Accelerating Ignorance
Do you ever feel like we’re all — individually and collectively — getting more ignorant? I believe we are, and the process is accelerating. Five things are falling apart at once: our information space, our education system, our government, our mentoring system for young employees, and our reliance on AI beyond its capabilities. These five problems reinforce one another, and the problem is greater than any one component would predict.
1. Information
The information available to all of us is declining in quality. There is a growing amount of disinformation — deliberate falsehoods by state actors and partisan groups aimed to sow discontent and polarization. As Kate Starbird and others have documented, many state actors share content on both sides of a controversial issue with the primary goal of creating tension between groups.
We are increasingly surrounded by misinformation. Even someone who is sincerely trying to be a careful consumer of information does not have enough information to make informed decisions on what to believe any more. Where did this story really come from? How well supported is it? Even some of our “gold standard” information sources (for example The New York Times and The Washington Post) are increasingly pushing partisan agendas. Owners of these venues have agendas beyond truthful reporting. More importantly, with financial support for journalism eroding, it’s no wonder that publications make more mistakes. We all know that you can’t rely on that blog post your cousin sent you — but if you can’t necessarily rely on professional journalism either, where do we turn?
2. Education
Our education system is declining due to an unfortunate collision of factors including: post-pandemic blues, the introduction of generative AI, and soaring class sizes for many institutions. Higher education students are attending class less, reading less, and using generative AI to cheat on assignments. No one yet knows how to redesign student assignments to assume the thoughtful use of AI. What we do know is that prompting an AI, skimming the result, and handing it in does not lead to learning.
An additional confounding problem is that at public universities in the United States, class sizes are swelling. If a student is just one of a hundred (in my computing and ethics class, one of over three hundred) people, it’s not surprising that they cut corners on their learning experience and do everything they can to get the desired grade with minimum effort.
3. Government
Checks and balances in our government in the US are no longer fully functional. Further, we seem to tolerate lying from our leadership more than before. Much misinformation is coming directly from elected leaders. Factors that led to the election of those leaders stem from the failures of our information environment and education system discussed above.
4. Mentoring
Next, there is a threat to our system of mentoring young professionals. Some companies have begun to hire fewer workers, asking the remaining workers to accomplish more work with the help of generative AI. But if we don’t have enough young professionals learning on the job, who becomes the senior professionals of the future? What are the skills needed for the future, and how do people acquire them?
5. Reliance On AI Beyond its Capabilities
Our final catalyst of ignorance is AI. Many kinds of AI perform reasonably well, and all forms of AI will keep getting better. However, people will keep trusting it beyond its abilities. And every single one of us is going to regularly be writhing in frustration at AI mistakes. AI is going to introduce ignorance and errors everywhere it is used.
What Can We Do?
My former Dean, Charles Isbell, is fond of saying that sometimes you need to let things fall apart before there is the “collective will” to fix them. We’ve definitely hit the falling apart stage — I hope the “collective will” part follows. In the meantime, we can all do our part. Our part involves demanding more of ourselves, our students, our colleagues, and our information systems — upholding high epistemic standards and countering lies, insisting that students really learn, mentoring those who will follow us, not tolerating unsupported claims for AI capabilities, and refusing to use poor-quality “smart” systems.
