In the mid-19th century, naturalist poet Henry David Thoreau thoroughly lamented the expansion of the railroad. Thoreau was concerned not only by how the railway was changing but also the changes to the way we lived. Living in Concord, Massachusetts, Thoreau argued that walking the 20 miles into Boston, rather than rail travel, was not only more efficient, but also more enjoyable. During Thoreau’s lifetime the average cost of a rail ticket was nearly equivalent to a full day of labor for the average worker. By extension, Thoreau argued that time was better spent walking: which took less than a day and allowed an individual to enjoy existing in nature. Thoreau recognized that the railroad could be a valuable tool, but he hoped that people might come to recognize the value in the journey rather than focusing solely on the outcome.
Like the average citizen in the 19th century, we too are faced with a rapidly changing landscape. Discourse on artificial intelligence (AI) has become an increasingly popular topic across major news outlets and college campuses over the past few years. The rate of innovation in the field means there is always a new development upon which we can reform and extend our current understandings of how AI will shape our world going forward. There are an abundance of issues in regards to AI, ranging from job loss to the impact of its incredibly high energy use on the environment. However, I want to examine one of the most widely purported benefits of artificial intelligence: time.
It has become increasingly common for business executives to rely on artificial intelligence to organize and optimize their daily work load: by allowing AI to handle their more “menial” and time consuming tasks, this allows them to dedicate more of their energy to higher-level projects. A similar strategy is often preached to students to optimize their learning: such as using AI to analyze practice problems to highlight areas for further study or generating lists of relevant sources at the beginning of a research project. The basic proposition of AI is that it allows us to skip over tasks of “low” value to focus our attention on “high” value tasks.
But what makes a task “low” value? These tasks are often necessary, but time-consuming. In other words, they are tasks that, in theory, are better to automate. This argument rests on the assumption that these tasks are only valuable for individuals based on the outcomes they produce rather than the processes they require. However, I find this to only rarely be true, particularly for students. Consider the example of allowing AI to analyze your practice problems or quizzes to suggest areas for improvement. I would argue that grading and evaluating your own work teaches you more than AI’s automation of the process by allowing you to discover the boundaries of your knowledge yourself.
Studying in this way also pushes you towards proven effective studying strategies such as those relying on active recall. Although AI might be helpful for some organizational study methods, such as planning spaced repetition, for methods that rely on students’ critical thinking, AI seems to fall short. If our goal as students is to learn and become more intelligent and thoughtful people, why should we let AI do the learning for us?
This logic can apply to the professional world as well. Although the goals of the corporate world are fundamentally different from those in education, over reliance on AI can replace critical analytical skills. This is particularly important when it comes to recognizing the current limitations of AI. If executives rely too heavily on artificial intelligence, they might lose the ability to recognize when and where it falls short.
Additionally, the use of AI as a time-saving tool continually relies on “best intentions.” Yet many people acknowledge that they use AI as a shortcut to doing work faster, rather than as a tool to make their work better. If we could fully trust everyone to use AI solely for good, we would not continually find ourselves amongst discourse regarding the boundaries of professional and academic dishonesty related to AI.
Much like Thoreau and the railroad, I have ambivalent thoughts on the use of AI. I think that artificial intelligence has the potential to do so much good. Its implications for scientific research are nearly infinite. At its best, AI will allow us to extend the boundary of human achievement. However, I worry that artificial intelligence will draw out the worst in us rather than the best.
I worry about the loss of critical analytical and technical skills developed overtime through the repetition of “menial” tasks we are quick to bypass through use of artificial intelligence. AI is changing the course of human history; however, we should be cautious that we don’t lose the skills to live and work meaningfully without it.
