Jargogled Impressions.

Generalization is the pathway towards achieving mastery. When you are able to apply learnings from one sphere to a different sphere of influence, you ‘generalize’ your learning. A specialist can generalize their learning into a different domain or sphere by taking active interest in dissecting the fundamental properties of his area of expertise. A generalist does not specialize in anything, but is able to draw from a portfolio of skills to apply to the challenge at hand. 

Magnus Carlsen generalized his specialty in chess to the philosophy of life or business. Ray Dalio famously did so with his voluminous tomes on the “principles” of life as derived from his principles of managing an investing business. Elon Musk is generalizing his maverick tendencies to the memetic culture of Twitter. Ye is generalizing his music repertoire to the world of world building and myth making.

Our learning is applied on a variety of vectors – a new stimulus, another environment, a different configuration. Learning how to cover drive in cricket and using it while actually playing is direct application. But cover drives also symbolize proactively taking head-on (on a front foot or back foot) an incoming challenge (the kookaburra ball for instance), identifying it as just in line for an attacking play, and then responding with a swift action (using the tool of choice) to deliver a reverse blow to the attacking object, right? 

Nope.

Generalization can be deceptive as overfitting our learnings can push you straight to authoring those cringe LinkedIn posts that dot the walls of most professionals. Drawing life lessons from an encounter, a chance conversation with a stranger, a tragic incident, someone passing away, etc used to be internalized and reflected on. Today these reflections happen over the medium of a mass propagation tool like social media, getting amplified. 

Underfitting our learnings can mean we are leaving valuable wisdom on the table. Or to put another way, not paying it forward. Most specialists fall into this category, some are able to escape the work trap and bring meta perspectives into their zones of operations. 

I have very limited understanding of machine learning models. But I used the definitions of overfitting and underfitting to overfit my commentary on generalizable learning.

That’s over-reach. Because generalization follows specific knowledge, it does not precede it. If I have not “worked” in AI or machine learning and built or helped build models, should I be venturing boldly to offer nuggets of wisdom using it as a metaphor? Well, yes. But purely as conceptual framing tools – that is, in a narrow enough construct. Using it to dispense life wisdom is, by definition, cringe. Unless you happen to be Walter Isaacson writing a biography of Virat Kohli.

When someone has specific knowledge, it bursts through their interactions. When they learn to articulate it, and generalize it, they become unstoppable forces and carve out a brand for themselves. Specific knowledge is learnt through practice. A mentor can point you to what a specific knowledge may look like, and how to go about designing your life to absorb specific knowledge. But gaining the knowledge is done through the process of doing. LinkedIn is filled with people with varied experiences across the startup, public policy, corporate, government, and media landscape. Their specific knowledge is baked into their implicit understanding of how their business works and not just from a commercial sense but also from a cultural, political, sociological, and ethical sense. 

Much of what you see influencers saying about the fantastic comeback of Virat Kohli in the T20 World Cup cricket is their motivation to use an in-news event to their advantage. Bear in mind, while you can certainly derive learning from your idols, specifically sports idols and their various comebacks, there needs to be a fermentation period for such learnings to be fruitful. Anything sooner is bound to be relegated to the history books as opportunism at best. 

When you are trying to become something (an influencer for instance) versus trying to help someone (through your generalized learning), you tend to “create” learning versus “experience” learning. The latter is more valuable, the former is more cringe.

When newspapers were born, I’d guess their purpose was to offer people a way to make sense of the different bits of news happening across their sphere of influence. As the commerce around newspapers grew, the purpose shifted from experiencing and imbibing these news and activities to bringing everything under the sun as news aka overfitting to the newspaper paradigm. Which ended in the newspapers (and the new media around it) entering the business of manufacturing news. More noise and more irrelevant.

Something similar is happening with the digital space, only in this case, the distribution is free with democratized access. But commerce is still driving these zealous souls, either directly through monetization, or indirectly through reputation. And LinkedIn is suffering as a result.

Not all generalizations are created equal. Learning generalization helped with our evolution as human beings. Fear generalization did so too – learning to fear a bear and generalizing the learning to the environment and your sensory inputs helps when dealing with tigers. But in the modern age, fear generalization can often lead to anxiety if overdone. Generalizing the learning from fear attained through a specific incident may not be translatable into another, so when overfitting to it, we can fall prey to anxiety and a constant fear mongering. 

We must be careful with generalizing our opinions too. As the scale of generalization expands, the level of precision dwindles. So any effort of generalization should consider the tradeoffs involved. 

Any generalization needs to be taken in context and with a recognition of its imprecision. The map is not the territory, and it should not be mistaken to be so. However, we often fall into the traps led by our own mental models which are basically a weaponization of generalization. Generalizations serve a need and we must be mindful that it does not overreach, else we will drown in our own conceptual pits.