Learning the Language of AI

The MSAI Intensive Program Design course offered incoming students the chance to learn a set of language-independent foundational programming skills.

Artificial intelligence (AI) is transformative interdisciplinary technology, and while there is a belief that AI is a thing of the future, as Northwestern Engineering Adjunct Professor Rachel Trana explained, the future is already here.

"We already rely on AI on a daily basis," Trana said, "from our Amazon recommendation choices to being able to unlock our phones with facial recognition software."

Trana taught Intensive Program Design this past quarter for select incoming students in Northwestern Engineering's Master of Science in Artificial Intelligence (MSAI) program. This introductory bootcamp provided students with an understanding of core computer science concepts — many of which will be relied upon throughout their time in the program.

Trana recently reflected on the course, the future of AI and the biggest misconceptions she hears about the technology.

How do you describe the goal of the course?

The biggest goal of the course is to provide students with a set of language-independent foundational programming skills, technologies and resources, as well as an experiential learning experience so that they are able to apply this knowledge to computational problems. 

What are the biggest challenges you face as a professor when it comes to this type of course?

There are two big challenges for this type of course. The first is creating appropriate content for a student body that has a broad range of programming backgrounds. At one end of the spectrum, there are students with fairly extensive programming experience due to having a Computer Science undergraduate background, and at the other end, there are students who have had some minor experience with programming and data analysis, but in an independent context. The second challenge is identifying the skills and concepts that are most important — and that can be reasonably taught in a 10-week period — for students to succeed in the specific field that they are entering.

How do you feel the course went overall?

Overall, I felt the course went well. This was in large part due to the students' willingness to provide feedback on the course content and structure as we progressed so that I could adapt components of it to address any potential concerns or needs, particularly since this was a fully remote course due to the pandemic.

In what ways do you hope students will be able to incorporate what they learned in your course into the rest of their time in MSAI?

Learning a particular programming language is not the most challenging aspect of entering a computer science field. Instead, it is learning how to work with others, how to read documentation, how to debug and understand code that you have not written yourself, and to sift through large amounts of information to determine what is relevant for the specific problem that is at hand. Through their pair-programming experience in this course, students become  more comfortable evaluating code and documentation, and very importantly, they are able to form a support network as they progress in the MSAI program.

What excites you about the field of AI?

I think the most exciting thing about AI is its potential to be used to transform entire industries, such as healthcare, in a positive way. For example, identifying diseases, such as cancer, with AI detection software at an earlier stage can help healthcare providers recommend better treatments, or can help them understand and suggest drug therapies that are the most effective for patients.   

What do you think is the biggest misconception about AI?

One of the biggest misconceptions about AI is that AI algorithms are objective. These algorithms are only as objective as the data on which they are based, and the people who created them. Issues with bias in AI algorithms have made headlines in recent years, such as Amazon's recruiting tool that displayed bias against women, or facial recognition software that systematically misidentifies and mislabels people of color. It is extremely important to be aware of these limitations, and the area of research that focuses on bias in algorithms needs to make significant gains before we can rely on AI in this capacity.  

What is your reaction when you hear people say they're worried that AI will replace humans in the job market?

To a certain extent, this is a valid concern. AI and the rise of automation will lead to the displacement of certain types of jobs. However, AI will likely create many new jobs and industries that allow for our uniquely human traits, such as creativity, innovation and compassion to be leveraged., a more important aspect of AI is not whether it will replace jobs or create them, but how it can be implemented to help humans in a constructive way so that we can be better and be more secure in our jobs. 

McCormick News Article