The dumbification AI era, or how Lionel Shriver made me think about AI

Today I finished a book that I enjoyed quite a bit, Mania by Lionel Shriver. The book tells the story of a US-based dystopia, where there exists a social campaign seeking to eliminate the prejudice against low intelligence. The book narrates the story of two friends: one who stands her ground, defending what she thinks, while the other goes with the flow, trying to stay afloat and climb the corporate ladder to reach her dream job. An analysis of positive and negative reviews about the book can be found on the Wikipedia page (Please, support Wikipedia if you can 😀). While this might sound too extreme, then I found this blogpost, by Prof. Cristina Lopes. It tells an anecdote she had with a student, where they used Claude Code to develop a full research project, without having any knowledge of what was happening under the hood. The LLM made up the data, the software, and the tests, leaving the student to act as mere observer, with no control of that was happening at all. This is something that is really concerning in today’s education, and also in the corporate world (as I have ranted about in the past).

These two things might not seem connected. What does a dystopia have to do to the overuse of AI in our world? I think that they are more connected than I would like to admit. While in Shriver’s world, mental parity leads to a point where people with no expertise can substitute everyone who is qualified in order to not offend the “alternative intelligence”, in today’s world we are reaching a point where we are substituting our deep understanding with AI. An NPR pool recently showed that 54% of polled teachers said AI makes it harder for students to learn critical thinking skills. There are a lot of AI enthusiasts who argue that AI can substitute our thinking, but not our understanding, that’s the red line that we should never cross. However, if during our most impactful years we do not learn to think critically, how can we expect newer generations to understand what the AI is doing? Humans try to minimize mental complexity, as outlined in this blogpost by Carl Hendrick. Therefore, can we expect students to take the difficult way and not use AI to serve as proxy for their brain to solve the tasks they have at hand?

Everyone, myself included, rely too much nowadays on AI for our daily tasks. However, mine might be one of the last generations that had to go through college without the use of AI. That does not make my generation better, but at least it forced us to learn the basics, even if we copied our work from other students at some point. “But we push students through exams, by doing so they are forced to learn that information, right?” I am not really sure. If I go back to my college years (mentally), all the information that I remember is the stuff that I had to “battle” with, not the things that I learnt by heart to pass the exam. Therefore, utilizing AI, students are missing one of the key parts of learning: failing and having to battle your way out of the problem you have at hand.

Now, going back to the connection with Shriver’s book, I think this might have some fatal consequences. While we are not going to reach a point where inexperienced people is going to rule the world and planes are going to fall out of the sky, we might reach a point where a lot of people might not know exactly what they are doing, and how it works (myself included). Sincerely, this scares me. I am positive and think that there will always be a preference for learning, understanding the concepts, and not just take the easy route. However, I cannot say that our world will not become a 5% version of Mania’s mental parity.