In the early days of the public internet, we had to log on using a dial-up modem.

For anyone too young to remember them:

A dial-up modem is a device that allows your computer to access the internet, using an analog phoneline.

No, not a smartphone, not even a mobile phone. Your grandma’s landline.

Whenever we went on the internet using our computer, that phoneline would temporarily be out of use, and if anyone in the house would pick up the phone, they’d hear ​a strange sound that still haunts the dreams of anyone who ever heard it​.

But there was another difference between dial-up modems and your average modern modem:

They had an incredibly low bitrate. Meaning: It could only process a small amount of data at once.

There’s a popular meme saying the primary purpose of the internet is porn, but that’s definitely not why it was made. Loading a single picture with a dial-up modem could take ages.

Because of that, most websites were incredibly lightweight. Just plain text, with perhaps a small logo or a simple image here and there.

After all, even if you could, there’s no point in making a highly advanced, gorgeous website that nobody can ever load because of bitrate limitations.

Later, when portable digital music players first came out, they had very limited hard drives. We’re talking 128mb. That’s megabytes. Not giga.

At the time, a 3 minute song took up about 30-35 megabytes. So effectively, a digital music player could carry your 3 favorite songs, the end.

The solution to that?

Compression.

What is compression in this context?

When you compress a digital file, you make it smaller by discarding unnecessary information. You basically “shave off” a bunch of the data while trying to still capture enough of the essential parts so that the file still contains the same thing.

The best way I can explain this is: If you take 4K video and you compress it, you have a regular HD video. It lost a bunch of data but it still looks great.

MP3 files were the solution to this. By shaving off lots of parts of the audio data, they could effectively shrink the size of music files to around 5mb. On top of that, you could have lower quality MP3 files which shrank the songs to 1MB or less—by leaving out so much data that every song pretty much sounded like ​that dial-up modem sound I linked to earlier​.

(We call such files “lossy“, because lots of data got lost. You could say that a lossy piece of information is way more efficient, but less effective.)

In the days of limited capacity, compression gave us a choice:

Do you want a high quality, accurate version of your songs, or do you prefer to have access to more songs and have each of them sound like a broken phoneline?

You may wonder “why would anyone choose the broken phoneline option”?

But this came with a lot of freedom: Remember, we were living in times when, if a file was too large, we couldn’t even download it! Sometimes it was broken phoneline or nothing.

The broken phoneline option is what paved the path to your porn today.

Now, before we transition to the main topic, let’s summarize 3 key concepts we’ll be referring to:

Bitrate: The amount of data processed per second.

Compression: Reducing the amount of data in a piece of information, so that it requires less capacity.

Lossy: A piece of information is “lossy” when it got compressed: Part of the information is lacking.

Data Loss In Communication

I think we’ve all experienced instances of lossy communication, simply by communicating through different mediums.

On a phone call, you can’t see each other’s body language or facial expressions. That information gets lost. You might occasionally think the other person’s angry when they are just tired.

On text, this applies even more. Texting is extremely lossy. We tried to introduce emojis to fix this, but it still applies: A lot of miscommunication happens over text because all the nonverbals are gone.

Loss can also happen in translation. When we think of pieces of text being translated, we generally think of it being a 1 to 1 process:

Take English word for “love”, look it up in the translation dictionary, replace it with the Spanish word for “love”.

But that’s not actually how it goes:

In Spanish, there are 2 different translations (amar and querer), which make distinctions that English doesn’t make.

The Greek differentiate between at least 6 different forms of love (agápē, érōs, philía, philautía, storgē, xenía and arguably 2 more).

There isn’t a single phrase in Dutch that literally means “I love you”. There is “ik hou van je”, “ik zie je graag” or— perhaps the closest—“ik heb je lief”, which no one ever says. But there’s no single phrase that means it in the exact same way the English do. There is always a slight difference.

That’s okay. They all come close. So we can loosely translate between them in different languages. But my point is, translation is lossy, it’s a lot of interpretation and reaching for near-synonyms that aren’t exact translations.

The only way to read a text with its literal meaning is to read it in its original language. But lossy translations do get most of the message across.

(Personally, I’d take it one step further and say that language in itself is lossy compression. But for the sake of time, let’s not go there today.)

If so far, today’s musings feel a bit abstract, worry not my friend! Let’s have a look at why all of this is highly applicable and relevant.

Bitrate In Communication

In digital communication, bitrate refers to 3 things:

1) The receiving bitrate (how much data can the system receive per second).

2) The sending bitrate (how much data can the system send per second).

3) The information bitrate (for example: how much data does a file use to convey 1 second of audio).

In interpersonal communication , the same applies:

1) How much data does a person’s system have the capacity to process in this moment? (receive)

2) How much data does a person’s system have the capacity to convey right now? (send)

3) How much data does this form of communication use to convey the message we’re trying to deliver?

A great example of this are letters written in ​legalese​:

A lawyer can take 2 pages of indecipherable language to simply state “sorry mate, you’re gonna have to pay up”.

The way the lawyer delivers it is, technically, much more accurate. But scroungy Jerry from the pub would’ve delivered it more effectively (if the purpose was for you to understand it).

Compression In Communication

Compression in communication is when we deliver a message that’s somewhat less accurate, but still preserves the core of what we’re trying to say, so that the other person can receive the message (instead of it, for example, sounding like legalese to them).

Skilled teachers are particularly great at this.

They may map their teachings to concepts you already understand (through metaphor, allegory, comparison, …) or simply cut out anything that is not relevant for you.

This is one of the reasons people like​ Neil deGrasse Tyson​ find such a large audience. It is not common to find someone who can compress what they teach in a way that preserves a degree of accuracy without alienating the receiver with lots of jargon.

This is also why people love platitudes. Platitudes are lossy as fudge. They’re way more lossy than a good teacher would be, but they manage to compress large amounts of data that would take years to download.

Books like the Bible are remarkable in their compression too. They can convey insane amounts of data with very short phrases. And the reason people argue about interpretation is exactly that: The phrases are compressed.

Compression is really useful when done well (as in the example of Neil deGrasse Tyson). It can also get really strange when too lossy, for example, in the case of the current understanding of dopamine:

On average, people now understand dopamine to be a “pleasure chemical”, some kind of substance that gets activated everytime you do something pleasurable: That scrolling Instagram gives you a dopamine hit, that having sex or eating chocolate gives you a dopamine hit, and that “unearned” dopamine is bad for you.

In reality:

1) Dopamine is one of many neurotransmitters and neurochemicals tied to pleasurable experiences (some others you may have heard of are endorphins, oxytocin, and serotonin—but there are many more).

2) Dopamine isn’t something you get “hit with” that gives you pleasure. It has a complex role in our motivation and reward system, influencing many things, including memory and movement. Dopamine is a good thing.

3) The things we colloquially call “dopamine hits” involve a complex cocktail of neurochemical processes. With social media feeds for example, the role of dopamine isn’t as much related to pleasure as it is to anticipation and desire for potential pleasure. While dopamine drives the craving, cortisol drives the compulsion and anxiety-driven part of the cycle.

…so why are people on major podcasts not explaining these things the way they actually work?

Because the message wouldn’t spread.

Let’s be honest, did all of the above help you make better decisions about social media than if I had just said “it gives you dopamine you didn’t earn and that’s bad”?

Not really. And that’s why it’s better to deliver this message compressed, even if that makes it inaccurate.

Bitrate vs. Processing Capacity

One ​distinction​ we didn’t make yet (and by the way,​ distinctions, which I wrote about here​, are exactly the sort of things we remove when we try to compress a message) is the difference between being able to receive a message and doing something with it.

As long as we have the time to download it, receiving a high-bitrate message is no problem for most of us.

But what can still happen is that our system currently doesn’t have the capacity to deal with so much information at once. And in that case, it doesn’t matter that we can receive the information: Less accurate information will be superior because we can actually use it.

Here’s a beautiful example of that:

How much data we have the capacity to meaningfully use is capped by ​how much of reality our nervous system has the capacity to experience​.

And that’s something which shrinks and expands according to your current ​window of tolerance​.

When you are in a very calm and centered place, your best friend can bring you the bad news of the decade in rich detail and you’ll find a way to process it.

But if you are highly stressed and there are lots of objections flying through the room that could hit you, you don’t want someone telling you “hey btw there’s a flying object there, and another one there, but this one’s a little softer, but that one from the back is pointy so watch out for that one, and…”. No, you just want someone to say “duck and cover your head, I’ll explain later”.

Many things are like this. There’s absolutely no reason to wait until the third date for anything. But it’s a rule that helps a lot of people because then they don’t need to process so much nuanced information at once before they make decisions.

When communicating, it’s vital to tailor our message to the audience’s capacity. That’s why I write to you in depth, but my twitter posts are lossy.

In fact, not respecting the receiver’s capacity is a major cause for conflict. Let me try something, which I’d have hard time explaining publicly, but since you read until here, I trust you’ll have the capacity to see it:

Racism is extremely low-bitrate thinking. It’s reducing the complex reality of humans to “What’s their skin-color? is it the one I dislike? ok, then they’re bad people.”

Racial profiling is when a person (say, a police officer), is more suspicious of a person purely because of their skin color.

Commonly, people would say racial profiling is racism. Because it follows the same, unfair, pattern of judgment.

But is that true? Not necessarily.

They often go hand in hand. A racist cop will definitely engage in racial profiling.

So might a non-racist one, though. If, say, hypothetically, they operate in a neighborhood that is dominated by a specific crime gang which is tied to only 1 ethnicity, and that ethnicity is otherwise underrepresented in that same neighborhood, then racial profiling can be useful tool for them….

…as long as they still treat each person they interrogate with the same degree of respect and empathy, regardless of their skin color. As long as they deeply understand that they are engaging racial profiling because of statistics, but avoid jumping to conclusion about any individual, they could theoretically do so without racism. In fact, it would be no different from saying “pay extra attention to teenage girls in the mall because some of them are prone to shoplift”.

But wait, isn’t that still prejudice?

Yes. And that’s where things get complicated:

It is prejudice. It’s looking at an individual with a degree of suspicion they did nothing to deserve. Racial profiling is extremely lossy compression, and therefore ignorant. That’s what ignorance is, lossy compression.

And yet…if a policeman were to operate in a neighborhood dominated by an Italian mafia, it would be ignorant of the police not to racially profile Italians in that particular neighborhood. It would be lossy compression too.

By racially profiling them, they would probably cause a lot of unpleasantness to a lot of innocent Italian people. By avoiding it, they would fail to protect a lot of innocent people who get hurt by the mafia.

It is extremely challenging to reconcile such things. So most people would rather receive a low-bitrate message (e.g. racial profiling is racism, or racial profiling is bad).

(I honestly don’t know what to do with it either, btw.)

As a result, nearly every way we can look at this topic will be a form of lossy compression. The whole topic is loaded with it because we can’t handle it:

Racism is lossy compression. Racial profiling is lossy compression. Seeing both as the same is lossy compression. Seeing racial profiling as good is lossy compression. Seeing it as bad is lossy too.

Interpreting everything I just shared as “police don’t racially profile because they’re racist, they do it because it’s statistically correct” would be lossy compression too. Because I didn’t say that, and it isn’t true either (systemic racism by cops does exist and is a serious problem—and is often statistically incorrect too! But my example still holds).

The funny thing (to me) is that racial profiling and seeing all people who do it as racist are expressions of the exact same mental process: They judge all people of 1 complex group as the same thing and relate to that thing negatively.

Because of this, it’s incredibly hard to talk about such topics and not be misunderstood. The last thing I’d want is for a racist to think I endorse them. Or for a person to think I will pre-judge them because of how they look. None of that is true. But I chose to tackle an incendiary topic to make this point:

We struggle with nuanced, high bitrate information because our nervous system can’t meaningfully process it.

We experience this seemingly opposing information as tension that needs to be resolved before we can be at peace with reality again.

Additionally, everyone’s vantage point makes them see different things.

If you are an innocent person being treated like a criminal because of your skin, that is absolutely terrible. And I wish nobody would have to go through that.

If you are a cold, impersonal system trying to regulate your well-being to the best of your ability,…then you probably care about individual suffering to the same degree that us, human individuals care about our individual cells.

We don’t see them as names, we see them as numbers.

When we feel stressed, we’ll eat a burger. Even if not all our cells are happy with that.

When we are sick, we take antibiotics, killing off all the good, upstanding citizens in our gut, just to make sure we get rid of the bad apples.

The system’s actions can be wrong to the individual but good for the system, just as our actions can be wrong for our gut bacteria but good for us.

Of course, a healthy system is on average better for the individuals and a healthy body is on average better for its cells. Ideally, nobody would ever have to take antibiotics.

And there are many cases of doctor’s prescribing antibiotics that weren’t needed, just as there are cases of governments implementing mass measures that weren’t needed.

Mostly, it appears that right now, neither our societal system nor most individuals in it have the capacity to look at our concerns in a way that holds both concerns simultaneously—and still see the shortcomings of both.

How to Use This to Improve Your Communication

We just went quite deep there, but there’s a reason I risked using an incendiary topic to illustrate this concept:

I wanted to show how compression operates at a scale where everyone can feel it. Both feel its necessity (a topic being too complex for easy consumption) and the problems it brings (prejudice, ignorance and the challenge with reconciling disconnected pieces of an interconnected web of causality).

So let’s take a breath, shake that off, and bring it back to a practical, day-to-day level:

Bitrate limitations, compressed communication, and data loss in messages impact our daily interactions in the same way as the above example. And they can feed or exacerbate conflict in the same way too.

If, for example, we approach disagreement with a very low-bitrate channel (who is wrong, who is right), there’s no way to reconcile. All we can do is fight.

If we approach it with a slightly higher bitrate (what do we both want, how are we impacting each other, what can we do about that to get along better from here on?), more solutions start to emerge.

The main challenge underneath all of this is developing the capacity to hold multiple seemingly opposing perspectives at once without seeing them as irreconcilable:

To ​increase our own capacity​ to the point that our ​map of reality​ can integrate both those perspectives.

Not only can this help us understand the other person better without demonizing them, it can allow us to communicate in a way that the other person understands.

We can communicate using things that are already on their map, instead of insisting they see ours.

We can compress our message to the smallest level of detail that still preserves all the important information we want them to receive.

We can let them know when we struggle receiving theirs (because it is too high bitrate for us at the moment, for example), and help them tailor their message to what we can receive.

In essence, the way to use these concepts is to take responsibility for our communication challenge.

To understand that communication can only happen when a sender and receiver are both sending their messages with a mutually compatible degree of compression.

When bitrate is near-zero, the best messages we can send (or ask the other to send us) are binary:

Yes/no, bad/good, true/false, healthy/unhealthy.

These are generally inaccurate, but receivable by everyone.

When both people’s capacity is high, a higher bitrate is generally preferred (outside of contexts that demand efficiency, like quick collaborations).

The highest bitrate is so rich and detailed that it looks fluid, that it’s hard to even point out the existence of 2 separate bits at all.

Because we can’t see where one ends and the other begins. In there, everything is connected and everything can coexist. Because it’s all part of the same tapestry.