Learning to Ride Your Brain

71

1

13

Share

Steve Jobs was wrong. A traditional computer was never a bicycle for the brain. It was never a coherent enough machine to require, or enable, the holistic mode of dynamically stabilized control that is “riding a bike” after learning to “balance” it. At best, a traditional computer is like a set of power tools. If it’s an Apple, it’s a coherently branded set of power tools in a nice toolbox, but not an integrated machine. At least not at the level of human user experience. You do not use a traditional computer the way you ride a bicycle.

AI has revealed this by revealing itself to be something that does require bicycle-riding type skills: a motorcycle for the brain. And just as learning to ride a motorcycle is about extending a previously acquired ability to ride a bicycle, learning to ride AI is about extending a previously acquired ability to ride your own brain.

The experience of using a regular computer is a largely unrelated set of skills. A giant red herring.

AI literacy is not an extension of computer literacy.

The thing is, the human brain itself is already a combination rider-and-bicycle type entity; one which not all humans know how to ride.

The “bicycle” part is loosely what we call intelligence, and is largely a cultivable set of native aptitudes, while the “rider” part is the part that has learned to manage attention. And that process is strikingly similar to learning to ride a bike. And as with riding a bike, not all humans learn it.

To understand how this works, we have to dig into the sociology of how “smart” is constructed, recognized, and socialized in human society. Because what we conventionally recognize as intelligence, which I’ll call smarts, is not quite the same thing as being able to ride your brain. This means it is dangerous to mistake “smart” as “competent to ride AIs.”

To appreciate this, you first have to appreciate what smarts are actually for socially, and why smarts can function without necessarily being accompanied by the trait required to ride AI well — cognitive balance. In fact that can be a liability if your societal role is to perform smarts.

Unless you’re active in one of the relatively few domains where savant performance is meaningful and legible, like chess, music, or number theory, “smart” tends to be highly socially constructed as some mix of academic performance, extracurricular depth, and social presentation.

Good grades, evidence of depth outside textbooks, and most importantly, a capacity for lively social interaction that subtly dominates conversation by topping others’ arguments, citing deeper knowledge sources, and referencing the right hooks into relevant authoritative discourses. That’s “smart” in conventional terms.

Generally, acquiring a “license to smart” amounts to demonstrating early and often to peers that almost any subject is deeper and more complex than the casual dilettante appreciates.

The “smart” person has a particular social role: Serving as ready proof of the shallowness of the mental life of the median person on the median topic, who is revealed to be, not stupid per se, but simply atrociously unaware of the meta of most topics they weigh in on, to the point that they cannot be taken seriously. This is primarily a social governance function, not a directly useful thinking function. The smart person is a useful instrument for the exercise of power. Sometimes the power is even their own, but more often it is the power of particular political, cultural, or social institutions.

Curiously, logic and reason play almost no role in this sort of performance of smartness, since smartness typically operates by subverting the interlocutor’s lazy presumption that they are competent to reason about something in the first place. The characteristic move of the smart person is to reveal to you that you haven’t yet earned the right to begin reasoning because you don’t understand the terrain. And to do so in a way that you cannot easily resist.

Smartness operates by vividly demonstrating the conversational incompetence of the interlocutor, with varying degrees of grace and kindness, to the interlocutor’s own satisfaction, even if they are too chagrined to admit it. This means cowing them into suitable deference, an awkward face-saving retreat, or visible absurdity. The effect is demonstrating to others that they are either cowards, frauds, or clueless. In all cases, they are revealed to be governable without force, which is why smarts are useful for power. This also explains why getting physically assaulted is the main risk of performing smartness, which is also a useful outcome to power, because it legitimates the use of violence to govern those who resist browbeating by smarts.

This model of smarts is what’s performed by Matt Damon in the Good Will Hunting bar scene (not the less interesting math-savant scenes). He doesn’t win by doing a harder calculation. He wins by revealing that his opponent’s apparent intellectual authority is borrowed, shallow, and unearned. And he does so within the opponent’s own preferred terms of reference, leaving him unable to defend himself on home turf. And he does so as a social outsider, adding insult to injury.

Real-life examples are less dramatically intense, but this is how people are “normally” judged to be smart. That’s why interviews are such a ubiquitous gatekeeping mechanism. They are stylized versions of a normal social process of establishing who deserves to be taken seriously.

There are several pathological smartness cultures that derive from this basic culture through degeneracy, but fail to function in quite the same way.

Virtuoso IQ performances on inadequately prepared or persuasive ontological ground. These separate the audience into uncritical future acolytes and skeptics who tag you as a solipsistic egghead with an unconvincing halo; someone who doesn’t realize they’re prematurely trying to “reason” without establishing persuasive common grounds and motives for doing so. You cannot persuade a non-English speaker of their own stupidity in English. You have to do so in their native language. Cults start this way.

Overdeveloped debate skill or wit that dazzles without persuading. Any wins fail to sustainably occupy nominally captured ground because later reflection by witnesses or opponents/victims typically reverses the gains, unless they’re locked in immediately by other means. Political and legal theaters operate this way precisely because such locking-in mechanisms exist.

Demonstrations of deep erudition that turn conversations into tedious lectures, presuming the authority of a tradition rather than establishing it. Ivory towers start this way, as do taste subcultures, fandoms, and various sorts of glass-bead games. A fan of rock music cannot establish that someone with no interest in rock music is uncool. A historian cannot demonstrate that someone with no interest in history is stupid.

All three modes create escaped realities that are only persuasive if you’re already inside them. They create diverging dissensus realities. By contrast, “normal” smartness performances are generally persuasive across such social boundaries. They create converging consensus realities.

Which mode dominates determines whether zeitgeist narratives are, overall, convergent or divergent. Whether we inhabit a consensus reality or a web of contending alt-realities.

But none of this is actually that interesting. It just demonstrates that functional and socially useful smarts, both normal and degenerate, do not obviously need cognitive balance in order to serve as governance mechanisms.

Performative smartness behaviors, both normal and degenerate, constitute a kind of consensus social-proof process layer of society. A form of sociological intelligence that underwrites societal governance, but has a very vague relationship with truth at best. Important as praxis, but peripheral to the pursuit of truth in any significant sense.

Performative smartness is what we are socialized into. Like getting a driving license for a four-wheeler, a vehicle that does not require a learned sense of balance to operate, at least at a basic level.

I recall when I was first tagged “smart” beyond my immediate family, sometime in fifth grade, with firm membership achieved by eighth grade. There was real signal there, but not in the visible signals.

The real signal was that between fifth and eighth grades, I learned to actually manage my attention and direct it where I wanted. Something that many of my peers, who were also tagged smart, did not learn.

“This is thinkable, and I can think about it by just paying steady attention, which I can do,” became a basic move for me. A move that moved the challenge frontier from how to think to what to think about. This does not mean I got equally good at thinking about everything. It means I got good enough at thinking about enough things, that the bottleneck moved to choice of attention targets.

I started coming across as “generally smart” to others as a side effect, and started to easily get good grades with barely any effort, and banter impressively at parties when in the mood. Control over attention was much more consequential than raw thinking ability. I was officially licensed to smart by age 14, with all the attendant privileges and vulnerabilities.

The underlying developmental event was not really the acquisition of knowledge, or even an increase in intelligence in any simple sense. It was the acquisition of a new kind of control over an intelligence that was already there.

This felt much more like learning to ride a bike, which revealingly does not have or require a licensing regime, unlike car driving.

When learning to ride a bike, your balancing circuits visibly cross a fairly sharp threshold, with just a narrow transient phase of wobbling. You’re suddenly able to operate as a dynamically stabilized human-plus-bicycle system.

Before that threshold, every little correction is a deliberate and effortful struggle. You wobble, overcorrect, lose the thread, and fall off. Afterward, the corrections happen largely beneath conscious awareness. You can steer, accelerate, slow down, and pay attention to where you’re going rather than merely trying to remain upright.

This is the key: learning cognitive balance does not mean learning a large collection of explicit rules for staying upright. It does not mean acquiring a lot of theoretical knowledge or even any theoretical knowledge. After all, monkeys can learn to ride bikes. It means acquiring an integrated regulatory capacity that makes an entire new class of behaviors possible.

You may still need to learn some safe riding behaviors for traffic, but nobody demands to see your biking license once you demonstrate you can ride. By contrast, it’s possible to fake being able to drive a car safely well enough to be dangerous, which is why licenses are needed.

In a similar sense, when learning to ride your brain, you’re suddenly able to operate as a dynamically stabilized generative-and-reasoning-brain-plus-attention-regulation system. It’s just not as visible. So we end up noticing and certifying varied epiphenomena that serve as unreliable proxies for being able to ride your brain. That’s smartness licensing. A fragile certification that you can avoid falling off in some situations.

The bike-riding quality of attention management becomes clearer in adulthood, when you start noticing people you thought were “stupid” suddenly turning “on” (usually because they landed in more suitable environments than school) and becoming seriously impressive in some way that does not present as conventional smarts.

They’ve learned to balance on their brain bikes. Their attention is controlled, directable, coherent, and steady. At least some of the time.

Sometimes this presents as “intelligence,” and other times it presents as other traits: drive, imagination, betting acumen, foresight, resourcefu