Understanding our Ubuntu: Mind your mentalising
This appeared as the introduction to the Deep Learning Indaba’s newsletter Satu Yetu Our Voice, in November 2024 and appears as a blog post on their website.
Ubuntu is ubiquitous for me as an African, and an avid Linux user. Broadly a philosophy of humanity and connectedness [1], I often see the term used as a self-evident truth or mantra of appreciation for a community. Yet, for this cherished and connected Indaba community, I want to reach beyond appreciation and attempt a touch of understanding our Ubuntu.
Although now a NeuroAI researcher, I started once upon a time as a psychology and biochemistry major. Two important lessons from these pursuits have stuck with me more than any other over a decade later. First, in psychology, the concepts of mentalising and the Theory of Mind. These metacognitive processes refer to our ability to attribute mental states – such as beliefs, intentions, desires, and emotions – to ourselves and others. This ability is fundamental for social interaction, empathy, and communication. To be clear, the concept itself was not what made the lesson impactful, but rather the shock of my classmates at how important it is to be introspective and try to be aware of others’ thoughts and feelings. Over the years, however, I have realised that mentalising is an active, intentional process that we all do but to highly variable extents. Furthermore, the number of times we “jump” minds is also highly variable: e.g. do we adapt our approach based on potential perceptions, and how will this change affect actual perceptions? Another way of putting this is the process of aligning what we want others to understand from what we say by reducing ambiguity and maximising mutual expectations.
Thus, beyond being passively aware of our connectedness - that we are dynamically interacting - I beseech us to actively pursue our own conscious thoughts and motivations, and those of others. Do this to seek understanding despite knowing many such things are unknowable even about ourselves. This neatly brings us to my second important lesson.
Biochemistry is the study of complex biological processes interacting in a fundamentally probabilistic manner. In lessons that spanned the gamut of protein folding, DNA interactions, Kreb’s cycle, and ion channel gating at the atomic level, the most “adhesive” lesson among these well-laid-out and complicated processes was the lesson on “it is more complicated than that™.” No matter what we learn or intuitively know (or “grok”) to an absurd level of detail, it is still more complicated than that. Although this lesson was taught in the context of genetics, molecules, and proteins, the natural extension is to life itself, and the emergent societies thereof. Egocentrically, despite being the person who knows the most about myself, arguably just in front of my wife or mother, I am still more complicated than I think I am. I act unexpectedly, explore new limits, and ultimately grow what I know. If I do not fully know my complicated, dynamic self, then I cannot pretend to understand others’ thoughts and feelings fully. The best I can do is build rough approximations (or “models”), which have many assumptions that will ultimately, definitely be wrong. Despite all models being wrong, many are useful [2].
We build many models. We make many connections. In deep learning as in life, some connections are fleeting and yet others are so strongly weighted that they inform how we process the world. For me, I have been privileged to form strong connections in particular through organising the Deep Learning Indaba𝕏 South Africa and the Simons Computational Neuroscience Imbizo. Along with building artificial models, these communities have been mutually built by inferring mental models of individuals and to a large extent, the broader community itself. With such a diverse audience, understanding the expectations of attendees is crucial to executing a successful conference (Indaba𝕏) and summer school (Imbizo). Preempting these expectations by mentalising with future participants, and being informed by past participants, is a crucial aspect of our teams. In some ways, we, as the organisers, can explicitly set expectations. Yet in many ways, there will always be a realm of uncertainty and novelty about the mental models we have built and the expectations we have set. When this inevitably occurs, then acceptance and adaptation update our understanding. In leading these teams, I have gained a better understanding of my own thoughts, beliefs, and feelings. In doing so, it has allowed me to improve my mentalising and contributions to the spirit of Ubuntu that these communities hold. It is with deliberate, intentional, and reflective care that I cultivated these communities to be more than any individual, including myself. Thus, as I move away from a steering hand and instead to a guiding voice for these initiatives, I have immense pride in the connections we have built, and imbue the next generation of leaders with the tools with which to not just fleetingly connect but build strong weights through mentalising.
It is my strong belief, therefore, that through mentalising we can build moldable models of minds [3] (our own, others’, artificial, and metaphysical alike), to get a step closer to a deeper understanding of our ubuntu.
- Christopher Brian Currin, Board of Directors on the Deep Learning Indaba𝕏 South Africa and the Computational Neuroscience Imbizo, steering committee for Sisonke Biotik and advisor for the African Computer Vision Summer School. November 2024.
[1] The African Journal of Social Work defined Ubuntu as: A collection of values and practices that people of Africa or of African origin view as making people authentic human beings. While the nuances of these values and practices vary across different ethnic groups, they all point to one thing – an authentic individual human being is part of a larger and more significant relational, communal, societal, environmental and spiritual world.
[2] More about “How to Make Decisions in an Imperfect World” and “Doing the Best We Can With What We Have” at https://jamesclear.com/all-models-are-wrong
[3] A shameless plug to my dear friend’s book “Models of the Mind: How Physics, Engineering and Mathematics Have Shaped Our Understanding of the Brain”
https://www.goodreads.com/en/book/show/50884536-models-of-the-mind