? “The process of developing AI is very important about how we humans define it,” says Assistant Professor Shiyan Jiang.

featured image

Technologies like Artificial Intelligence (AI) may seem strange to many, but Shiyan Jiang, an assistant professor of design and technology learning at North Carolina State University School of Education, teaches students to learn about technology. I believe it will help you without having to acquire important skills. Pursue the STEM profession.

AI is often seen as a sophisticated, high-level technology. However, Jiang, who is conducting research focused on expanding access to STEM in areas such as artificial intelligence, data science, and computing, says STEM is very human-oriented.

For example, in a subfield of artificial intelligence called machine learning, computers learn patterns from training data collected and generated by humans and use these patterns to make predictions based on the knowledge gained from the dataset. increase.

“Many of us may think that artificial intelligence is a mystery that can cause magic, but it’s not. The process of developing AI is how we humans define AI. That’s it, “she said. “I want students playing with these technologies from the beginning to see the decisions that humans need to make, not as magic or mystery, but as a very human-oriented experience. I want to see even the potential bias we have put on those AI technologies. ”

Bring diversity to the field

According to Jiang, the bias in AI technology may not be intentional, but it can be due to the lack of diversity in the field and can adversely affect the people who use the technology.

For example, studies have shown that facial recognition AI technology on most smartphones works more often and better in white men than in people with other racial backgrounds. This is the result of the training data used to develop the AI ​​model, which contains an imbalanced number of images of white men.

“Initially, we are not always aware of these issues when developing technologies that make our lives better, but in the end, some people can only use the features, so how many? I found out that there was a problem. Everyone, “said Jiang.

Having different opinions and perspectives can challenge the assumptions that may have been made in developing AI technology, so students with historically undervalued backgrounds in AI Encouraging them to consider their career potential is very important. She said people with different life experiences could find problems with the dataset used to build the AI ​​model, or with variables that someone from a different background might not have considered. ..

To bring the diversity needed to this area, Jiang believes it is important to give high school students of all backgrounds the opportunity to take up the AI ​​profession.

“As educators, we want more students to explore their potential and stay out of reach. Finally, to AI engineers and scientists. If not, they can see “I have”. Here’s an option, and this is what I can do, “she said. “The goal is not that everyone wants to be ready to become an AI engineer or AI scientist, but what kind of career they want to get after having them see their potential and get inspiration. You can decide whether to go inside. ”

Merge AI and ELA

Because even students who are not interested in pursuing a career related to artificial intelligence can learn skills related to the efficient use of technology and critical thinking about whether to use advanced technology. , Said Mr. Jiang, which is beneficial.

“They need to have a critical perspective on how to use technology in their areas of interest,” she said.

Jiang is helping high school students expose high school students to artificial intelligence with the “Narrative Modeling with StoryQ: Integrating Mathematics, Language Arts, and Computing to Create Pathways to Artificial Intelligence Careers” project launched in 2020.

The StoryQ project allows students to participate in the process of developing a test classification model and ask how decisions made along the way affect the appearance of the model. You also need to think about what your model has goals and who will be affected by the AI ​​model.

“We want to open the AI ​​black box so that we can experience AI as a decision maker and see decisions that affect how the model looks,” Jiang said. Says.

StoryQ is unique in that it introduces AI through integration of various disciplines and is intended for use in English art classes. English language classes are mandatory for all students, so using AI in this area means that all students have the opportunity to learn about technology. This often does not apply to optional computer science classes.

Students can write stories using their stories, perspectives, and experiences as a dataset of AI models. Because they use their own words as datasets, discover and discuss that they are making mistakes by misunderstanding what the AI ​​model intended, and understanding how to make the model work better. I can.

Languages ​​can vary by region, culture, and community, so different perspectives can also be used to increase the diversity of AI models.

“StoryQ naturally attracts students and helps them access AI. [through] Text and language, we all have something to say. Often barriers aren’t like math or computer science, “says Jiang. “They built a text classification model and saw how the language was used in different contexts, while incorporating a unique perspective. Text is this kind of with AI. A wealth of space for creating conversations. “

In addition to exposing students to the concepts of AI and data science in a new context, according to Jiang, using StoryQ in ELA classrooms will allow students to invest more in English lessons, making ELA teachers a role model. It was reported to see.

“Learning isn’t just about content. It’s also about building relationships, and ELA teachers who love writing, literacy, and poetry can teach AI in an effective way and want to learn more about this topic. I found that it would be like this, “she said.