While my work is in library IT, I am also a PhD student in Computer Science, with research specifically in the area of machine learning (ML) and data science. As such, I came to the IDEA institute with a good knowledge and understanding of machine learning algorithms and issues. However, I’ve really enjoyed the wider variety of perspectives presented at the institute. It has been wonderful meeting other people in libraries that are interested in and working with artificial intelligent (AI) and learning about their ideas for projects as well as their concerns and the personal pain points they’ve encountered.
I’ve specifically noticed how enthusiastic this cohort is to learn new things and explore diverse ideas. For a group of librarians, this isn’t surprising. The people here have a wide variety of skills and expertise, and everyone has been eager to learn from others and share their own knowledge. This isn’t always the case. I’ve often encountered people who believe they’re just not capable of understanding AI, so why bother. While it is refreshing to spend a week with a group that is confident in their ability to learn, I do understand the other view. Not a lot of effort is spent to really make AI accessible and understandable to everybody. Too often, AI is presented as a mystical solution that users are supposed to accept as good and correct.
One of the things I have really enjoyed about the IDEA institute this summer is the variety of AI examples and tools that have been presented, many of them that don’t require any computer science or programming skills to use. Not surprisingly, none of these have really come up in my formal education, where the focus has been on programming algorithms and models from scratch. While it’s understandable that computer science education doesn’t shy away from the highly technical, AI is increasingly touching many aspects of everyday life and AI and data science professionals are not the only voice that should be heard. It’s important that anyone who is interested in AI can access information and gain a general understanding of how these technologies work and impact people, without the barrier of too much math or computer code.
I think many of the tools presented during the institute, particularly tools such as the Teachable Machine, provide a lower barrier of entry than the options traditionally taught in computer science courses. With such tools available, it’s possible for most people to understand how machine learning algorithms work, how they use data, and the importance of data quality to the final results. These tools can make it clear how historical and implementation bias can impact the decisions made by AI. With AI being incorporated into more and more daily decision making, it’s increasingly important that everyone understands the advantages and disadvantages of relying on ML systems. If more and more people can become familiar with AI, there is a greater chance that decisions for AI policies and practices will be made by groups of people with a variety of perspectives who are more able to identify a wide variety of issues and solutions.
