Abstraction Diseases

IMPORTANT NOTE:

Please take 2 minutes to fill out this form, it would really help me out.

I'd like to understand what kinds of content you're looking for and make the blog even better.

Thank you so much!

-d

I have been lying to you all along.

See, I'm very goal-oriented.

I've always set goals.

On this blog and others, I have reviewed my goals diligently year after year, breaking down where I succeeded and where I fell short. Setting goals in the new year is something I actively look forward to. My first ever blog was called "New Year's Resolution."

And I am just now slowly realizing that I had it all wrong - terribly, terribly wrong.

Chances are, if you're reading this, you have been terribly wrong as well.

In fact, you have not only been holding yourself back - underachieving relative to what is possible - you have been inexorably moving away from the things that will make you happy, secure, and successful.

I have had a lie eating at the very heart of my life.

And my bet is that you do, too.

It has taken me a long time to get here. Fair warning: I am still unsure. Part of me doesn't want to write this essay, for fear that if I say this out loud, my entire life will have to change, forever.

But I've been taking steps for months now and have already seen positive signs - less anxiety, more gratitude and better financial, physical, and emotional results.

So it's time. Let me lead you through the argument that's been banging around in my head.

You can tell me if I'm crazy when we're done.

We start, as with so many things do, with Freud.

Central to Freud's understanding of human life was the drive.

Drives exist at the level of the organism - they are evolutionarily selected for, deeply programmed forces. We have a drive towards reproduction, for example, and a drive eat and drink.

A drive is never satisfied; it is a kind of tension that yearns for a temporary release, only to build again. The drives propel us into motion, constantly seeking new ways of relieving the tensions within us.

What is critical about understanding the drive is that we do not pursue the drive itself. We do not walk around thinking "I want reproductive success!" Instead, we seek an object - a sexual partner, money, a powerful position in the social hierarchy - which we believe will lead to a reduction of the tension caused by the drive.

Hence, we are only ever seeking proxies for what we really want. We have no direct access to happiness, survival or reproductive success, only to the things that we think will get us there.

When the drive rises to the level of consciousness and is aimed at a specific object, we call it desire. Desire, and the process by which it finds its object, is problematic for a number of reasons, but for now let us leave it at: the energy of the drive focused on a particular object is what we call "desire."

Biological beings exist to get what we desire. That is what we do: our entire cognitive apparatus is there specifically for this purpose. How does this occur?

Human beings are expert predictors. That is what our brains specialize in, and it is what has allowed us to become the dominant species on earth (with the possible exception of ants!). Our single evolutionary advantage has been to predict more effectively - and at greater levels of abstraction - than any other living thing. This predictive capacity is so important that we are born entirely helpless, simply to allow our gigantic noggins to safely pass through our mother's birth canal. Evolution has selected for a being that cannot even lift its own head for months after entering the world, simply to enable our advanced predictive processing.

Our brains are prediction machines. They do this by constantly building and updating causal models of how the world works. We take action, perceive its result, and add that as one more piece of evidence to our mental model of what to expect moving forward.

This is happening at all times along multiple levels of abstraction.

My brain has a causal model and a set of predictions about physical space, for example: how I expect things to react if I touch them, how far away they are likely to be given how they look, etc. This is considered a low-abstraction model.

When these low-abstraction models prove to be wrong - when we experience what is called a prediction error - it can be quite disorienting. A good way of experiencing this for yourself is through the Hollow Mask illusion, a well-known optical illusion in which the concave side of a mask is perceived to be convex.

You know for a fact that the inside of the mask is concave, that its contours are pointed away from you. And yet, you cannot help but perceive them as pointed towards you. Why? Because the predictive mechanism of your brain has lots of evidence that faces are never inside-out. This mental model of the world has so much evidence that your brain concludes your sensory inputs are inaccurate and "corrects" them for you in real time.

Our predictive mental models are not restricted to this level of abstraction. Groupings of predictions are bundled together, creating increasingly more abstract models. We have predictive models for how our partner will respond to a specific kind of comment; we have predictive models for whether the "cuisine" of different "countries" will be "delicious."

These levels of abstraction continue all the way until we get quotes like the one from Jacques Lacan below:

"What is realized in my history is not the past definitive of what it was, since it is no more, or even the present perfect of what has been in what I am, but the future anterior of what I shall have been for what I am in the process of becoming."

...Which is comprehensible to some people, based on their causal models. Not me, but I take their word for it.

How does this predictive nature of our brain guide our behavior? By physically editing the structure of the brain as it goes.

When the predictive brain senses an adaptive pattern emerging (a behavior that is regularly followed by a reward, for example) the orbitofrontal cortex releases a spike of dopamine. That dopamine is a signal to the neurons that just fired to do so again, a kind of "more like this" button. Thus, desired behavior is reinforced, in the same way that giving a dog a treat after it sits reinforces that behavior.

Do this long enough and the structure of our brain changes. For example, neurons that often fire in concert literally strengthen their connections, making it easier for them to activate together in the future. Physical changes occur at the synapse, such as growth of new dendritic spines or increased synaptic efficiency. In short, they become associated with one another. This is known as Hebbian Learning.

To review our argument so far:

• Drives are evolutionarily-selected, genetically hard-wired end goals we have no control over (reproductive success, survival, etc.)
• Drives are inextinguishable; they are cycles of the buildup and release of tension. As such, we do not seek the end of the drive itself, but rather an object that is predicted to temporarily satisfy the drive.
• Our brain is a predictive machine; it is hard at work building a causal model for how the world works. It does this so it can better predict and pursue the objects of our desire.
• As the brain detects adaptive patterns - things which lead to rewards (the objects of our desire) - it edits itself to reinforce those patterns, in the short term with the release of dopamine, and over repeated reinforcements, through Hebbian learning.

Note that none of these processes are happening consciously. All of this architecture exists in monkeys just as it exists in us. This is already-always happening, a pre-requisite to our existence, something we can never opt-out of. It is inevitable.

And this is good! It is adaptive in the environment in which this architecture developed.

But I would argue we are no longer in that environment.

Instead, we are in the Anthropocene, an age defined by massive amounts of environmental manipulation and flux. Our day-to-day lives do not reflect the day-to-day lives of our ancestors.

The primary way in which our lives differ is in the amount of abstraction we must respond to and predict.

Take what I’m doing right now. For one, I’m using language - itself extremely new on the evolutionary timeline. The brain does not have a dedicated, newly evolved region solely for language. Instead, language processing relies on a network of pre-existing brain areas, like Broca’s and Wernicke’s areas, which have been adapted and repurposed for different aspects of language comprehension and production. I’m using old equipment for an entirely new and unexpected purpose.

But what's more, I am not speaking my language - I'm writing it down, an even newer development. Not to mention I'm writing it on a computer, hooked up to the internet, for an audience of thousands of people I have never met. The size and the anonymity of that group (you, dear readers!) would be completely impossible for our ancestors to grasp.

But why am I doing all this in the first place? To learn? To build my personal brand? To engage in the world economy? Literally nothing I am doing has any corollary to the world in which my cognitive architecture developed.

My brain is chugging along with the end goal of "make babies and don't die." It is dedicated to that purpose. Meanwhile, I'm "blogging" about "neuroscience" to "get subscribers" that will "pay me for coaching someday" and because I want to be "fulfilled" and enjoy "personal knowledge management."

It's like taking a Buick to the bottom of the Mariana Trench - and, in spite of everything, succeeding!

While an incredible testament to the power possessed by the three-pound piece of meat inside our skulls, these rapid changes in our environment and unprecedented levels of abstraction can lead to problems.

It's one thing to ask the brain to predict things like how to move through physical space, or what other people are feeling.

It's quite another to ask our brains to predict the behavior of massive groups of people we've never met. How does one begin to model and predict the behavior of people in a city, for example? What about large groups of sports fans? Entire nations?

It doesn't stop with others. We demand our brains help us interpret the brain's own biochemistry ("should I go on antidepressants?"), linguistic abstraction ("Are 'autistic' people more likely to have 'ADHD'?"), vast webs of causation ("Will oil futures go up or down in the wake of Israel's air strikes in Iran?"), and technologies which completely warp our perception of what is possible (VR, AI, and so on).

Our world is unrecognizable to the machine we ask to predict it.

And for the most part we do all right.

Until we don't.

With that groundwork laid, next week we will turn our attention to the ways in which our predictive brain falls short in our world of ever-increasing abstraction, and the ways in which goals exacerbate and accelerate this trend.

Yours,

Dan

Something I'm Reading:

Joint Over- and Underdiagnosis

Today I had several more terrible lectures on ADHD. In one of them, I was informed that America is medicalizing normal childhood mischief and loading anyone who gets worse than a B+ up with Ritalin…

Slate Star CodexScott Alexander

An oldie, this time from 2014 - but an excellent illustration of a statistical phenomenon that is very hard to grasp intuitively (for me, anyway):

How can something be both over and under-diagnosed?