top of page

Disrupting Our Algorithms



In this era of isolation and loneliness, I count myself among the fortunate who have a number of friend groups. One of these groups consists of about a dozen guys I have known for more than forty years - some of them longer than 50. A casual observer might question whether we are actually doing “friendship” right. We pile on, make light of each other's misfortunes, retell every embarrassing story, and more. Even when we are talking about something of substance, the conversation often devolves into something you would expect out of high school sophomores. Still, the time we spend is cherished. When we are together, we are the funniest people on the planet, we relive the most colorful adventures (over and over), and we sing way too loudly way too often. Fortunately, our spouses humor us, and the reason I think that they do is that they can see the deep affection we have for each other - that and the fact that they know we are better, more complete people through our association.


I belong to this tribe. I also belong to the tribe of my family with all of its ancestral roots. Most of us belong to at least one tribe, and constant across all of them are their predictability and idiosyncrasies. We meet at certain places. We celebrate certain ways. We expect certain responses in certain situations. And even when we complain about the crazy, irrational sides of our tribes, we recognize that these represent OUR brand of crazy. In short, each tribe has a set of norms and values that play a role in constructing the customized lenses through which we see and interact with the world around us. Tribes produce algorithms.


Algorithms have purpose. We do certain things in certain ways in order to attain certain, relatively consistent outcomes. In healthy relationships, this enables us to establish trust and manage expectations. This, in turn, creates a degree of psychological safety which provides fertile ground for learning. Whatever wisdom we come to acquire over our lifetimes is colored by our most influential relationships and, by extension, the algorithms employed.


Similarly, algorithms are the underpinning of technologies. The one big difference is that tech does algorithms much better than we humans. Take, for instance, generative AI such as ChatGPT. Where in tribal settings trust can be tested when someone goes off script, generative AI is programmed to return the most predictable response to any question in any situation. While there may be flaws and biases in the algorithms, the AI itself acts exactly as it was designed every time. In this way, technology is cleaner, simpler, and less likely to produce disagreeable results (assuming the right inputs).


Just as the algorithms of our tribes create the conditions for learning, so too do the algorithms of educational technologies and generative AI. (See these previous posts on ed tech and AI here and here.) Your preference may be for tribal algorithms over technology algorithms or vise-versa, but at their core they are very similar: both sources act like funnels, channeling life’s complexities and uncertainties into nuggets of insight and action. This is invaluable to the learning process…until it isn’t. So much depends on the algorithms themselves and, even more importantly, the inputs we provide.


Looking at the algorithms of my above-mentioned friend group, they are influenced by the fact that we are a bunch of middle-aged, midwestern, white males from middle-class Catholic families. This is neither a good thing nor a bad thing. It’s reality. While the group’s algorithms are certainly influential in my life, they are not well suited to produce value in many situations. The temptation, however, is to ask for advice on such matters simply because they are trusted friends. In such cases, you are placing your inputs/questions in the wrong funnel. You likely will get a tidy, predictable response that resonates, but it will do little to deepen your level of understanding of the original issue.


Similarly, regarding artificial intelligence, we all remember the 2016 presidential election when every major media outlet thought they had cracked the algorithmic code on voter predictive modeling. They hadn’t. In that instance, the algorithms and inputs failed to account for one immutable principle, the capriciousness of life. It was like playing poker with a joker card included in the deck. With every new round of exit poll reporting, more data was poured into the funnel slowly revealing the winning hand - that is until someone else played the joker. The humble lesson learned by media outlets is that the “exit poll” funnel is not a direct proxy for the “votes cast” funnel. This isn’t meant to suggest that the exit poll funnel doesn’t provide value. One just needs to be cognizant of what’s really being measured.


We’ve reached an inflection point in the information age where networks, data, and algorithms combine to provide more access to more information more rapidly than ever before. The pace of change from here forward is likely to be exponential. Does this mean that humanity is getting smarter? No. When speaking on this subject recently, Brother Guy Consolmagno, S.J. rightly pointed out that, “Data is not information. Information is not knowledge. Knowledge is not understanding. Understanding is not wisdom.”


In spite of our technological advances, learning remains a highly personal, highly imperfect human pursuit based on experiences, tribes, tendencies, motivations, and more. A person’s access to quality learning resources is also an important factor. Where in the past the access barriers were physical in nature, today they are primarily digital. Because of the ubiquity of data and information, many view the need for traditional knowledge to be less critical. In some ways, that might be true, but what seems to have emerged as a result is a heightened need to understand because much of the world’s information conflicts.


This shift from a knowledge age to an understanding age seems to have caught us off guard. What I sense is that there is a spark of a metacognitive desire to understand within most people. The struggle is that many of the algorithms of our go-to funnels either haven’t been updated, or they are programmed to confirm, affirm, and validate. Thus, our attempts at understanding are incomplete though we don’t recognize them as such. Predictably, we are left wondering why someone else’s understanding could be so different from our own.



In order to better prepare young people for this new reality, many are calling for a revolution within our formal education system. I would suggest a more measured approach. At its core, learning has always been about questions and sources of information, and I believe that remains the case today. The shift comes in the types of questions being asked, the persons posing them, and the exploration of sources that not only yield knowledge and understanding but unearth more questions. This constructivist approach is not new. Its roots include the socratic method of ancient Greece as well as 450 years of Ignatian pedagogy. The change comes in the ownership of the process.


In many ways our current classrooms are about orchestration and navigation. Teachers expertly map out a plan for reaching a defined destination, then students are loaded on the bus, the GPS is set to “fastest route”, and they’re off! Occasionally, the teacher pulls the bus over at predefined stops to allow students to explore, then students are herded back on the bus to proceed. This worked well when the students placed greater value on information and knowledge. Now, armed with their cellphones, Google, and AI, it seems that students are less willing to board or re-board the bus. Why? Because this model is more akin to something that happens to our students rather than through our students.


In this model, the classroom is owned by the teacher with relevance and student agency being defined and prescribed through the lens of the teacher. What might a classroom look like under new management where ownership resembled a 60-40 partnership between teacher and students? The teacher would largely retain control of overarching objectives/destinations and pacing, and then accompany students, collectively and individually, in defining questions and exploring sources of information.


In this new model, the keys are turned over to the students who make their way to the predefined stops along routes they find most suitable. Through their accompaniment, teachers continue to point the way through expert demonstration, prompting, and course correction when necessary. They also call attention to the processes of discovery and skill activation in which the students are engaged so students can recognize the origins of their efforts and build upon them.



Tribes. AI. Algorithms. The age of understanding. Questions and sources. Relevance, agency, and accompaniment. There is still so much more here to unpack and I’ve yet to touch on the thread that ties it all together - a process of discernment that connects body, mind, and spirit. I will humbly attempt to address all in the weeks and months ahead. Know that as I do this, I am going through my own process of discovery and would value greatly any guidance you have to provide.


Blessings this Advent Season.





 

CONTRIBUTOR: Jeff Hausman, AVLI President


vol 6 issue 4

bottom of page